Polycyclic aromatic hydrocarbons (PAHs), which have at least two fused aromatic rings, are ubiquitous in the environment. PAHs are typically generated by anthropogenic activities, such as the incomplete combustion or pyrolysis of fossil fuel and biomass, vehicle emissions, and petroleum leakage (Gan et al. 2009; Yu et al. 2016; Zhang et al. 2017a), and are introduced into the environment by processes, including atmospheric deposition, wastewater discharge, and surface runoff (Hu et al. 2017a; Johnsen et al. 2005; Pheiffer et al. 2018). Due to their distinct physicochemical properties, PAHs interact to different extents with air, water, soil, sediment, and biota (Fan et al. 2019; Montuori et al. 2016; Pérez-Fernández et al. 2015). Low-molecular-weight (LMW; three or fewer rings) PAHs typically predominate in water (Zeng et al. 2018), whereas high-molecular-weight (HMW; 4 or more rings) PAHs tend to be absorbed in soil phase for a longer period of time (Chen et al. 2018; Yang et al. 2017). PAHs exerts serious harmful effects on ecosystems and public health due to their toxicity (carcinogenic, teratogenic, and mutagenic), persistence, long-distance transport, lipophilicity, and capacity for bioaccumulation (Hu et al. 2017b; Kim et al. 2013; Sarria-Villa et al. 2016; Gope et al. 2018). Thus, the United States Environmental Protection Agency (USEPA) lists 16 PAHs as priority pollutants, 7 of which are designated as potential human carcinogens (IARC 2010; Zheng et al. 2016).

The environmental fate and behavior of PAHs are of great concern. Persistent organic pollutants, including PAHs, are detected worldwide in areas of high human activity up to the remote polar regions (Doong and Lin 2004; Wania and Mackay, 1993). Therefore, PAH pollution has become a global issue. Soil is the major terrestrial sink for PAHs (Chen et al. 2004; Wang et al. 2017; Yang et al. 2012). Wild and Jones (1995) estimated that approximately 90% of environmental PAHs was stored in soil in the United Kingdom. Peng et al. (2013) reported that the concentrations of PAHs in soil were significantly correlated with the level of urbanization. Zhang and Chen (2017) analyzed > 6000 topsoil samples from China and found that the PAH concentrations in soil decreased along the urban–suburban–rural gradient. Moreover, the soil physicochemical properties also affect the distribution of PAHs (Heywood et al. 2006; Li et al. 2010; Bucheli et al. 2004). PAHs in soil can be diverted into water by flow scouring. PAHs are not stored in water but can directly enter and accumulate within aquatic organisms, particularly when solubilized (Countway et al. 2003; Zhang et al. 2017b). Humans tend to be indirectly exposed to PAHs in aquatic ecosystems upon ingesting aquatic food, which may represent health risk (Zeng et al. 2018).

Guangzhou is one of the most developed cities in China; it has undergone rapid economic development and urbanization over the last several decades, resulting in increased discharge of organic contaminants and deterioration of environment quality. The majority of the local water resources are polluted, and thus unsuitable for drinking, resulting in an increasingly serious scarcity of clean water (Peng et al. 2017); this has constrained the economic development of Guangzhou. As the longest river in Guangzhou (171 km), the Liuxi River is of great importance to the city and in fact is regarded as the “mother river” of Guangzhou. The Liuxi River basin is the major source of drinking water within the administrative area of Guangzhou and so plays an important role in the exploitation and utilization of water resources in the city. However, contaminants originating from local human activity, and from atmospheric deposition (both dry and wet), have exerted an adverse effect on the environment in the basin. According to monitoring data, nitrogen and phosphorus pollution have resulted in class III–V water quality in the lower reaches of the Liuxi River (GZEMC 2012). To date, few studies have investigated PAHs in the Liuxi River basin. Therefore, information on the species and concentrations of PAHs in the area would facilitate the prevention and control of pollution.

In this study, the main objectives are to (1) evaluate the spatial distribution and composition of PAHs in topsoil and surface water of the Liuxi River basin, (2) identify potential sources of PAHs, and (3) assess the risks posed by PAHs in topsoil and surface water to the environment and public health.

Materials and Methods

Study Area

The Liuxi River basin (23°12′30″–23°57′36″N, 113°10′12–114°2′00″E) is located in the north of Guangzhou, the capital city of Guangdong Province, China (Fig. 1) and has a total area of approximately 2300 km2. The basin is elongated from northeast to southwest and has a high elevation in the northeast and low elevation in the southwest. Accordingly, the Liuxi River flows from northeast to southwest. The upper reaches of the basin are mountainous and dominated by forest; the middle reaches are hilly and encompass orchards, forests, croplands, and grasslands; and the lower reaches are plains (mainly built-up land and cropland). The area has a southern subtropical monsoon climate. The mean annual air temperature is 21.4–21.8 °C, with average maximum and minimum temperatures of 38.7 °C in July and −7 °C in January, respectively. The mean annual precipitation is 1800 mm, 81.3% of which falls between April and September. The basin spans four county-level districts (Conghua, Huadu, Huangpu, and Baiyun), with a total residential population of approximately 5.0 million.

Fig. 1
figure 1

Sampling sites of topsoil and surface water in the Liuxi River Basin

Full size image

Collection and Preparation of Samples

The levels of PAHs in topsoil and surface water may be higher between April and September due to the impact of wet deposition in the study area. Also, the month of April, as the beginning of the flood season, the dilution effect of precipitation on PAHs may be not significant. Thus, a total of 24 topsoil samples (0–20 cm) were collected using precleaned stainless-steel augers in April 2018: six from rural areas (S1–S6); ten from suburban areas (S7–S16); and eight from urban areas (S17–S24) (Fig. 1). Of the 24 samples, 16 were from agricultural land (9 from cropland and 7 from orchards) and 8 were from green spaces (4 each from forests and grasslands). At each site, five subsamples (4 corners and the center) were collected from within a 100 m2 area (10 × 10 m) and mixed thoroughly to generate a composite sample (Jia et al. 2017; Tong et al. 2018). The positions of the sampling sites were recorded using a global positioning system (GPS). The samples were placed in plastic automatic-sealing bags and immediately transported to the laboratory. The topsoil samples were lyophilized at −80 °C, passed through a 60-mesh sieve after removing stones, twigs, and other materials, and stored in dark glass bottles at −4 °C until analysis.

A total of 24 surface water samples were taken from the top layer (0–20 cm) of the waterbody in April 2018. The sampling locations were distributed along the Liuxi River (13 sites, L1–L13) and its main tributaries (11 sites, T1–T11) (Fig. 1), which were located using GPS. The surface water samples were collected into individual 2-L precleaned brown glass bottles with PTFE screw caps. After collection, the samples were immediately transported to the laboratory and stored in desiccators at 4 °C before analysis.

Reagents

The standards of the following 16 PAHs (US EPA priority control): naphthalene (Nap), acenaphthylene (Acy), acenaphthene (Ace), fluorene (Flu), phenanthrene (Phe), anthracene (Ant), fluoranthene (Fla), pyrene (Pyr), benz[a]anthracene (BaA), chrysene (Chry), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[a]pyrene (BaP), indeno[1,2,3-c,d]pyrene (InP), dibenz[a,h]anthracene (DaA), and benzo[g,h,i]perylene (BgP), and five deuterated PAH standards (D8-naphthalene, D10-acenaphthene, D10-phenanthrene, D12-chrysene, and D12-perylene) were used, which were purchased from o2si Smart Solutions (Charleston, SC). The internal standard (hexamethylbenzene) was obtained from Aldrich Chemical (Gillingham, Dorset, UK). The organic solvents used for extraction and purification were purchased from CNW Technologies (Düsseldorf, Germany) and were of pesticide-grade purity. Anhydrous Na2SO4 (Guangzhou Chemicals Inc., Guangzhou, China) was heated at 500 °C for 5 h and stored in a sealed container. Oxides were removed from copper powder using 10% dilute hydrochloric acid, which was then washed using MeOH and n-hexane.

PAH Extraction and Analysis

The methods for topsoil sample extraction and analysis were described previously (Ping et al. 2007; Rachwał et al. 2015). To extract PAHs, 5 g of freeze-dried soil, spiked with five deuterated PAH standards, was mixed with 5 g of anhydrous sodium sulfate and 2 g of copper powder, and extracted in 15 mL of dichloromethane (DCM) at 40 °C for 25 min in an ultrasonic shaker. The extracts ware centrifuged and harvested three times. The resulting three extracts were combined and concentrated to ~ 2 mL. Subsequently, 2 mL of n-hexane was added and the samples were concentrated to 2 mL. Next, the samples were cleaned by being passed over a silica column (CNW) and eluted using 30 mL of n-hexane/DCM (7/3, v/v). The eluent was concentrated to 0.2 mL and transferred to 2-mL capped Teflon-lined vials. Then, internal standards were added and the volume was made up to 1 mL using n-hexane before analysis by gas chromatography-mass spectrometry (GC–MS).

Water samples were filtered through a 0.45-μm organic membrane. To the filtered samples was added 20 mL of methanol, followed by spiking with five deuterated PAH standards. A solid-phase extraction system was used to extract water samples. MIP-PAHs cartridges (500 mg, 6 mL; CNW) were preconditioned with 6 mL of DCM, 6 mL of methanol, and 6 mL of ultrapure water. The water samples were passed through the cartridge at a flow rate of 5 mL/min and eluted using 12 mL of DCM. Next, water was removed from the eluent, which was concentrated to 0.2 mL and transferred to 2-mL, capped Teflon-lined vials. Next, the internal standards were added, and the volume was made up to 1 mL using n-hexane before analysis by GC–MS.

The concentrations of 16 PAHs were quantified using a model 7890 gas chromatograph with a 5975 mass-selective detector (MSD; Agilent Technologies, Santa Clara, CA) equipped with an HP-5MS capillary column (30 m × 0.32 mm × 0.25 μm) and helium carrier gas (1 mL/min). The GC oven temperature program was 50 °C for 1 min, followed by heating at 20 °C/min to 200 °C, and at 10 °C/min to 290 °C, which was maintained for 15 min. The injector temperature was 280 °C. Sample (1 µL) was injected in spitless mode. The MSD was operated in the electron impact mode at 70 eV; the ion source temperature was 300 °C and mass spectra were recorded in selected ion-monitoring mode. PAHs were identified based on their retention times and selection ions and quantified by using internal standards. The concentrations of PAHs were determined at the Instrumental Analysis and Research Center of Sun Yat-sen University.

Determination of Total Organic Carbon Content and pH

The total organic carbon (TOC) content of topsoil was determined using a TOC analyzer (Vario TOC; Elementar, Germany). Topsoil pH was measured in a 2.5: 1 (v/w) water: soil slurry using a Compact pH Meter (Horiba, Kyoto, Japan).

Quality Assurance and Control

A procedural blank, spiked blank, spiked samples, and sample duplicates were analyzed for quality assurance and control. Internal calibration based on a five-point calibration curve was used to quantify the concentrations of the 16 PAHs. Procedural blanks were analyzed after every five samples; no PAH was detected in any of the blanks. The average recovery of 15 standard PAHs was 75.6–93.7% (except for NaP; 57.3%) in spiked water blanks (n = 5) and 78.3–99.2% (except for NaP; 62.4%) in spiked soil blanks (n = 5). The mean recoveries of the five deuterated PAH standards were 58.1–91.1% and 64.9–96.3% in water and soil samples, respectively. The detection limits of the 16 PAHs were 0.15–0.5 ng/L and 0.20–1.2 ng/g in water and soil samples, respectively.

Data Analyses

The concentrations of PAHs were assessed by one-way analysis of variance (ANOVA) and the least-significant difference test; a value of P < 0.05 was taken to indicate statistical significance. Spearman’s correlation analysis was performed to assess the relationships between pairs of variables. Principal component analysis (PCA) was used to identify the major sources of PAHs. Statistical analysis was carried out using SPSS for Windows software (ver. 17.0; SPSS Inc., Chicago, IL) and figures were generated using Origin 9.0 (Origin Lab, Northampton, MA) and ArcGIS 10.1 (Environmental Systems Research Institute, Inc., Redlands, CA) software.

Results and Discussion

Concentrations of PAHs

The levels of 16 PAHs in topsoil and surface water of the Liuxi River basin are listed in Table 1. The detection rates of the 16 PAHs in topsoil were 100.00%, with the exception of BbF (95.83%), BkF (95.83%), DaA (91.67%), and BgP (95.83%). Therefore, PAHs were ubiquitous in topsoil in the study area. The coefficients of variation of BbF and BgP were 100.59% and 100.76%; those of the other 14 PAHs were < 100% (mean, 53.51%), indicating little variation in their concentrations in the topsoil samples. BbF, InP, and BgP were detected in 8.33% of the surface water samples, whereas DaA, BaP, BbF, Chry, and BaA were detected in 29.17–66.67%. The other eight PAHs were detected in all of the surface water samples.

Table 1 Summary data on 16 PAHs in topsoil and surface water of the Liuxi River Basin
Full size table

The total concentration of the 16 PAHs (∑16PAH) in topsoil of the Liuxi River basin ranged from 296.26 to 888.14 ng/g, with a mean of 490.68 ng/g. The concentrations of the seven carcinogenic PAHs (∑7CarPAH) (BaA, Chry, BbF, BkF, BaP, InP, and DaA) ranged from 1.81 to 359.20 ng/g, with a mean of 92.05 ng/g (1.81–40.44% of the ∑16PAH concentration [mean, 18.76%]), indicating that ∑7CarPAH accounts for a small proportion of ∑16PAH. Compared with topsoil in other areas, the mean ∑16PAH concentration in topsoil of the Liuxi River basin was comparable to or higher than that in agricultural or rural topsoil in South Korea (236 ng/g; Nam et al. 2003), Norway (150 ng/g; Nam et al. 2008), and the Huanghuai plain (129.5 ng/g; Yang et al. 2012), Chongming Island (192.8 ng/g; Lu et al. 2012), Taizhou (466.5 ng/g; Tang et al. 2010), and Shangdong (407.4 ng/g; Chai et al. 2017), China. However, the ∑16PAH concentration was markedly lower than that in urban topsoil from London, United Kingdom (18,000 ng/g; Vane et al. 2014), Erbil city, Iraq (2296.1 ng/g; Amjadian et al. 2016), Dhanbad, India (3488 ng/g; Suman et al. 2016), and Ji’nan (23,250 ng/g; Dai et al. 2008), Shanghai (3290 ng/g; Jiang et al. 2009), Beijing (1228 ng/g; Peng et al. 2011), Chengdu (3234 ng/g; Xing et al. 2011), Nanjing (3330 ng/g; Wang et al. 2015), and Lanzhou (2360 ng/g; Jiang et al. 2016), China. Moreover, compared with the ∑16PAH concentrations in more than 6000 topsoil samples from China (Zhang and Chen 2017), those in topsoil were lower than the mean in China (730 ng/g) but higher than that in South China (349 ng/g). The levels of PAHs in soil are not regulated in China (Chen et al. 2018); therefore, to assay the level of contamination, the contamination classification proposed by Maliszewska-Kordybach (1996) was used. This showed that 87.50% of the topsoil samples were slightly contaminated (200–600 ng/g), 12.50% were contaminated (600–1000 ng/g); no samples were uncontaminated (< 200 ng/g) or heavily contaminated (> 1000 ng/g).

The ∑16PAH concentration in surface water ranged from 156.73 to 422.03 ng/L, with a mean of 268.83 ng/L, whereas the ∑7CarPAH concentrations ranged from 1.30 to 70.76 ng/L (not detected in L13), with a mean of 12.16 ng/L (0.55–18.98% of ∑16PAH [mean, 4.16%]). The ∑16PAH concentrations in surface water were compared to those in other areas worldwide, which vary considerably. The mean ∑16PAH concentration in this study was significantly lower than that in surface water from the Gomti River, India (10,330 ng/L; Malik et al. 2011), the Almendares River, Cuba (2512 ng/L; Santana et al. 2015), the Nile River, Egypt (1877.5 ng/L; Mohamed and Mohamed 2010), and the Humen River mouth (652.54 ng/L; Liu et al. 2014), Wuhan section of the Yangtze River (2095.2 ng/L; Feng et al. 2007), Lanzhou reach of the Yellow River (1375 ng/L; Zhao et al. 2015), Taizi River (1818 ng/L; Song et al. 2013), Daliao River estuary (748.76 ng/L; Zheng et al. 2016), and Songhua River (934.62 ng/L; Zhao et al. 2014b), China. In contrast, they were higher than those in the Tiber River and estuary, Italy (90.5 ng/L; Montuori et al. 2016), the Danube River, Hungary (122.6 ng/L, Nagy et al. 2014), and the Yellow River estuary (52.49 ng/L; Li and Li 2017) and rivers in Shanghai (112.92 ng/L; Liu et al. 2016), China. The PAH concentrations in surface water were therefore lower than most values reported previously.

Spatial Distribution of PAHs

The spatial distribution of PAHs in topsoil of the Liuxi River basin is presented in Fig. 2. The rural, suburban, and urban areas were selected, which approximately corresponded to the upper, middle, and lower reaches, respectively, of the river basin. The mean ∑16PAH concentration was 624.78, 427.80, and 468.71 ng/g in the topsoil of the rural, suburban, and urban areas, respectively. The concentrations of ∑16PAH or two- to six-ring PAHs did not significantly vary among the three zones (ANOVA, P > 0.05), indicating that PAHs in topsoil are not influenced by urban activities. To evaluate the effect of agricultural activities on the PAH concentrations in topsoil, the sampling sites were divided into agricultural land (vegetable fields and fruit gardens) and green space (grassland and woodland). No significant difference in the concentrations of ∑16PAH or two- to six-ring PAHs was found between the two land types (ANOVA, P > 0.05), indicating that the PAH concentrations in topsoil are little affected by agricultural activities: e.g., use of pesticides, chemical fertilizers, and membranes. Also, the 16 PAHs in topsoil were evenly distributed, suggesting that the PAHs in the Liuxi River basin are derived mainly from dry and wet deposition of transported pollutants. This result is in agreement with the circumstances in the study area. Because the basin is a crucial source of drinking water, there is at present no point source of pollution. Zhang and Chen (2017) found that the PAH concentrations in China decreased along the urban-suburban-rural gradient, and the mean ∑16PAH concentrations therein were 1243, 525, and 278 ng/g, respectively. Our results were not in agreement with previous reports of the distribution of PAHs in topsoil, possibly due to the absence of serious point sources of pollution in the Liuxi River basin.

Fig. 2
figure 2

Mean concentrations of PAHs in topsoil of the Liuxi River basin in different functional zones (a) and land use types (b). (Same alphabet means no significant differences among concentrations of PAHs from different locations according to ANOVA test, P > 0.05)

Full size image

Five- and six-ring PAHs were not detected in some surface water samples. The mean ∑16PAH concentration was 281.18, 278.97, and 253.52 ng/L in the surface water in rural, suburban, and urban areas, respectively (Fig. 3). The concentrations of ∑16 PAH and two- to four-ring PAHs in the surface water did not differ significantly among the three zones (ANOVA, P > 0.05), similar to the distribution of PAHs in topsoil. No significant difference in the concentration of two-ring PAHs was observed between the water of the Liuxi River and its tributaries (ANOVA, P > 0.05). The concentrations of three- and four-ring PAHs in the water of the tributaries were significantly higher than those in Liuxi River water (ANOVA, P < 0.05). The ∑16PAH concentration was higher in the water of the tributaries (ANOVA, P < 0.05). The differences in PAH concentrations between the water of tributaries and the main stream may be due to the physicochemical properties of HMW PAHs, which are preferentially adsorbed by particulates and sediments in aquatic systems, hampering their migration (Zhu et al. 2008; Adeniji et al. 2019). Thus, the concentrations of two-ring PAHs (lowest MW) were similar in surface water from the tributaries to the main stream of the Liuxi River, but the concentrations of three- and four-ring PAHs in the main stream were lower than those in the tributaries.

Fig. 3
figure 3

Mean concentrations of PAHs in surface water of the Liuxi River basin in different functional zones (a) and in surface water of the Liuxi River and its tributaries (b). (Same alphabet means no significant differences among concentrations of PAHs from different locations according to ANOVA test, P > 0.05)

Full size image

Proportions of individual PAHs

In topsoil, three-ring PAHs accounted for 59.68% (39.39–78.94%) of the total amount of PAHs, followed by two-, four-, and five-ring PAHs (13.00% [5.88–22.49%], 10.51% [6.56–16.55%], and 10.43% [1.06–26.15%], respectively) (Fig. 4). Six-ring PAHs, which accounted for 4.61% (0.36–17.76%) of all PAHs, were the least prevalent. Regarding individual PAHs, Phe predominated (36.76% [25.23–47.52%]), followed by Nap (13.00% [5.88–22.49%]) and Ant (12.13% [7.73–16.45%]). Overall, compared with HMW (4 or more rings) PAHs (mean, 27.32%), the proportions of LMW (3 or fewer rings) PAHs (mean, 72.68%) in topsoil were higher. The contribution of LMW PAHs was similar to that in Xinzhou (60%, Zhao et al. 2014a), but different from that in Beijing (30–34%, Ma et al. 2005), Hangzhou (≤ 20%, Yu et al. 2014), and Nanjing (15%, Wang et al. 2015), China. LMW PAHs in the vapor phase can be transported long distances due to their high saturated vapor pressure. Correspondingly, HMW PAHs, particularly five-ring and six-ring PAHs, which have lower saturated vapor pressures, readily bind to particles and can be transported only a short distance (Zhang and Chen 2017). The PAHs in topsoil originate mainly from atmospheric deposition after distant transport, which explains the higher proportions of LMW compared with HMW PAHs.

Fig. 4
figure 4

Percentage concentration (a) and percentage mean concentration (b) of PAHs in topsoil of the Liuxi River basin

Full size image

An occupation analysis of PAHs in surface water demonstrated that the proportions of PAHs decreased in the following order: two rings (36.59–75.81%; mean, 55.07%), three rings (19.49–54.81%; mean, 35.31%), four rings (3.30–16.51%; mean, 6.62%), five rings (0–18.34%; mean, 2.75%), and six rings (0–3.42%; mean, 0.25%) (Fig. 5). Moreover, Nap predominated (55.07% [36.59–75.81%]), followed by Phe and Flu (19.26% [9.60–37.86%] and 9.21% [5.45–14.20%], respectively). Therefore, LMW PAHs accounted for ~ 90.38% (74.53–96.25%) of the 16 PAHs; this is consistent with the values reported for the Tiber River and estuary, Italy (Montuori et al. 2016), the Danube River and its tributaries, Hungary (Nagy et al. 2014), the Gomti River, India (Malik et al. 2011), the Chenab River, Pakistan (Farooq et al. 2011), the Almendares River, Cuba (Santana et al. 2015), and the Songhua River Basin (Hu et al. 2017a), Liaohe River Basin (Bai et al. 2014), Henan reach of the Yellow River (Sun et al. 2009), Songhua River (Zhao et al. 2014b), drinking water source in Henan Province (Ma et al. 2008), Hun River (Zhang et al. 2013), and Poyang Lake (Zhi et al. 2015), China, likely due to the higher vapor pressure and solubility in water of LMW compared to HMW PAHs (Lan et al. 2016).

Fig. 5
figure 5

Percentage concentration (a) and percentage mean concentration (b) of PAHs in surface water of the Liuxi River basin

Full size image

Correlation Analysis

Generally, after PAHs enter the topsoil, a variety of factors modulate their sorption, sequestration, and level, among which soil organic matter (SOM) is one of the most important (Duan et al. 2015). The concentrations of PAHs are positively correlated with the TOC of soil (Tang et al. 2005; Jiang et al. 2011; Li et al. 2014; Jiang et al. 2016; Cai et al. 2017). Also, no correlation will be established until an equilibrium is reached by continuous input of fresh PAHs (Nam et al. 2008). In this study, the topsoil TOC ranged from 0.17% to 2.59% with a mean of 1.11%. The topsoil TOC was not correlated with the concentrations of PAHs or ∑16PAH (Table 2), suggesting that TOC is not an important determinant of the spatial distribution of PAHs in topsoil. An increase in the soil pH enhances the polarity of SOM, hampering adsorption of PAHs (Li et al. 2014). A decrease in soil pH influences microbial degradation and may affect the levels of PAHs (Bucheli et al. 2004). The concentrations of LMW PAHs are negatively correlated with the soil pH (Chai et al. 2017; Schlautman and Morgan 1993). In this study, topsoil pH ranged from 4.64 to 7.18, with a mean of 6.26, and was not significantly correlated with the concentrations of two- to six-ring PAHs or ∑16PAH; this indicated that pH did not influence the PAH distribution in topsoil.

Table 2 Correlations among PAHs, TOC, and pH in topsoil
Full size table

The correlation coefficients of the PAHs in topsoil are listed in Table 2. The concentrations of three- to six-ring PAHs were strongly positively correlated. However, there was no correlation between the concentration of two-ring PAHs and those of other PAHs. The reduced concentration of two-ring PAHs in different topsoil sites over time was likely difference, because it is the most degradable, volatile, and soluble in water among the 16 PAHs. Therefore, the PAHs detected in topsoil could have common sources.

Identification of the Sources of PAHs

Because of their different physiochemical properties, the concentrations of PAHs change after emission, but the ratios of PAHs of similar molecular weight are relatively stable and can be used to identify sources thereof (Yunker et al. 1996; Ping et al. 2007; Lv et al. 2014). The Ant/(Ant + Phe) and Flu/(Flu + Pyr) concentration ratios are used to distinguish combustion and petroleum sources (Montuori et al. 2016; Yunker et al. 2002). An Ant/(Ant + Phe) ratio of < 0.1 indicates a petroleum source, whereas a ratio of > 0.1 indicates combustion. A Flu/(Flu + Pyr) ratio of < 0.4 is indicative of petroleum, a ratio of 0.4–0.5 implies petroleum combustion, and a ratio of > 0.5 indicates combustion of coal or biomass. The Ant/(Phe + Ant) ratios in 83.33% of the topsoil and surface water samples were > 0.1 (Fig. 6). In addition, the Fla/(Fla + Pyr) ratios in all of the topsoil samples, and all but four of the surface water samples, were > 0.5 (0.4–0.5 in four surface water samples). Therefore, combustion of coal and biomass was the major source of PAHs in topsoil, and emission from combustion and petroleum was the major source in surface water.

Fig. 6
figure 6

Cross plot for the isomeric ratios of selected PAHs in topsoil and surface water of the Liuxi River basin

Full size image

The BaA/(BaA + Chry) and BgP/(BgP + InP) ratios are used to identify the sources of PAHs in soil (Devi et al. 2016; Thiombane et al. 2018). A BaA/(BaA + Chry) ratios of < 0.2 indicate petroleum sources; ratios of 0.2–0.35 imply petroleum or combustion, and ratios of > 0.35 indicate combustion. An InP/(InP + BgP) ratio of < 0.2 is indicative of petroleum; ratios of > 0.5 indicate combustion of coal or biomass, and ratios of 0.2–0.5 are markers of petroleum combustion. The two topsoil samples had BaA/(BaA + Chry) ratios of 0.2–0.35, while those of the other topsoil samples were > 0.35. With a single exception, the InP/(InP + BgP) ratios of the other topsoil samples were > 0.5 (Fig. 6), suggesting that the major source of the PAHs was combustion of coal and biomass.

PCA, a type of multivariate analysis, reduces the number of input variables and extracts a small number of latent factors to analyze relationships among variables (Yuan et al. 2014). To explore the sources of PAHs in topsoil, PCA of the PAH concentrations was performed. PC1, PC2, and PC3 (eigenvalues > 1) were extracted by varimax normalized rotation and explained 91.26% of the total variance in the dataset (Table 3). PC1, which accounted for 45.02% of the total variance, was characterized by high loadings of five- and six-ring PAHs (InP, BbF, BkF, BaP, DaA, and BgP), and moderately high loadings of four-ring PAHs (BaA, Fla, Pyr, and Chry). Among these, InP, DaA, and BgP are associated with traffic emissions (Fraser et al. 1998; Devi et al. 2016; Valavanidis et al. 2006). BbF and BkF are mostly generated by combustion of fossil fuels (Kavouras et al. 2001; Rogge et al. 1993) and BaP by coal combustion and traffic emissions (Li et al. 2017; Wilckel 2000). BaA and Chry are produced during incomplete combustion of diesel and gasoline (Bi et al. 2016; Lee et al. 2005). Fla and Pyr are generally generated by coal combustion (Liu et al. 2018; Luo et al. 2016). Hence, the PAHs in PC1 mainly originated from fossil fuel combustion, particularly vehicular emissions. PC2 accounted for 25.69% of the total variance and had high loadings of Flu, Ace, Phe, and Ant, which are markers of coking sources (Jiang et al. 2016; Simcik et al. 1999; Yuan et al. 2014). PC3 contributed 20.55% of the total variance and had high loading of Acy and moderately high loadings of Chry, BaA, Nap, and Pyr. Acy is indicative of wood combustion (Khalili et al. 1995). Nap is related to incomplete combustion or petroleum sources (Chen et al. 2016; Simcik et al. 1999). Thus, PC3 was identified as a mixed combustion source.

Table 3 Rotated factor loadings for PAHs in topsoil on PCA with varimax
Full size table

Risk Assessment

The carcinogenicity of the PAHs relative to BaP was quantified by estimating the BaP-equivalent concentration (BaPeq). The toxic equivalency factors (TEFs) (Nisbet and LaGoy 1992) were used to estimate the BaPeq of PAHs in topsoil. BaPeq was calculated as BaPeq = Ci × TEFi, where Ci is the concentration of a certain PAH and TEFi is its TEF value. The BaPeqs of PAHs in the topsoil of the Liuxi River basin are shown in Table 4. The total BaPeq of the 16 PAHs was 3.73–105.66 ng/g, with a mean of 30.93 ng/g. The BaPeq value of the seven carcinogenic PAHs was 3.30–104.39 ng/g, with a mean of 30.22 ng/g (97.70% of the total), suggesting that these carcinogenic PAHs are the major contributors to the carcinogenic potential. Among the PAHs, BaP and DaA were the predominant contributors, accounting for 48.39% and 29.75% of the total BaPeq, respectively, followed by BbF (8.17%) and InP (6.88%). The Canadian Soil Quality Guidelines for the protection of the environment and human health suggest a safe BaPeq value of 600 ng/g in soil (CCME, 2010). In this study, the BaPeq value of all of the topsoil samples was < 600 ng/g, indicating that the topsoil does not pose a risk to the environment or public health.

Table 4 BaPeq of PAHs in topsoil of the Liuxi River basin
Full size table

Kalf et al. (1997) devised the risk quotient (RQ) to estimate the environmental risk posed by PAHs in aquatic ecosystems, as well as the negligible concentrations (NCs) and maximum permissible concentrations (MPCs) of PAHs. The risks posed by PAHs were quantified by calculating the RQs. The RQ(NCs) and RQ(MPCs) values were calculated as follows:

$$\begin{aligned} {\text{RQ}}_{{ ( {\text{NCs)}}}} & = {\text{C}}_{\text{PAHs}} /{\text{C}}_{\text{QV(NCs)}} \\ {\text{RQ}}_{{ ( {\text{MPCs)}}}} & = {\text{C}}_{\text{PAHs}} /{\text{C}}_{\text{QV(MPCs)}} \\ \end{aligned}$$

where CPAHs is the concentration of a certain PAH in surface water and CQV(NCs) and CQV(MPCs) are its NC and MPC, respectively. The RQ of Kalf et al. (1997) can be used to assess the ecological risks posed by only 10 individual PAHs; thus, to evaluate the ecological risk of the 16 individual PAHs listed by the USEPA, Cao et al. (2010) inferred the NCs and MPCs of Acy, Ace, Flu, Pyr, BbF, and DaA according to the TEFs (Nisbet and LaGoy 1992) reported previously (Wang et al. 2016; Yan et al. 2016; Zhang et al. 2017a). The quality values of the 16 individual PAHs in surface water are listed in Table 5.

Table 5 RQ(NCs) and RQ(MPCs) of PAHs in surface water of the Liuxi River Basin
Full size table

The RQ∑PAHs(NCs) and RQ∑PAHs(MPCs) values were defined as follows (Cao et al. 2010):

$$\begin{aligned} {\text{RQ}}_{{\sum {\text{PAHs(NCs)}}}} & = \sum\limits_{i = 1}^{16} {{\text{RQ}}_{\text{i(NCs)}} \quad \quad ({\text{RQ}}_{{{\text{i(NCs)}} \ge 1}} )} \\ {\text{RQ}}_{{\sum {\text{PAHs(MPCs)}}}} & = \sum\limits_{i = 1}^{16} {{\text{RQ}}_{\text{i(MPCs)}} \quad ({\text{RQ}}_{{{\text{i(MPCs)}} \ge 1}} )} \\ \end{aligned}$$

The RQ(NCs) and RQ(MPCs) of 16 PAHs with values of ≥ 1 were summed to calculate the RQ∑PAHs(NCs) and RQ∑PAHs(MPCs) values of total PAHs. The ecosystem risk posed by the PAHs is shown in Table 6.

Table 6 Risk classification of individual PAHs and ∑ PAHs
Full size table

In surface water, the mean RQ(NCs) values of 13 PAHs were > 1.0 (except for Chry, BkF, and InP), and those of Nap, Ace, Flu, Phe, Ant, BaA, and BaP were > 10.0 (Table 5). Also, the RQ(MPCs) values of the PAHs were < 1.0, except for BaA in a water sample from T2 (RQ(MPCs) = 1.73), and the mean RQ∑PAHs(NCs) and RQ∑PAHs(MPCs) values in surface water were 158.35 and 0.07, respectively. Therefore, PAHs in surface water of the Liuxi River basin posed a relatively low ecological risk.

Conclusions

The occurrence, sources, and risk levels of 16 PAHs in the topsoil and surface water of the Liuxi River basin were evaluated in this study. The ∑16PAH concentration was 296.26–888.14 ng/g in topsoil and 156.73–422.03 ng/L in surface water, indicating a low level of pollution. The seven carcinogenic PAHs comprised 1.81–40.44% of all PAHs in topsoil and 0.55–18.98% of all PAHs in surface water (not detected in L13). The PAH concentrations in topsoil did not significantly vary and were not influenced by urban or agricultural activities. The spatial distribution of PAHs in surface water was similar to that in topsoil, but the concentrations of three- and four-ring PAHs, and of ∑16PAH, differed significantly between the Liuxi River and its tributaries. LMW PAHs, particularly three-ring PAHs (Phe, Nap, and Ant) predominated in topsoil, whereas Nap, Phe, and Flu predominated in surface water. There was no significant relationship between PAHs and the TOC or pH of topsoil, but strong positive correlations were exhibited in three- to six-ring PAHs. Pyrogenic sources were the major sources of PAHs in topsoil, whereas PAHs in surface water originated from a variety of sources. The PAHs in topsoil were derived primarily from fossil fuel combustion (especially vehicular emissions) and coking. The total BaPeq of the 16 PAHs in topsoil was 3.73–105.66 (mean 30.93) ng/g, indicating a low carcinogenic risk. The RQs indicated that the total PAHs in surface water posed a low ecological risk, while some of the individual PAHs were associated with a moderate level of risk. Therefore, Some concerns and control measures should be conducted to the contamination sources of PAHs of the Liuxi River basin.