3.1 The occurrence of PFASs in surface water
A total of 12 PFASs were detected, including 3 PFSAs (C4, C6, and C8), 8 PFCAs (C4-C10, C13) and F-53B (Fig. 2) The concentration of total PFASs (sum of the 12 PFASs, ΣPFASs) in the Caohai Wetland ranged from 1.78 ng/L to 112.21 ng/L (mean = 19.43 ng/L). The detection frequency of each analyte varied widely between sampling sites; PFTriDA was observed at the lowest number of sites (52.38%). The predominant PFASs in the Caohai Wetland were PFBA (0.81–17.38 ng/L, 48.12%), followed by PFOA (0.12–2.10 ng/L, 14.61%) and PFBS (0.00-4.57ng/L, 11.59%). PFOS exhibited a relatively lower concentration (0.00-0.14 ng/L, 0.76%). In the Caohai Wetland, the short-chain PFASs were detected at higher concentrations and detection frequencies compared to long-chain PFASs. This observation can be explained by the significant usage and high solubility of short-chain PFASs(Zhou et al. 2013). In fact, short-chain PFASs, especially PFBA, have become the most important PFASs under the restrictions of PFOA and PFOS.
As a substitute for PFOS, F-53B has been widely used as a mist suppressant in China's electroplating industry for decades. Research shows that F-53B is the most bioaccumulative PFAS, with a high frequency of detection in wastewater, rivers and surface water around electroplating industrial parks(Wang et al. 2013; Wei et al. 2018). Surprisingly, the concentration of F-53B in the Caohai Wetland was low, in the range of 0.14–1.48 ng/L (mean 0.48 ng/L); however, it was detected at all sampling sites. The occurrence and concentration of F-53B was similar to that observed for the Xiaoqing River (mean 0.419 ng/L) and Tangxun Lake (0.149 ng/L) (Shi et al., 2015).
To determine the relative degree of pollution, the concentration of PFBA、PFOA、PFOS in Caohai wetland were compared to other sites investigated by other researchers (Table 5). The concentration of PFBA in Caohai wetland was much lower than that in Tangxun Lake (1820–6280 ng/L), which is near a production base of the fluorochemical industry inWuhan, China, and was similar to Baiyangdian (3.0-14.6 ng/L) (Zhou et al., 2012), but slightly higher than that found by other researchers (Table 5). The concentration of PFOA and PFOS was both lower than other lakes. Overall, the concentration of PFASs in Caohai wetland is at a lower level compared with other lakes.
Table 5
Comparison the concentration (ng/L) of PFBA, PFOA and PFOS with previous study
Study area | PFBA Min-Max (mean) | PFOA Min-Max (mean) | PFOS Min-Max (mean) | References |
Caohai wetland | 0.81–17.38 (5.37) | 0.12–2.10 (1.61) | n.d.-0.14 (0.09) | This study |
Chaohu Lake | 0.31–6.77 (2.04) | 1.32–23.51 (8.62) | n.d.-0.82 (0.08) | (Liu et al. 2015) |
Tangxun Lake | 1820–6280 (4580) | 70.5–1390 (372) | 73.4–1650 (357) | (Zhou et al. 2013) |
Taihu Lake | 0.6–4.06 (0.86) | 2.15–73.9 (28.2) | 0.5–10.5 (3.54) | (Guo C. et al. 2015) |
Poyanghu Lake | n.a. | 0.30–1.89 (1.10) | n.d.-0.71 (0.35) | (Wang et al. 2015) |
Baiyangdian | 3.0-14.6 | 6.8–56.8 | 0.1–17.5 | (Zhou et al. 2012) |
Dianchi | | 3.41–35.4 (10.31) | 1.71–15.1 (7.78) | (Zhang et al. 2012) |
Yangtze River | 0.45–8.38 (1.67) | 1.02–15.8 (8.44) | n.d.-3.93 (0.79) | (Pan et al. 2014) |
Yellow River | 6.35–53.5 (18.8) | 2.01–41.8 (11.5) | 2.65-41.0 (10.8) | (Zhao et al. 2016) |
Pearl River | 0.20–3.34 (1.65) | 0.71–8.70 (3.70) | 0.52-11.0 (3.30) | (Zhang et al. 2013) |
Huaihe River | n.a. | 6.20–47.0 (18.0) | 1.40–25.0 (4.70) | (Yu et al. 2013) |
Hunhe River | n.a. | 1.52–6.32 | n.d.-0.91 | (Jin and Zhu 2016) |
Daliao River | n.a. | 5.81–7.12 | 1.54–1.65 | (Gong et al. 2016) |
Lake Superior | n.a. | 0.178–0.95 | 0.051–0.271 | (De Silva et al. 2011) |
Lake Michigan | n.a. | 1.876–2.573 | 0.326–0.901 | (De Silva et al. 2011) |
Lake Erie | n.a. | 1.564–2.11 | 0.972–1.427 | (De Silva et al. 2011) |
Lake Huron | n.a. | 0.493–3.84 | 0.507–1.192 | (De Silva et al. 2011) |
Lake Ontario | n.a. | 1.706–3.724 | 1.102–5.669 | (De Silva et al. 2011) |
River Rhine upstream | 1.60–2.48 (2.13) | 0.61–3.44 (2.13) | 1.41–6.38 (3.70) | (De Silva et al. 2011) |
River Rhine downstream | 75.5–188 (117) | 2.26–4.07 (3.11) | 3.03–7.34 (4.13) | (Möller et al. 2010) |
“n.d.” means not detected; “n.a.” means not available. |
3.2 The distribution of PFASs in Caohai Wetland
The sampling sites were divided into four sides: the inlet included S1—S3; town side is near urban land and some ports and industrial parks, which included S4 --- S13; the outlet included S14 --- S17; village side is far away from Weining town and is the mainly cultivated land, which included S18 --- S21. The highest concentrations of both PFBA and ΣPFASs occurred at S1 (Dazhong River, ΣPFASs: 22.67 ng/L), followed by S14 (ΣPFASs: 20.11ng/L). The Dazhong River is a tributary of the upper reaches for the Caohai Wetland, and receives domestic sewage and agricultural irrigation water. Therefore, the disorderly discharge of rural domestic sewage to the Dazhong River has been considered the one source of the high concentrations of PFASs in the wetland, especially in the absence of other point sources. Samples from S14 were collected from the only outlet of the Caohai Wetland. The change of water flow rate causes PFASs to be absorbed and intercepted by a large amount of vegetation along with the suspended solids(Tang et al. 2020). And the specific reasons need to be further confirmed.
Weining is the largest region and possesses the highest elevation in Guizhou Province; its main population lives in the Winning town. The economy of this region is mainly based on traditional agriculture, supplemented by the handicraft industry, resulting in the disorderly use of pesticides. In addition, several automotive repair facilities and electroplating factories are scattered throughout the caohai town. Therefore, PFASs input to the wetland adjacent to the town are probably related to emissions from human and agricultural activities. However, it is notable that the concentration of PFASs varied slightly between the sampling sites in the town side region (mean 10.81 ng/L). During field investigations, we found that the largest concentration of artificial purification wetlands were located near S5 and S6. These artificial wetlands may have had a purification effect on the discharged pollutants(De Martis et al. 2016; Zhu et al. 2010). Consequently, despite the serious problems of disorderly discharge from untreated sewage and the dumping of waste, the concentration of PFAS near the town was relatively low. In addition, the concentration of PFASs in the village side region (mean 6.72 ng/L) was lower than town side. Within the village side area, rural domestic sewage is thought to have contributed significantly to PFAS concentrations. In summary, although the concentration of PFASs is lower than other lakes, its pollution problem cannot be ignored, especially in a town that is under urban construction and has inadequate wastewater treatment facilities.
3.3 Source tracing using a PCA-MLR model
The multiple sources of PFASs complicate the identification of PFAS sources in the aquatic environment. Therefore, a PCA-MLR model was applied to analyze the contributions of the 10 main PFASs in the Caohai Wetland from the predominant sources. Three principal components were extracted, accounting for 72.18% of the total variance (Fig. 3).
PC1 explained 41.93% of the total variance with a high loading of PFNA, PFHxA and PFOA. PFNA has often been used as a surfactant to produce perfluoropolymer polyvinylidene fluoride, which is a typical marker for the degradation of FTOHs. PFHxA has been widely used in paper food packaging and water/oil repellent paper coatings as a substitute for PFOA, which can also be produced by the conversion of precursor substances during sewage treatment(Qi et al. 2016; Xiao et al. 2012). Thus, PC1 was interpreted to represent the degradation of FTOHs and food-packaging emission sources.
PC2 contributed 18.15% of the total variance and is characterized by high loadings of PFBS, F-53B and PFHxS. These species are mainly used in textile treatments and metal plating as substitutes for PFOS(Qi et al. 2017; Xie et al. 2013).
PC3 contributed 12.1% of the total variance with high loadings for PFBA and PFHpA, which are indicators of human and livestock excretion and atmospheric precipitation, respectively (Chen et al., 2016). Wang et al. (2018) confirmed that semi-volatile precursors of PFOA are more susceptible to degrade at high altitude sites due to their lower temperatures(Wang et al. 2018). As noted earlier, the Caohai Wetland is located on the Tibetan Plateau where it is subjected to the southeast blowing monsoon. The “Global Fractionation Effect” and “Mountain Cold-trapping Effect” of POPs enables this area to vary in response to vertical changes in ocean currents. Hence(Daly and Wania 2005), PC3 is thought to represent the combined effects of atmospheric precipitation and surface runoff.
3.4 Ecological risk assessment
A total of 22 and 24 acute toxicity values were collected for PFOS and PFOA, respectively. SSD curves were then fit to the data (Fig. 4.a.b). The PFOS curve is on the left of the PFOA curve (Fig. 4.c).
Indicating that ecological risk of PFOS is higher. The HC5 values can be calculated using the SSD curves. Once calculated, the HC5 data can be used to analyze the toxic effects of different pollutants on the same freshwater organism. The acute HC5 for PFOS (1.7 mg/L) is two orders of magnitude lower than that for PFOA (101.5 mg/L). However, the results demonstrated that the toxicity of PFOS is much larger than that of PFOA. Thus, there is a significant difference in the HC5 values between this study and Liu et al(Liu et al. 2015), because different types of toxicity data were selected. Liu et al selected 12 and 10 chronic toxicity data (NOEC) for PFOA and PFOS, respectively. The toxicity data of algae were not included because those data would cause a relatively high uncertainty in the SSD curve. In general, chronic toxicity data are more consistent with actual long-term environmental exposure, which is more meaningful for the derivation of water quality criteria(Xing et al. 2014).
In the use of SSD, the detected concentration of a contaminant is usually used for calculating the PAF (Potential affected fraction) for ecological risk assessment(Xu et al. 2015). The estimated PAF of PFOA and PFOS at every sampling site is shown in Table 6.
The concentration of PFOA (0.12–2.10 ng/L) was one or two orders of magnitude higher than that of PFOS (n.d.-0.14 ng/L), whereas the PAF of the PFOS (which ranged from 8.13E-15 to 1.02E-9) was much higher than that of PFOA (which ranged from 6.31E-129 to 1.27E-84), indicating that the adverse effect of PFOS is more serious than that of PFOA on the ecosystem.
At present, multiple PFASs have been detected in the surface water in Caohai Wetland, which increases the overall risk. Given that an individual target contaminant may represent a relatively low ecological risk to aquatic organisms, it is necessary to conduct a combined ecological assessment(Du et al. 2017). The SSD model can calculate the msPAF (multi-substance potentially affected fractions) based on the toxic mode of action (TMoA) to evaluate the combined ecological risk(He et al. 2014). Owing to the different TMoA of the different contaminants, the msPAF can be calculated via concentration addition or response addition(Gregorio et al. 2013). Considering the unclear TMoA of PFOS and PFOA, the response addition approach was applied using the following equation:
msPAF = 1- (1-PAFpfoa)(1-PAFpfos) (1)
Interestingly, the combined ecological risks (ranging from 1.44E-87 to 1.02E-9) were approximately equal to the risks of PFOS in the detectable sampling sites, although the concentration of PFOS was low. Generally, the spatial distribution of combined ecological risks was related to the distribution of PFOS, which exhibited a relatively higher risk inside of the town side. Risk assessment results for the individual and combined PAF of PFOS and PFOA suggest that current concentrations in the Caohai Wetland are not likely to bring adverse effects to aquatic organisms. However, due to the wide presence and potential biomagnification of PFASs in the food web, PFASs may be a potential threat to precious species in the Caohai Wetland. Therefore, it is essential to monitor PFASs in the Caohai Wetland.