1 Introduction

In recent years, perfluoroalkyl substances (e.g., perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS)) have become emerging contaminants due to their persistence and bioaccumulation characteristics (Li et al. 2022a, 2022b; Li et al. 2021a, 2021b; Yu et al. 2021). It is found that PFOA and PFOS have widely detected in various environmental media, such as surface water (3.52–45.54 mg·L−1 for PFOA, and 0.25–3.78 mg·L−1 for PFOS) (Cai et al. 2018; Miao et al. 2017), groundwater (2–9 μg·L−1 for PFOA, and 38–381 μg·L−1 for PFOS) (Bräunig et al. 2017; Martin et al. 2019), and soil (< 0.05–1.54 mg·kg−1 for PFOA, and < 0.05–0.74 mg·kg−1 for PFOS) (Choi et al. 2017; SUN et al. 2017). Extensive research has shown that perfluoroalkyl substances (PFAS) are characterized by high thermal and chemical stability and difficult to biodegrade because of strong C-F bonds (Wang et al. 2015). And PFOA and PFOS could be released into soil through various channels, such as biosolids application, pesticide use, wastewater irrigation, and then be extracted by plants (Xiang et al. 2020a; Yu et al. 2021). It was reported that the introduction of PFOA and PFOS into soils had considerable adverse effects on living organisms owing to their chemical stability, bioaccumulation properties (Li et al. 2021a, 2021b). The bio-accumulation of PFOA and PFOS in vegetable crops has caused toxic effects on the health of human (Xiang et al. 2020b). Several previous studies have confirmed that PFOA and PFOS could cause various health problems (Knutsen et al. 2018), including hepatotoxicity (SUN et al. 2017), kidney failure (Kumar et al. 2017), immunotoxicity (Li et al. 2021a, 2021b), and even oncogenic potential (Chen et al. 2020). By now, most of previous studies concentrated on the distribution and behavior of PFOA and PFOS in the environment (Bräunig et al. 2017; Qian et al. 2017; Wang et al. 2012), and its potential adverse effects on plants (Xiang et al. 2020a; Zhou et al. 2019) and animal health (Chen et al. 2013; Han and Fang 2010; Huang et al. 2015). Few studies focused on the impacts of PFOA and PFOS on soil ecosystems and their effects on soil carbon emission (Bao et al. 2018; Groffen et al. 2019; Lv et al. 2022). Consequently, understanding the dynamics of soil mineralization of soil organic carbon (SOC) with PFOA and PFOS contamination is critical for predicting the future of carbon cycle under widespread PFAS contamination. Based on chemical stability and resistance to biodegradation (Wang et al. 2015), PFOA and PFOS could exist in soil for a long time. In recent years, most of the mechanistic studies on soil PFOA and FPOS have focused on short-term effects, rather than long-term exploration. Therefore, it is essential to understand the dynamics of soil physicochemical processes and biological properties under the long-term PFOA and PFOS contamination.

As a natural and heterogeneous medium, soil was subjected to multiple stresses, such as natural and anthropogenic contamination. Under stress conditions, the join effect can be synergistic and/or antagonistic, which usually results in the dynamic variations in soil carbon turnover. It has been found that terrestrial ecosystems store a mass of soil organic carbon (SOC) in the epigeosphere, which plays a critical role in regulating atmosphere-soil carbon turnover (Zeng et al. 2022). If the SOC is largely released back into the atmosphere, it could accelerate climate change with catastrophic consequences for life on global geographic scale (Wang et al. 2019). The wide distribution of PFOS causes changes in soil properties (e.g., microbial biomass, soil respiration, metabolic efficiency), which could further affect soil carbon turnover directly or indirectly (Zeng et al. 2022). It was reported that microorganisms enhanced soil carbon emission by increasing the degradation of SOC (Wieder et al. 2013). In addition, soil respiration is considered to be the main pathway through which soil organic carbon emission is released into atmosphere (Schlesinger and Andrews 2000). However, microbial biomass could also have potential positive correlation with SOC through the production of microbial residual carbon (Liang et al. 2017). And it was reported that higher carbon use efficiency of microorganism could lead to higher microbial biomass and therefore could increase organic carbon fixation via microbial residual carbon accumulation in soil (Doetterl et al. 2018; Wang et al. 2021). It is indispensable to understand the influence of PFOA and PFOS on soil microorganisms, enzymes and organic carbon mineralization.

In this paper, a laboratory-scale long-term culture experiment was performed to explore the micro-ecological effects of PFOA and PFOS on contaminated soils. The aims of this article are as follows: (1) to demonstrate the effects of PFOA and PFOS on the dynamics of soil organic carbon emission; (2) to clarify the relationship among the soil properties, enzymatic activity, microbial diversity, PFOS and PFOA introduction; (3) to demonstrate the effects of PFOA and PFOS on the soil ecosystems and carbon turnover. This study aims to gain insights into the influence of PFOA and PFOS on soil carbon turnover and provide a theoretical basis for the remediation and sustainable development of FPOA and PFOS contaminated soil.

2 Materials and methods

2.1 Soil preparation

The soil sample was collected from the suburbs of Wuhan, Hubei Province, China. The soil was air dried, crushed, and sieved using a 2-mm stainless steel mesh after collection. Then, PFOA/PFOS (Aladdin, Shanghai, China) aqueous solution was used to adjust the soil PFOA/PFOS concentration to 0, 0.1, 1.0, and 10 mg·kg−1, respectively. PFOA/PFOS-contaminated soils were thoroughly mixed and incubated for 180 d with 60% of the maximum water content. Triplicate subsamples were processed for each treatment. After incubation was completed, soil samples were collected and stored at −20 °C for analysis of physiochemical properties, enzymes and high-throughput sequencing. In addition, soil organic carbon mineralization rates were detected at different times.

2.2 Analysis of soil physical and chemical properties

The soil pH was measured with a pH meter (FE20, Shanghai Mettler Toledo Company, Shanghai, China) after mixing the soil and water at a 1:2.5 ratio. Soil total carbon (TC) was calculated by total carbon analyzer (multi N/C 3100, Jena, Germany). SOCCr was quantification using wet oxidation with potassium dichromate (Li et al. 2022a, 2022b). Soil enzyme activities (including urease, sucrase, amylase, neutral phosphatase, nitrate reductase, and nitrite reductase) were measured by adapting a reported procedure (Li et al. 2008). Testing procedures are provided in supporting information Text S1-Text S7. Details of determination of soil organic carbon mineralization rate procedures are presented in Text S8. The extraction and characterization for dissolved organic carbon (DOC) are given in the Text S9. The PFOA/PFOS content in the soil was assessed by using ultrasound-assisted extraction method with methanol as an extractant, followed by a purification with SPE (Agela, 150 mg, 6 mL) purification (Text S10). In addition, soil bacterial and fungal communities were evaluated by amplifying the V3V4(a) region of bacterial 16S rRNA and the ITS1 region of the fungal ITS rRNA gene. This process was carried out by Shanghai Personalbio Technology Co., Ltd. (Shanghai, China). Testing procedures are listed in the supporting information Text S11.

2.3 Data analysis

The experimental data were collected in Excel 2019 (Microsoft) and One-way ANOVA of soil physical and chemical properties was processed in SPSS 25 (US, Chicago, IBM company). Origin 9.5 (OriginLab Company) was used to plot the graphs.

3 Results and discussion

3.1 Soil organic carbon mineralization rate

The introduction of PFOA and PFOS directly affected the soil organic carbon mineralization rates by priming effect (PE). In this study, it was observed that the soil organic carbon mineralization rate was significantly depressed in PFOA and PFOS contaminated soils after a long-time incubation (≥180 d) (Fig. 1c). Whereas the short-term incubation experiments showed that the significant promotion was observed in the soil organic carbon mineralization (Fig. 1a). The newly added PFOA and PFOS significantly enhanced soil organic carbon mineralization rate (Fig. 1a). A possible explanation for this might be that PFOA and PFOS pollution increased the demand of microorganisms for carbon and energy, which accelerated soil organic carbon mineralization. It was found that the PFOA increased the number of bacterial and fungal OTUs (Fig. 5a), and PFOS caused the increases of fungal OTUs (Fig. 5b), suggesting that PFOA and PFOS could enhance soil microbial abundance. This finding is consistent with the previous study which reported that PFOS can increase soil bacterial abundance (Qiao et al. 2018). This could explain that short-term cultured PFOA and PFOS contaminated soil showed higher mineralization rate of soil organic carbon (Fig. 1a).

Fig. 1
figure 1

Change in the soil organic carbon mineralization rate. a PFOA and PFOS were added for 30 d; (b): PFOA and PFOS were added for 90 d; (c): PFOA and PFOS were added for 180 d; (d): PFOA and PFOS were added for 360 d. Different letters on the bar indicate significant differences between the treatment groups (p < 0.05)

Full size image

As the incubation continued, high levels (10 mg·kg−1) of PFOA exhibited an inhibitory effect on soil respiration rate (Fig. 1b). Different from PFOA treatment, the mineralization of SOC in soil decreased in PFOS treated groups (Fig. 1b). In addition, after a long period of pollution (≥180 d), the mineralization of SOC in both soils with different pollutants (PFOA/PFOS) significantly decreased as the PFOA and PFOS content increased (Fig. 1c and d). Compared with the CK (without PFOA and PFOS), the groups treated with 0.1 mg·kg−1–10 mg·kg−1 PFOA presented 58.30%–60.72% lower mineralization rates of SOC, and 62.77%–65.24% lower mineralization rates were observed in the PFOS treated groups (Fig. 1c). Results of soil DOC and hexose could explain this phenomenon (Fig. 3a and b). It was found that the pollution of PFOA and PFOS had slight influence on TOC and SOC (Fig. S1), whereas significantly decreased the content of DOC in soil (Fig. 2). For further characterization of DOC, carbon content (DOC), hexose, protein, nitrogen (DTN), and dissolved protein were determined in this work. It was found that DOC decreased as the PFOA and PFOS concentration increased. With the 0.1, 1.0, and 10 mg·kg−1 PFOA pollution, the DOC decreased by 8.34%, 21.10%, and 22.19% compared with the 0 mg·kg−1 PFOA treatment. The similar results were obtained for PFOS polluted soils, where the DOC content decreased by 10.78%, 13.00%, and 17.49% for 0.1, 1.0, and 10.0 mg·kg−1 PFOS treatments, respectively (Fig. 2a).

Fig. 2
figure 2

Characterization of dissolved organic carbon after incubation for 180 days. a dissolved organic carbon content; (b): hexose content in DOC; (c): dissolved protein in DOC; (d): dissolved total nitrogen. Different letters on the bar indicate significant differences between the treatment groups (p < 0.05)

Full size image

In addition, the similar trend to DOC was observed for hexose contents. PFOA significantly reduced the hexose in soil. The hexose contents decreased by 22.98%, 34.61%, and 30.11% compared with the CK after applying 0.1, 1.0, and 10.0 mg·kg−1 of PFOA, respectively (Fig. 2b). At the same time, PFOS treatments resulted in a significant decrease in hexose contents in soils compared with the CK (Fig. 2b). For the DOC in soils, there was a significant negative correlation between its content and soil PFOA (− 0.640*) and PFOS (− 0.683*) (Table S2). This result could be explained by the fact that PFOA and PFOS contamination resulted in an increase in soil microbial abundance (Fig. S2 and Fig.5), which increased the consumption of carbon sources by microorganisms. Figure 2c indicates that PFOA and PFOS had no significant effect on soil protein content. Overall, the decrease of DOC (including hexose) contents in PFOA and PFOS polluted soils (Fig. 3b and c) further limited the activity of soil microorganisms and ultimately inhibited soil organic carbon mineralization (Fig. 1c).

Fig. 3
figure 3

a Growth rate of soil substrate-induced respiration rate (SIR) after glucose supplement for 3 d; (b): the relationship between soil substrate-induced respiration rate and PFOA/PFOS concentration. Different letters on the bar indicate significant differences between the treatment groups (p < 0.05)

Full size image

It was also found that soil organic carbon mineralization rate was significantly positively associated with DOC (0.741**) and hexose (0.884**) (Fig. S4). DOC is the most mobile and bioavailable substance in soil (Adam Zsolnay 2003), which accounts for less than 0.25% of the SOC content (Temminghoff et al. 1997). In this study, DOC accounts for 1.9%–2.4% of SOC in the soil used in this experiment, and a laboratory-scale cultivation experiment was completed without the addition of organic fertilizers. And withered leaves and residual roots in the soil were screened and removed before soil samples were used in this study. Thereby, the introduction of exogenous DOC from atmospheric deposition or human activities and plant residues was excluded, and thus the DOC and hexose were continuously consumed by the microorganisms in soils without being supplemented. If organic matter is not supplemented, the soil carbon mineralization rate decreases with the consumption of most soil decomposable organic matter such as DOC (Carter and Gregorich 2022). In this experiment, the mineralization rate decreased from 0.24 ± 0.01 mg·d−1·g−1 at 90 days to 0.12 ± 0.01 mg·d−1·g−1 at 180 days (Fig. 1) due to DOC shortage.

It was found that with the 0.1, 1.0, and 10.0 mg·kg−1 PFOA applications, the DTN content decreased by 4.65%, 10.29%, and 6.09% compared with the CK, respectively. While the increase of PFOS content from 0.1–10.0 mg·kg−1, DTN decreased by 6.54%–10.47% compared with the CK (Fig. 2d). Interestingly, no significant differences were found between DTN and protein in DOC, while a significant correlation was found between DTN and hexose (0.888**) (Fig. S4). This result could be attributed to the presence of amino sugars in soils, such as glucosamine, galactosamine, mannosamine, and muramic acid (Carter and Gregorich 2022). In addition, it was observed that PFOA and PFOS had no significant influence on soil pH. There was a significant negative correlation between pH and DOC, hexose, carbon mineralization rate, and SIR (Fig. S4 and Table S2), and one possible explanation for these results could be the release of carbon dioxide during the mineralization of these in DOC (Carter and Gregorich 2022). Further analysis showed that PFOA and PFOS had no significant impact on SOC and protein while reducing DTN content and slightly enhancing specific UV absorbance of DOC. Specific UV absorbance can be used to evaluate the concentration of aromatic compounds in DOC (Chin et al. 1994; Novak et al. 1992; Traina et al. 1990). Studies have found that the specific UV absorbance was positively correlated with the aromatic hydrocarbon polymer (XAD-8) (Dilling and Kaiser 2002; Kalbitz et al. 2003), and phenol content in DOC (Carter and Gregorich 2022), while aromatic hydrocarbons are the relatively stable part of DOC compare to hexoses (Kalbitz et al. 2003). As mentioned earlier, the introduction of PFOA and PFOS promoted the mineralization of easily decomposable organic carbon in the soil, thus increasing the proportion of refractory substances such as aromatic compounds in DOC (Fig. S5).

3.2 Changes in soil organic carbon mineralization rate after glucose supplement

It was observed that hexose had a significant response to PFOA and PFOS contamination in soil (Fig. 2). Nonetheless, it was still unknown that whether PFOA and/or PFOS accelerated the consumption of hexose by microbial metabolism. To further reveal this phenomenon, glucose was supplemented to explore the role of sugars in soil organic carbon mineralization in soil polluted with PFOA and PFOS. Soil substrate-induced respiration rate (SIR) of various treatments was detected and the growth rate of SIR in various treatments was calculated using the corresponding soil without added sugar as a control (growth rate = (m2-m1)/m1*100, where m1 and m2 are the SIR before and after glucose supplement). It was obvious that the growth rate of SIR was increased after glucose was supplemented in the PFOA and PFOS treated groups for 3 days (Fig. 3a). Compared with the CK, the growth rate of SIR increased by 6.10%, 45.26% and 81.62% after applying 0.1, 1.0, and 10 mg·kg−1 of PFOA, respectively. Similarly, the growth rates of SIR increased by 41.50%, 59.50% and 100.21% for the 0.1, 1.0 and 10 mg·kg−1 of PFOS treatments compare with the CK (Fig. 3a). The most interesting aspect of Fig. 3b is that the supplement of glucose significantly relieved the inhibition of soil organic carbon mineralization by PFOA and PFOS (Fig. 1c). It was found that the growth rate of soil organic carbon mineralization showed a strong correlation with the content of PFOA and PFOS (Fig. 3b). The results indicated that PFOA and PFOS both presented notable priming effects on hexose in soil. The analysis showed that soil SIR was negatively correlated with the soil PFOA (− 0.615*) and PFOS (− 0.547) content in the long-term process. This result suggested that microbial activity was significantly reduced under long-term PFOA/PFOS contamination due to the lack of carbohydrate (Table S2). It was obvious that SIR was decreased with the addition of PFOA and PFOS (Fig. S6). Interestingly, 3 days after the soil was supplemented with glucose, the SIR increased rapidly in PFOA and PFOS contaminated soils (Fig. 3). Overall, these results suggested that PFOA and PFOS increased soil microbial activity and accelerated the consumption of DOC (including hexose) in a short-term period (Figs. 1a, 2, and 3). It was reported that PFOS briefly enhanced soil respiration rates (Lv et al. 2022). These could explain why the PFOA and PFOS contamination enhanced the soil carbon mineralization in the short-term incubation (Fig. 1a).

3.3 Effects of PFOA and PFOS on soil enzymes

Understanding soil quality dynamics is essential for managing soil resources and achieving sustainable soil development. The presence of soil microorganisms and enzymes could accelerate the cycling of nutrients in the soil (Burns et al. 2013), promote plant growth, and directly reflect the characteristics of the soil ecosystem (Tang et al. 2019). Therefore, it is indispensable to understand the effects of contamination pressure on soil microorganisms and enzymes. Previous studies have shown that some pollutants, such as trace metals (Khan et al. 2010), pesticides (Yue et al. 2016), and PAHs (polycyclic aromatic hydrocarbons) (Shen et al. 2006) would promote or inhibit soil urease, phosphatase, and sucrase, etc. In this study, sucrase, amylase, urease, nitrate reductase, nitrite reductase, and neutral phosphatase were selected and detected according to previous literature for a comprehensive evaluation of soil health (Li et al. 2008). Figure 4a reveals that PFOA and PFOS contamination activated the soil sucrase. Sucrase shown significantly increased in soils treated with PFOA and PFOS (p < 0.05). The S-SC of the PFOA treatment increased by 33.79%, 66.27%, and 53.73% compared with the CK for the 0.1, 1.0, and 10.0 mg·kg−1 PFOA addition, respectively. Likewise, the pollution of various concentrations of PFOS resulted in elevated S-SC. As the results shown in Fig. 4a, the S-SC decreased by 41.92%, 62.03%, and 54.29% compared with the CK for the 0.1, 1.0, and 10.0 mg·kg−1 FPOS treatments, respectively (Fig. 4a). This outcome is contrary to that of previous literature that PFOS inhibited the soil sucrase activity (Qiao et al. 2018). Further, the sucrase activity was significantly negatively correlated with the soil DOC (− 0.745**) and hexose (− 0.759**) contents (Fig. S4). A possible explanation for this might be that the demand for sugars by soil microorganisms accelerated the hydrolysis of sucrose by sucrase. No significant reduction or increase in soil amylase (S-AL) in PFOA and PFOS polluted soils were observed compared with the CK (Fig. 4b).

Fig. 4
figure 4

Changes in soil enzymes after incubation for 180 days. a sucrase; (b): amylase; (c): urease; (d): nitrate reductase; (e): nitrite reductase; (f): neutral phosphatase. Different letters on the bar indicate significant differences between the treatment groups (p < 0.05)

Full size image

Urease is an amidase that enzymatically promotes the breakdown of peptide bonds in organic molecules (Li et al. 2008), which promotes the conversion of organic nitrogen to inorganic nitrogen. The analysis of soil urease (S-UE) showed that PFOA and PFOS had a significant adverse impact on S-UE even in the low levels (Fig. 4c). This finding was also reported by previous study that urease activity was inhibited by PFOA (Lv et al. 2022). After applying 0.1, 1.0, and 10.0 mg·kg−1 of PFOA, the S-UE decreased by 51.33%, 49.24% and 65.22% compared with the CK, respectively. In particular, the same concentration of PFOS treatment significantly reduced the S-UE by 43.86%, 39.70%, and 35.18% (Fig. 4c). The results of nitrate reductase (S-NR) were shown in Fig. 4d. There was a rapid increase in the soil nitrate reductase (S-NR) with the contamination of PFOA and PFOS (Fig. 4d), indicating that PFOA and PFOS contamination increased soil denitrification potential and increased the contribution of nitrogen emissions. With the application of 0.1, 1.0, and 10 mg·kg−1 PFOA, the S-NR increased by 63.00%, 67.66%, and 95.62% compared with the CK. Similarly, PFOS pollution resulted in a significant increase in S-NR (59.12%–95.58%) (Fig. 4d). Correlation analysis showed that the S-NR was significantly negatively correlated with the soil DTN (− 0.623**) and hexose (− 0.735**). A possible explanation was that DTN and carbohydrate were consumed as substrate and energy sources, respectively, in the denitrification process. Soil nitrite reductase (S-NiR) activity increased at first and then decreased as the PFOA concentration increased, and the lowest S-NIR activity (208.89 ± 28.81) was obtained for the treatment with 0.1 mg·kg−1 PFOA (Fig. 4e). No significant differences for S-NIR were found in the PFOS treatment soils with various concentrations (Fig. 4e). As for soil neutral phosphatase (S-NP), the contamination of PFOA and PFOS significantly enhanced its activity. After applying 0.1–10.0 mg·kg−1 PFOA, the increase in S-NP ranged from 58.71%–125.47%. Obviously, PFOS also had a dramatic effect on the S-NP: S-NP increased by 156.72%, 154.83%, and 247.16% compared with the CK for the 0.1, 1.0, and 10.0 mg·kg−1 PFOS treatments, respectively (Fig. 4f). It was found that S-NP showed a significant negative correlation with DTN, hexose, and protein in PFOA and PFOS contaminated soils, with coefficients of − 0.579**, − 0.475*, and − 0.561**, respectively (Fig. S4). In summary, these results showed that soil PFOA and PFOS contamination limited the conversion of soil organic nitrogen to inorganic nitrogen while increasing the potential for soil organic carbon mineralization and nitrogen emission.

3.4 Diversity and structure composition of microbial communities

Soil microbes affect the formation and stabilization of the soil structure (Sun et al. 2016) and participate in important ecological processes such as soil nutrient cycling and organic matter degradation (Ludwig et al. 2015; Uroz et al. 2009). Thus, the dynamics of microbial activity can effectively reflect the health of polluted soils (Lessard et al. 2014). Therefore, it is important to assess the effect of PFOA and PFOS on soil microorganism diversity. Studies have reported that the pollution of PFOA and PFOS has negative effects on soil microorganisms (Lu et al. 2020). In this study Alpha diversity indexes of bacteria and fungi in soils including Chao 1, Shannon, Observed_species and Simpson are shown in Fig. S2. For alpha diversity indexes (Chao 1, Shannon, Observed_species and Simpson), it was observed that PFOA and PFOS sightly increased all the indices of bactera and fungi in soils. Nevertheless, no significant difference (p < 0.05) was observed among various soils except for the Observed_species between CK and PFOA treatment (Fig. S2).

The dominant bacterial phyla across all treatments were Actinobacteria, Proteobacteria, Bacteroidetes, Acidobacteria, Gemmatimonadetes, Firmicutes, Chloroflexi, Planctomycetes, WPS-2, and Verrucomicrobia, together accounting for a large proportion (> 99%). The proportion of Actinobacteria in PFOA and PFOS treated soil was lower than that in CK and decreased by 14.20% and 12.22% in PFOA and PFOS treatments, respectively. It was found that PFOA and PFOS significantly increased the Proteobacteria abundance by 16.26% and 12.87%, while the abundance of Bacteroidetes decreased by 15.32% and 14.13% for PFOA and PFOS polluted soil compared with CK. Contamination with PFOA and PFOS significantly increased the abundance of Acidobacteria by 6.10% and 8.39% in soils, and similarly increased the soil Gemmatimonadetes by 8.8% and 6.04% (Fig. S3a). The results from Fig. S3b show that the DOCinant fungal classes (> 99%) in soils were Ascomycota, Basidiomycota, Mucoromycota, and Mortierellomycota. PFOA and PFOS pollution significantly reduced the relative abundance of soil Mucoromycota and increased soil Mortierellomycota relative abundance (Fig. S3b).

The relative abundance of top 20 bacterial and fungi genera are shown in Fig. 5. The significant changes were caused by the exposure of PFOA and PFOS (Fig. 5). It was observed that PFOA significantly enriched some of bacterial genera, including Bacillus, Streptosporangium, Rhodanobacter, Dyella, while abated most of bacterial genera, such as Nonomuraea, Pseudosphingobacterium, Kribbella, Actinomadura, JG30-KF-CM45, Cohnella, Amycolatopsis, Parapedobacter, Micromonospora, Streptomyces. For PFOS treatment, Paenibacillus, Nitrolancea, Bryobacter, Pedobacter, Arthrobacter were significantly enriched, and Actinomadura, JG30-KF-CM45, Cohnella, Amycolatopsis, Parapedobacter, Micromonospora, Streptomyces were markedly decreased (Fig. 5c). In addition, for fungi genera, PFOA enriched Purpureocillium, Tausonia, Humicola, Aspergillus, Penicillium, Pseudogymnoascus, Trichoderma, PFOS enriched Mucor, Rhizopus, Cephalotrichum, Oidiodendron, while PFOA and PFOS contamination decreased soil Solicoccozyma, Mycosphaerella, Paraboeremia, Fusarium, Cladosporium (Fig. 5d). Heatmap (Fig. S3c) indicated the relationship between soil properties and species (phylum) by Pearson correlation. Actinobacteria had significantly positive correction with DOC, hexose, SOC-MlR, and SIR (Fig. S3c). This result could be explained by the fact that the degradation of soil organic matter by Actinobacteria increased DOC content. It was reported that the Actinobacteria phylum played a key role in terrestrial ecosystem function. Actinobacteria phylum was believed to contribute to the global carbon cycle by decomposing soil organic matter, enhancing plant productivity (Araujo et al. 2020). Thus, the decrease of actinomycetes in PFOA and PFOS contaminated soils (Fig. S3a) could be another reason for the decrease of DOC content (Fig. 3b). Interestingly, PFOA and PFOS resulted in a significant increase of the phyla Proteobacteria, Acidobacteria and Gemmatimonadetes (Fig. S3a). These phyla all showed significant positive correlations with soil pH, S-SC, S-NR, and S-NP, and significant negative correlations with soil DOC, DTN, hexose, S-UE, S-NIR, SOC-MlR, and SIR (Fig. S3c). It was reported that Acidobacteria could play an important ecological role in the degradation of polysaccharides produced from plant and fungal (Lladó et al. 2016). The Ascomycota phylum is an important driver of carbon and nitrogen cycling in terrestrial ecosystems. These fungi play a role in soil stability, plant residue decomposition, and endophytic interactions with plants (Challacombe et al. 2019). In this work, PFOA and PFOS had no significant adverse impact on soil Ascomycota, while decreasing the abundance of Basidiomycota (Fig. S3b).

Fig. 5
figure 5

The ASV/OTU Venn diagram and relative abundance of the top 20 bacterial and fungi genera in the soil exposed to PFOA and PFOS. a and (c): bacteria; (b) and (d): fungi. The green color means positive correlation, and brown color means negative correlation

Full size image

3.5 Microbial communities with statistically significant differences

In addition to alpha and beta diversity, another main purpose of comparing microbial communities is to identify specific communities in soils (Huhe Chen et al. 2017). As shown in Fig. 6, 55 bacterial and 47 fungal clades exhibited significant enrichment among the soil samples from phylum to genus (LDA threshold: 2). It was observed that there were 45 clades including two phyla (Bacteroidetes and Chloroflexi), two classes (Bacteroidia and Chloroflexia), six orders (Micromonosporales, Propionibacteriales, Sphingobacteriales, Thermomicrobiales, Rhodobacterales, Oceanospirillales), eight families (Micrococcaceae, Micromonosporaceae, Nocardioidaceae, Propionibacteriaceae, Sphingobacteriaceae, Rhizobiaceae, Rhodobacteraceae, Pseudohongiellaceae), and 22 genera (Brevibacterium, Isoptericola, Micromonospora, Brevibacterium, Isoptericola, Micromonospora, Micromonosporaceae, Actinopolymorpha, Kribbella, Nocardioides, Marinilutecoccus, Chitinophaga, Olivibacter, Parapedobacter, Pseudosphingobacterium, Sphingobacterium, Aminobacter, Aureimonas, Pseudaminobacter, Paracoccus, Achromobacter, Advenella, Bordetella, Variovorax, Pseudohongiella) (Fig. 6a). And in PFOA contaminated soils, 43 clades were significantly enriched, namely, Rokubacteria, Gemmatimonadetes, and Proteobacteria (phylum), Thermoleophilia, Gemmatimonadetes, Longimicrobia, Deltaproteobacteria, and Gammaproteobacteria (class), Solirubrobacterales, Cytophagales, Gemmatimonadales, Longimicrobiales, Micropepsales, Reyranellales, and Xanthomonadales (order), Microscillaceae, Rhodospirillaceae, Acidothermaceae, 67_14, Solirubrobacteraceae, Flavobacteriaceae, Gemmatimonadaceae, Longimicrobiaceae, Reyranellaceae, Xanthobacteraceae, Sandaracinaceae, and Rhodanobacteraceae (family), Acidothermus, Microtetraspora, 67_14, Conexibacter, UTBCD1, Flavobacterium, ZOR0006, Gemmatimonas, Longimicrobiaceae, Micropepsis, Reyranella, Afipia, Sandaracinus, amlibacter, and Dyella (genus). For the PFOS treatment, fewer bacteria were significantly enriched (Fig. 6a and b), and there were only 16 clades significantly enriched, including one phylum (Acidobacteria), one class (Acidobacteriia), two orders (Solibacterales, and Acetobacterales), four families (Cyclobacteriaceae, Acetobacteraceae, Dermacoccaceae, and Promicromonosporaceae) and eight genera (Bryobacter, Flexivirga, Arthrobacter, Sinomonas, Promicromonospora, Pedobacter, Chujaibacter, and Luteimonas).

Fig. 6
figure 6

Cladogram shows the biomarker microbes of the microbial and fungal lineages from Domain to genus among three different treatments. Circle midpoints in (a) and (c) represent indicator bacteria and fungi with LDA scores greater than two in microbial groups. Circles indicate phylogenetic level from phylum to family. The diameter of each circle is proportional to the abundance of the group. Hollow dot represents the microbes with no statistical differences among three treatments; (b) and (d) histogram of LDA scores computed for differentially abundant bacteria and fungi among three treatments

Full size image

In addition, for fungi (Fig. 6c), Sordariomycetes (class), Hypocreales (order), Hypocreales_fam_Incertae_sedis, Sporormiaceae, Nectriaceae, Mycosphaerellaceae, and Pleosporales_fam_Incertae_sedis (family), Mycosphaerella, Nigrograna, Westerdykella, and Fusarium (genus) were significantly enriched in CK. Eurotiomycetes, and Leotiomycetes (class), Eurotiales, Thelebolales, and Cystofilobasidiales (order), Aspergillaceae, Pseudeurotiaceae, Hypocreaceae, Chaetomiaceae, and Mrakiaceae (family), Aspergillus, Penicillium, Pseudogymnoascus, Trichoderma, Mariannaea, Humicola, Staphylotrichum, and Tausonia (genus) were significantly enriched in PFOA contaminated soils. And in PFOS polluted soils, 18 clades were enriched including Mucoromycota (phylum), Mucoromycetes (class), Helotiales, Microascales, and Mucorales (order), Herpotrichiellaceae, Trichocomaceae, Myxotrichaceae, Microascaceae, Mucoraceae, and Rhizopodaceae (family), Exophiala, Talaromyces, Oidiodendron, Cephalotrichum, Gamsia, Mucor, and Rhizopus (genus). Bacteroidetes was found to be the dominant phylum in soil (Keshri et al. 2013), and it was usually associated with organic carbon-rich substrates in soils (Shi et al. 2019). Previous studies reported that Mucoromycota and Planctomycetes were positively correlated with soil pH, nitrogen, total phosphorous, available phosphorous, carbon and available potassium (Tarin et al. 2021). In addition, it was reported that the microbial clades of Sphingobacteriales showed a highly significant correlation with the degradation of lignocellulosic (Abrusci et al. 2009). These active biomarkers reflect differences in abundance among various treated soils (PFOA and PFOS), suggesting that the bacteria and fungi in the different treatments are perhaps due to functional drivers. Furthermore, LEfSe analysis showed that control soils (CK) enriched more bacterial and fungal clades, it was suggested that PFOA and PFOS might have adverse effects on the soil microbial community especially PFOS treatment (Fig. 6).

4 Conclusions

The present study was designed to systematically evaluate the effect of PFOA and PFOS on the soil ecosystem and soil organic carbon mineralization. In the study, divergent organic mineralization of short-term and long-term PFOA/PFOS contamination in red soil has been observed. It was found that soil organic carbon mineralization and carbon emissions were significantly facilitated by PFOA and PFOS short-term pollution, even at a low level of contamination (0.1 mg·kg−1), and there is no doubt that PFOA and PFOS must contribute to the loss of global soil carbon pools and global climate problems. For the microorganism community, PFOA and PFOS had a significant impact on the diversity and richness of soil microorganisms. The reorganization of microbial and fungal communities produced more Proteobacteria, Acidobacteria, Gemmatimonadetes, and Mucoromycota, and reduced Actinobacteria, Proteobacteria, Chloroflexi, and Basidiomycota. LEfSe analysis showed they might have adverse effects on the soil microbial community especially PFOS treatment. Overall, these findings contribute in several ways to our understanding of the impacts of PFOS and PFOS on soil carbon cycling and ecosystem, and provide a basis for soil PFOA and PFOS pollution remediation. However, this work was performed in a laboratory scale, and it is difficult to comprehensively evaluate the ecological effect in PFOA and PFOS polluted soil. In the future, further large-scale field experiments are necessary to comprehend.