Soil microbiomes divergently respond to heavy metals and polycyclic aromatic hydrocarbons in contaminated industrial sites.

Contaminated sites from electronic waste (e-waste) dismantling and coking plants feature high concentrations of heavy metals (HMs) and/or polycyclic aromatic hydrocarbons (PAHs) in soil. Mixed contamination (HMs + PAHs) hinders land reclamation and affects the microbial diversity and function of soil microbiomes. In this study, we analyzed HM and PAH contamination from an e-waste dismantling plant and a coking plant and evaluated the influences of HM and PAH contamination on soil microbiomes. It was noticed that HMs and PAHs were found in all sites, although the major contaminants of the e-waste dismantling plant site were HMs (such as Cu at 5,947.58 ± 433.44 mg kg-1, Zn at 4,961.38 ± 436.51 mg kg-1, and Mn at 2,379.07 ± 227.46 mg kg-1), and the major contaminants of the coking plant site were PAHs (such as fluorene at 11,740.06 ± 620.1 mg kg-1, acenaphthylene at 211.69 ± 7.04 mg kg-1, and pyrene at 183.14 ± 18.89 mg kg-1). The microbiomes (diversity and abundance) of all sites were determined via high-throughput sequencing of 16S rRNA genes, and redundancy analysis was conducted to investigate the relations between soil microbiomes and contaminants. The results showed that the microbiomes of the contaminated sites divergently responded to HMs and PAHs. The abundances of the bacterial genera Sulfuritalea, Pseudomonas, and Sphingobium were positively related to PAHs, while the abundances of the bacterial genera Bryobacter, Nitrospira, and Steroidobacter were positively related to HMs. This study promotes an understanding of how soil microbiomes respond to single and mixed contamination with HMs and PAHs.

electronic waste

heavy metals

polycyclic aromatic hydrocarbons

organic carbon

naphthalene

acenaphthylene

fluorene

phenanthrene

anthracene

fluoranthene

pyrene

benzo[A]anthracene

chrysene

benzo[b]fluoranthene

benzo[k]fluoranthene

benzo[a]pyrene

dibenzo[a,h]anthracene

benzo[ghi]perylene

indeno[1,2,3-cd]pyrene

16S ribosomal RNA

amplicon sequence variants

principal coordinate analysis

linear discriminant analysis (LDA) coupled with effect size measurements

redundancy analysis

Introduction

Soil ecosystems are important parts of terrestrial ecosystems upon which humans rely [1]. Anthropogenic activities such as mining, coking, and electronic waste (e-waste) dismantling processes have led to widespread pollution of soil with heavy metals (HMs) and refractory organics such as polycyclic aromatic hydrocarbons (PAHs) [[2], [3], [4], [5], [6], [7]]. The contamination by HMs and PAHs poses a huge threat to soil microbiomes, human health, and natural ecosystems [[8], [9], [10], [11], [12], [13]]. Soil microorganisms are sensitive to pollutants and can be used as ecological indicators to evaluate soil contamination [14]. The mixed contamination of HMs and PAHs shows different influences on soil microorganisms compared to the single contamination [15]. An intensive understanding of how soil microbiomes respond to single and mixed contamination with HMs and PAHs is essential for evaluating soil health and developing effective bioremediation processes [16].

HMs such as Cd, Cr, and Pb are widely present in soil, which are toxic to organisms [[17], [18], [19], [20]]. Studies have reported changes in microbial metabolic activity, population diversity, and abundance of soil microbiomes in response to HMs [[21], [22], [23]]. Bacteria such as Proteobacteria, Acidobacteria, and Bacteroidetes are found in soil that is contaminated with HMs [9,24,25]. Bacteria may be able to eliminate the toxicity of HMs via adsorption, oxidation, and reduction mechanisms [26,27]. In addition to HMs, PAHs are also common soil pollutants produced from industrial processes and human activities, e.g., incomplete combustion and pyrolysis of organic substances, waste transportation, and incineration [[28], [29], [30]]. PAHs cause great concerns due to their high hydrophobicity, toxicity, and low bioavailability, especially those with more than four aromatic rings [31]. Soil contamination with PAHs has been reported to result in decreases in soil microbial diversity, abundance, and metabolic function [32]. Bacterial genera such as Rhizobacter, Sphingobium, Mycobacterium, Bacillus, and Pseudarthrobacter have been found to increase in abundance in response to PAHs [8,33,34].

Besides, many contaminated sites are affected by both HMs and PAHs [35,36]. Bourceret et al. [37] found that Proteobacteria, Actinobacteria, and Bacteroidetes were the dominant phyla in soil that was contaminated with HMs and PAHs. Gran-Scheuch et al. [38] reported that genera such as Sphingomonas, Ferruginibacter, Pseudomonas, Rhodanobacter, and Sphingobium were abundant in the soil of King George Island (Antarctica), where HMs and PAHs were detected. Although the soil microbiomes in soil contaminated by HMs and PAHs have been reported, the differences of the soil microbiomes with single and mixed contamination of HMs and PAHs are still confusing.

In this study, we aimed to explicate the characteristics of soil microbiomes from an e-waste dismantling plant and a coking plant with single and mixed contamination with HMs and PAHs. Microbial communities of four soil samples, including one with higher concentrations of HMs, one with higher concentrations of PAHs, and two with higher concentrations of HMs and PAHs, were compared to a control sample from a paddy field by the Illumina MiSeq sequencing and redundancy analysis. We found that the microbial community structure of soil microbiomes with high concentrations of HMs was different from that with high concentrations of PAHs. This study promotes an understanding of how soil microbiomes respond to single and mixed contamination with HMs and PAHs and provides insight into soil health evaluation.

Materials and methods

Materials

Chemical reagents, including hydrochloric acid, nitric acid, acetone, n-hexane, and acetonitrile of analytical grade, were purchased from Sinopharm Chemical Reagent Co., Ltd. (China). Decafluorobiphenyl was purchased from Shanghai Eon Chemical Technology Co., Ltd. (China), and dichloromethane was purchased from Shanghai Macklin Regent Co., Ltd. (China). A standard solution of metals (National Institute of Metrology, China) was used to measure the concentrations of K, Ca, Na, Mg, Fe, Al, and Ti; multielement calibration standard 3 (PerkinElmer, USA) was used to measure those of Mn, As, Cr, Cd, Pb, Cu, Zn, Ni, Ba, Sr, V, Li, and Co; and Wave Cal solution (PerkinElmer, USA) was used to measure those of La, Mo, and Sc. Acetonitrile of HPLC grade was obtained from Concord Technology (Tianjin) Co., Ltd. (China). Certificated Reference Materials for the Chemical Composition of Soils (GBW07556, China) from the Institute of Geophysical and Geochemical Exploration was used to measure the recoveries of metals. A standard solution of 16 PAHs was obtained from TMstandard Co., Ltd. (China). Standard fluorene from Shanghai Aladdin Bio-Chem Technology Co., Ltd. (China), fluoranthene from sigma-aldrich Co., Ltd. (USA), pyrene from Shanghai Meryer Chemical Technology Co., Ltd. (China), and benzo[a]anthracene from Shanghai Macklin Regent Co., Ltd. (China) were used to evaluate the recoveries of PAHs. A DNeasy® PowerSoil® Kit (QIAGEN, Germany) was used to extract DNA from soil samples.

Soil sampling

An initial investigation of soil contamination with HMs and PAHs was carried out at an e-waste dismantling plant and a coking plant. The two plants were located 48.8 km from each other and showed similar climatic features. Five samples (annotated samples A-E) were selected for further investigation according to their concentrations of HMs and PAHs. Sample A was collected from a paddy field near the e-waste dismantling plant and served as the control in this study. Sample B was obtained from the e-waste dismantling plant, and samples C-E were obtained from the coking plant. Sample C was collected from the field where coal and iron were stacked, sample D was obtained near the coking workshop, and sample E was significantly polluted with tar oil. Compared to sample A, which had relatively low concentrations of HMs and PAHs, samples B and E were characterized by high concentrations of either HMs or PAHs, and samples C and D featured high concentrations of both HMs and PAHs.

Sampling was conducted on December 12th, 2020. The local temperature was 5–9 °C. The upper 30 cm of the soil from each site was collected within a 50 × 50 cm square area after removing the root mat. Five soil subsamples were mixed to obtain one composite soil sample. The soil samples were stored in sterile sampling bags at 4 °C and transported to the laboratory. After drying and sieving (0.15 mm), triplicate soil samples of 100 g for each sampling site were used for soil characterization. The analysis methods of soil properties, including moisture, organic carbon (OC), ammonia-nitrogen, and nitrate-nitrogen, are specified in the Supporting Information.

Chemical analysis

Analyses of HMs

The concentrations of HMs were determined using an inductively coupled plasma optical emission spectrometer (ICP-OES; Optima 5300 DV, PerkinElmer, USA) after acidic digestion [39]. Triplicate soil samples (approximately 1.0 g each) in 7.0 mL of aqua regia (HCl:HNO3, 3:1, v/v) were digested at 185 °C for 40 min using a microwave digestion system (Topwave, analytikjena, Germany). After digestion, the acid solution was boiled gently on a hot plate in a fuming cupboard until a volume of 2–3 mL remained. The residual acid solution was dissolved in ultrapure Milli-Q water to a final volume of 50 mL. The concentrations of metals (Al, Fe, Ca, Mg, Na, K, Ti, Cd, Mn, Cu, Ni, Pb, Zn, Ba, Sr, Co, Mo, Cr, As, V, La, Li, and Sc) in the solutions were measured using ICP-OES. The certified soil GBW07556 (GSS-65) in triplicate was digested and measured at the same condition [34,40]. Recoveries of the metals in the certified soil were 85.02–87.39% (Al), 86–104.65% (Fe), 85.3–91.37% (Ca), 89.71–97.37% (Mg), 88.78–92.78% (Na), 83.58–93.41% (K), 86.06–94.83% (Ti), 93.39–98.3% (Cu), 89.47–94.07% (Zn), 88.8–90.37% (Mn), 101.64–102.41% (Pb), 89.91–102.06% (Ba), 91.97–93.64% (Ni), 90.8–93.11% (Co), 94.72–98.11% (Sr), 86.21–97.37% (Cd), 92.53–100.41% (Mo), 94.67–100.68% (As), 92.78–101.36% (Cr), 94.58–97.06% (La), 99.13–105.02% (Li), 90.5–93.15% (V), and 94.56–98.93% (Sc).

Analyses of PAHs

PAHs in the soil were extracted using the Soxhlet method and analyzed by high-performance liquid chromatography (HPLC; 1260 Infinity, Agilent Technologies, USA) [41]. Triplicate soil samples of 10.0 g each were extracted with 100 mL of acetone and n-hexane (1:1, v/v) using a Soxhlet extractor and spiked with 200 ng of decafluorobiphenyl. Then, the extract was filtered through fiberglass meshes and concentrated and cleaned up using a rotary evaporator and Si SPE cartridge (CNW, ANPEL Laboratory Technologies (Shanghai) Inc., China), respectively. A mixture of n-hexane and dichloromethane (1:1, v/v) was used to dissolve the PAHs. The extract was concentrated by a rotary evaporator and dissolved in acetonitrile to a final volume of 5.0 mL. The concentrations of PAHs in the concentrated acetonitrile solution were analyzed using HPLC with an Eclipse plus C18 column (Agilent, USA) and acetonitrile-water as the mobile phase according to the following gradient: 6:4 acetonitrile:water from the start to 18 min; only acetonitrile from 18 to 28.5 min; and 6:4 acetonitrile:water from 28.5 min to the end. The analyzed PAHs were naphthalene (NAP), acenaphthylene (ACY), fluorene (FLO), phenanthrene (PHE), anthracene (ANT), fluoranthene (FLA), pyrene (PYR), benzo[a]anthracene (BaA), chrysene (CHR), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[a]pyrene (BaP), dibenzo[a,h]anthracene (DhA), benzo[ghi]perylene (BgP), and indeno [1,2,3-cd]pyrene (IcP). The concentrations of the 15 PAHs in the soil samples were determined, and the total concentration (ΣPAHs) was calculated. Soil samples A and E with the lowest and highest concentrations of PAHs in duplicates, respectively, were used for quality control [40,42]. The recoveries of the 15 PAHs in sample A were 68.5–72.95% (NAP), 48.72–53.56% (ACY), 46.44–49.55% (FLO), 84.34–85.34% (PHE), 79.04–80.75% (ANT), 110.05–132.29% (FLA), 77.08–77.08% (PYR), 80.65–82.59% (BaA), 82.56–85.37% (CHR), 75.25–75.25% (BbF), 82.61–82.76% (BkF), 80.43–82.74% (BaP), 85.57–85.63% (DhA), 75.87–77.94% (BgP), and 86.51–87.01% (IcP), respectively. The recoveries of FLO, FLA, PYR, and BaA in sample E were 70.51–108.91%, 97.10–115.30%, 75.65–78.08%, and 110.62–119.34%, respectively. The methods are specified in the Supporting Information.

Characteristics of microbial communities

Soil DNA was extracted from triplicate samples using a DNeasy® PowerSoil® Kit. The primer set 515F (5‘-GTGCCAGCMGCCGCGG-3’)/907R (5′-CCGTCAATTCMTTTRAGTTT-3′) was used to amplify the V4–V5 region of bacterial 16S ribosomal RNA (16S rRNA) genes [[43], [44], [45]]. The PCR products were sequenced by Shanghai Majorbio Biopharm Technology Co., Ltd. (China) using the Illumina MiSeq Platform (San Diego, USA). The microbial diversity and composition were analyzed using QIIME2 (v2020.2). The Shannon index was calculated using MOTHUR software (v1.30). One-way analysis of variance (ANOVA) was used to establish significant differences between the groups with p < 0.05. Beta diversity at the level of amplicon sequence variants (ASVs) was evaluated by principal coordinate analysis (PCoA). Linear discriminant analysis (LDA) coupled with effect size measurement (LEfSe) analysis was performed to screen for various bacterial species. Redundancy analysis (RDA) was conducted with R software (v.3.3.1) using the Vegan package to determine the correlation between the soil microbiome diversity and the level of PAHs/HMs.

Results

The e-waste dismantling plant and the coking plant sites feature HMs and PAHs

Soil samples are collected from the coking plant (samples C, D, and E), the e-waste dismantling plant (sample B), and the nearby paddy field (sample A). The soil properties (moisture, pH, organic carbon, ammonia nitrogen, and nitrate nitrogen) of the five soil samples are presented in Table S1. The moisture of the soil samples was approximately 8.31–26.49%. The soil pH was in the range from 5.96 ± 0.08 (control) to 8.56 ± 0.28 (sample C). The concentration of organic carbon in sample D (51.93 ± 0.88 mg kg−1) was significantly higher than those in the other samples (13.28–31.59 mg kg−1). The concentrations of ammonia nitrogen and nitrate nitrogen in samples A and E (5.77–10.38 mg kg−1 and 14.16–24.12 mg kg−1, respectively) were higher than those in samples B, C and D (<2.21 mg kg−1 and <1.06 mg kg−1, respectively).

The contamination of the soil samples was evaluated based on the concentrations of metals (Al, Fe, Ca, Mg, Na, K, Ti, Cd, Mn, Cu, Ni, Pb, Zn, Ba, Sr, Co, Mo, Cr, As, V, La, Li, and Sc) and PAHs (NAP, ACY, FLO, PHE, ANT, FLA, PYR, BaA, CHR, BbF, BkF, BaP, DhA, BgP, and IcP), which are presented in Fig. 1 and Table S1. Metals Al, Fe, and Ca were present in all the soil samples at high concentrations (1,000–10,000 mg kg−1); metals Mg, K, Ti, Mn, Zn, and Ba at moderate concentrations (100–1,000 mg kg−1); and most of the remaining metals, such as Cd, Mo, La, Li, and Sc, at concentrations of less than 100 mg kg−1. HMs such as Mn, Zn, and Cu showed significant influences on the microbial community structures, in contrast to common metals such as Al, Fe, and Ca, according to RDA results (Fig. 4), and were therefore mainly considered in this study. In addition, most kinds of PAHs were present at a concentration of 100 mg kg−1. Various kinds of PAHs, such as ACY, FLO, and PYR, in Sample E were present in concentrations that exceeded 100 mg kg−1. Pearson correlation analysis (Fig. S1) showed that HMs such as Mn, Zn, Cu, and Mo were highly related to each other (p < 0.5) and that PAHs such as ACY, PYR, and FLO were significantly related to each other (p < 0.5).

Fig. 1

Concentrations of HMs (left) and PAHs (right) in the soil samples. The number in each cell is the average concentration. ND denotes not detected.

Fig. 4

Redundancy analysis (RDA) of environmental factors and the community structure of bacteria in soil with single contamination (a) and mixed contamination (b) of HMs and PAHs.

The concentrations of HMs and PAHs in sample A were relatively low compared to those in the remaining soil samples. For example, the concentrations of Zn (307.7 ± 21.4 mg kg−1), Ba (229.62 ± 25.95 mg kg−1), Co (15.06 ± 1.37 mg kg−1), and As (7.12 ± 0.27 mg kg−1) in sample A were much lower than those in other samples. The total concentration of PAHs (ΣPAHs) was 4.40 ± 0.84 mg kg−1, which was the lowest among these samples. In addition to FLO (3.00 ± 0.68 mg kg−1), the other PAHs were present in concentrations of below 0.30 mg kg−1.

Sample B showed the highest concentrations of HMs among the five samples but low concentrations of PAHs. The concentrations of HMs, including Cu (5,947.58 ± 433.44 mg kg−1), Zn (4,961.38 ± 436.51 mg kg−1), Mn (2,379.07 ± 227.46 mg kg−1), Pb (1,314.33 ± 55.92 mg kg−1), and Ba (1,156.47 ± 103.39 mg kg−1), were approximately tenfold higher than those in the other groups. Cd was only detected in sample B (27.39 ± 3.94 mg kg−1). The ΣPAHs in sample B (10.36 ± 0.74 mg kg−1) was slightly higher than that in sample A. In contrast, sample E showed a low concentration of HMs but a high concentration of PAHs. The concentrations of HMs such as Mn (672.61 ± 7.13 mg kg−1), Ni (31.05 ± 1.95 mg kg−1), and Sr (43.08 ± 4.09 mg kg−1) were even lower than those in sample A. The concentration of FLO was 11,740.06 ± 620.1 mg kg−1, followed by those of ACY (211.69 ± 7.04 mg kg−1), PYR (183.14 ± 18.89 mg kg−1), and BbF (107.98 ± 4.44 mg kg−1), which resulted in ΣPAHs of as high as 12,558.06 ± 611.19 mg kg−1.

Most of the concentrations of HMs and PAHs in samples C and D were between those in samples B and E. The concentrations of HMs such as Mn, Zn, and Ba in sample C (954.34 ± 93.97 mg kg−1, 430.7 ± 15.79 mg kg−1, and 268.76 ± 11.99 mg kg−1, respectively) were lower than those in sample D (1,306.82 ± 48.57 mg kg−1, 451.5 ± 7.66 mg kg−1, and 353.08 ± 41.14 mg kg−1, respectively). The ΣPAHs in sample C (208.00 ± 52.21 mg kg−1) was also lower than that in sample D (698.74 ± 176.62 mg kg−1). The main PAHs in sample C were CHR (29.65 ± 9.71 mg kg−1), BkF (28.72 ± 0.28 mg kg−1), and BaP (24.29 ± 7.35 mg kg−1), and the main PAHs in sample D were BbF (95.32 ± 20.40 mg kg−1), BkF (94.20 ± 20.76 mg kg−1), and FLA (90.83 ± 29.15 mg kg−1).

Compared to sample A, samples B and E featured single contamination with HMs and PAHs, respectively, and samples C and D were characterized by mixed contamination with HMs and PAHs. The high concentrations of HMs in sample B could be related to the e-waste dismantling process, whereas the high concentrations of PAHs in sample E should be related to the oil tar [46]. The coking and steeling processes may have led to increased concentrations of both HMs and PAHs in samples C and D [47].

α- and β-diversities of the soil microbiomes

The 16S rRNA genes of bacteria were sequenced to analyze the soil microbiomes. The rarefaction curves (Fig. S2) indicated that the measurement of the sequences was sufficient. After quality control, 831,803 sequences of 16S rRNA genes were obtained. These sequences were clustered into 28,746 ASVs and assigned to 1,271 genera of 50 bacterial phyla. The α-diversities of the soil microbiomes (the Shannon, Simpson, Ace, and Chao1 indices) were evaluated (Table S2). The Shannon indices of samples A, C, and D reached as high as 7.10–7.33 and were close to one another (p > 0.05). Compared to that of sample A, the Shannon indices of samples B (6.31 ± 0.15) and E (5.01 ± 0.20) decreased significantly (p < 0.05), thereby showing that the richness of soil microbiomes in samples B and E was apparently lower than that in samples A, C, and D. This result showed that the soil microbiomes could survive at certain contamination levels of HMs and PAHs (samples C and D), but that high concentrations of either HMs or PAHs (samples B and E) decreased the α-diversities of the soil microbiomes. The β-diversity of the soil microbiomes was evaluated by PCoA (Fig. S3). The samples were distant from one another, which suggested that the community structures of the soil microbiomes were different from one another.

Features of the soil microbial community at the phylum and genus levels

The microbial community structures of the soil microbiomes were analyzed at the phylum and genus levels (Fig. 2). There were twelve bacterial phyla with relative abundances of greater than 1% in the soil samples (Fig. 2a and Fig. S4). The top ten phyla that were found in the soil samples were Proteobacteria, Acidobacteriota, Actinobacteriota, Bacteroidota, Chloroflexi, Planctomycetota, Myxococcota, Gemmatimonadota, Firmicutes, and Cyanobacteria. Actinobacteriota (21.56%) and Firmicutes (4.31%) were relatively abundant in sample A but less abundant in the other samples. Acidobacteriota (24.71%) and Myxococcota (6.90%) showed the highest relative abundances in sample B. The relative abundance of Proteobacteria increased significantly from 28.24% to 78.69% from sample A to sample E. These results were similar to those of previous studies in which Acidobacteriota and Myxococcota were more frequently detected in HM-contaminated environments [48] and Proteobacteria were found to be more abundant when the concentration of PAHs was higher [49].

Fig. 2

Bacterial phyla (a) and genera (b) in the soil samples. Unclassified genera in various domains (d), phyla (p), classes (c), orders (o), and families (f) are specified.

Thirty-nine genera had relative abundances of above 1.5%, including twenty-two classified genera (56.41%) and seventeen unclassified genera (43.59%) (Fig. 2b and Fig. S5). This high proportion of unclassified genera showed that many bacteria still need to be characterized in co-contaminated sites. The genera in sample A were quite diverse, and each of them was present with a similar relative abundance. In sample B, MND1 (7.95%), Bryobacter (7.54%), and an unclassified genus of Vicinamibacteriales (6.66%) showed high relative abundances. Flavobacterium showed the highest abundance in sample C (5.39%), and Pseudomonas was well represented in sample D (6.01%). Sulfuritalea was detected in samples D (0.36%) and E (21.34%) but barely detected in samples A, B, and C. In addition, Rhodanobacter (9.06%), an unclassified genus of Comamonadaceae (9.35%), Sphingobium (4.03%), Pandoraea (3.70%), and Pseudoxanthomonas (2.92%) showed higher relative abundances in sample E.

LEfSe with an LDA score of >4.0 was used to identify the main genera that were differently represented in the microbiomes of the soil samples (Fig. 3a). There were eight distinctive genera with LDA scores of >4.0 in sample B, six in sample E and three in samples A, C, and D. The three most distinctive genera of each sample are presented in Fig. 3b, and the others are presented in Fig. S5.

Fig. 3

LEfSe results of bacterial genera (a) and relative abundances of the distinctive genera (b).

Massilia (2.18%) and unclassified genera of Gaiellales (3.93%) and Acidobacteriales (3.85%) were present mainly in sample A, and all of them showed relative abundances of <1% in the other samples, which implied that these genera were not capable of resisting HMs or PAHs. Genera such as MND1, Bryobacter, and an unclassified genus of Vicinamibacteriales showed the highest relative abundances in sample B (6.66–7.95%). Genera including Sulfuritalea, Rhodanobacter, and an unclassified genus of Comamonadaceae were the most abundant in sample E (9.06–21.34%), whereas the three genera showed relatively lower abundances <2.26% in samples A, B, C, and D. Flavobacterium in sample C (5.39%) and Pseudomonas in sample D (6.01%) showed higher abundances than other genera in the soil with high concentrations of HMs and PAHs [37,38]. The results implied that the bacteria which were tolerant of high concentrations of HMs were also found in soil with mixed contamination with HMs and PAHs but that the bacteria which were dominated by high concentrations of PAHs hardly survived in soil with mixed contamination with HMs and PAHs, which has rarely been reported.

HMs and PAHs divergently impact the soil microbiomes

The impacts of HMs and PAHs on the soil microbiomes were evaluated with RDA (Fig. 4, Table S3, and Table S4). The results showed that the microbial community structures of these soil samples were influenced mainly by PAHs (such as ACY, FLO, and BbF) and HMs (such as Mn, Pb, and Ni) and slightly by soil properties (such as pH, moisture, and ammonia). Sample A was far from the others and less influenced by HMs and PAHs. The microbiomes of sample B were influenced mainly by HMs, and the microbiome of sample E was highly influenced by PAHs. In contrast, the microbial community structure of sample D was influenced by both HMs and PAHs, and the influences were more significant than those of sample C. In addition, the community structure of the soil bacteria in sample D was also significantly influenced by soil organic carbon. The RDA results were in accordance with soil properties and the concentrations of HMs and PAHs.

The correlations between HMs and PAHs and bacterial phyla and genera are presented in Fig. 5 and Fig. S6. Impressively, the bacteria that showed positive relations with HMs or PAHs were obviously different. At the phylum level (Fig. 5a), Acidobacteriota, Myxococcota, and Nitrospirota were positively related to HMs but negatively related to PAHs, whereas Proteobacteria and Bacteroidota were positively related to PAHs. At the genus level (Fig. 5b), MND1, Gaiella, Nitrospira, and Steroidobacter were positively related to HMs and negatively related to PAHs [9], whereas Rhodanobacter, Sphingobium, and Pseudoxanthomonas were positively related to PAHs and negatively related to HMs [34]. The results were in accordance with the LEfSe results.

Fig. 5

Correlations between the contaminants and bacterial phyla (left) and genera (right).

Discussions

The results demonstrated that the microbiomes divergently responded to single contamination with HMs or PAHs and mixed contamination with HMs and PAHs in soil. The dominant bacteria in this study were in accordance with previous reports (Table S5). For example, Rogiers et al. and Qin et al. reported that bacterial phyla such as Proteobacteria (23.8–43.0%) and Acidobacteria (14.6–34.5%) and bacterial genera such as Nitrosospira (0.27–3.48%) were abundant in soil that was contaminated with HMs such as Pb (19–225 mg kg−1), Cr (180–410 mg kg−1), and Cd (2.4–110 mg kg−1) [9,24]. Geng et al., Liu et al., and Miao et al. reported that bacterial phyla such as Proteobacteria (20.86–81.37%) and bacterial genera such as Pseudomonas, Sphingomonas, Flavobacterium, and Pseudoxanthomonas were abundant in soil that was contaminated with PAHs (0.11–56.11 mg kg−1 in total) [8,34,50]. In addition, bacterial phyla such as Proteobacteria (35.3–98.45%), Bacteroidetes (0.3–18.6%), and Actinobacteria (0.6–13.4%) and bacterial genera such as Pseudomonas (0.7–25.0%) and Flavobacterium (5.2%) were found in soil that was contaminated with HMs (such as Pb at a concentration of 19–1,133 mg kg−1, Mn at 64–435 mg kg−1, and Cr at 0.04–1,033 mg kg−1) and PAHs (15–2,002 mg kg−1 in total) [37,38,51]. Compared to previous reports, the concentrations of HMs and PAHs in soil samples in this study were much higher, and the divergent responses of microbiomes to such high concentrations of HMs and PAHs were rarely reported. Our study may facilitate understanding of how microbiomes respond to HMs and PAHs.

Bacteria may interact with HMs and PAHs via multiple strategies. For example, bacteria could use HMs as electron donors or acceptors for energy production [52] or use PAHs as both carbon and energy sources for growth [53]. Bacterial genera such as MND1 and Bryobacter decreased in abundance with decreasing concentrations of HMs but remained abundant in the presence of PAHs. In contrast, bacterial genera such as Sulfuritalea and Sphingobium failed to survive when the concentrations of PAHs were reduced. Therefore, it is inferred that PAH degraders are more sensitive to the concentrations of HMs and PAHs than HM-resistant bacteria.

In addition to HMs and PAHs, microbiomes can also be influenced by soil physicochemical properties [36]. Eutrophic Pseudomonas, Sulfuritalea, and Pandoraea were abundant in sample D. In contrast, the relative abundance of a putative oligotrophic and unclassified genus of Oxalobacteraceae increased in sample C.

Bacteria surviving with HMs, PAHs or HM and PAH mixtures may serve as potential resources for bioremediation and for investigation of bacterial evolution. Various studies reported that Bryobacter, Gaiella, and Nitrospira were able to resist HMs [[54], [55], [56]], and Pseudomonas, Sphingobium, and Pseudoxanthomonas were able to degrade PAHs [[57], [58], [59]]. There were also reports that Pseudomonas and Sphingobium strains were able to resist HMs and degrade PAHs [58,60]. For example, Pseudomonas putida UW4 could degrade FLA in the presence of Pd [60], and the Cu-tolerant Sphingobium sp. PHE-1 could degrade PHE [61]. However, the knowledge on Sulfuritalea, Bryobacter, and Pseudoxanthomonas is still limited to the bacterial community level [38,62,63], and their adaptation mechanism to HMs/PAHs and application potential for bioremediation need to be explored.

Conclusions

In our study, soil microbiomes divergently responded to single and mixed contamination with HMs and PAHs. The bacterial genera Sulfuritalea, Pseudomonas, and Sphingobium were positively related to PAHs, and the bacterial genera Bryobacter, Nitrospira, and Steroidobacter were positively related to HMs. This study showed how hazardous HMs/PAHs changed the soil microbiome.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Author contributions

Methodology, data analysis and original draft preparation, Z. N. Yang and Z. S. Liu; measurements of PAHs and HMs, K. H. Wang and Z. L. Liang; sample preparation and DNA extraction, R. H. Wang, H. L. Ma, X. K. Wang, M. L. Yang, and B. G. Zhang; bacterial analysis, R. Abdugheni and Y. Huang; review and editing, C. Y. Jiang, D. F. Li, P. F.-X. Corvini, and S. J. Liu; and funding acquisition, S. J. Liu and C. Y. Jiang. All authors have read and agreed to the published version of the manuscript.