TY - JOUR AU1 - Lemmel,, Florian AU2 - Maunoury-Danger,, Florence AU3 - Leyval,, Corinne AU4 - Cébron,, Aurélie AB - ABSTRACT Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous soil organic pollutants. Although PAH-degrading bacteria are present in almost all soils, their selection and enrichment have been shown in historically high PAH contaminated soils. We can wonder if the effectiveness of PAH biodegradation and the PAH-degrading bacterial diversity differ among soils. The stable isotope probing (SIP) technique with 13C-phenanthrene (PHE) as a model PAH was used to: (i) compare for the first time a range of 10 soils with various PAH contamination levels, (ii) determine their PHE-degradation efficiency and (iii) identify the active PHE-degraders using 16S rRNA gene amplicon sequencing from 13C-labeled DNA. Surprisingly, the PHE degradation rate was not directly correlated to the initial level of total PAHs and phenanthrene in the soils, but was mostly explained by the initial abundance and richness of soil bacterial communities. A large diversity of PAH-degrading bacteria was identified for seven of the soils, with differences among soils. In the soils where the PHE degradation activities were the higher, Mycobacterium species were always the dominant active PHE degraders. A positive correlation between PHE-degradation level and the diversity of active PHE-degraders (Shannon index) supported the hypothesis that cooperation between strains led to a more efficient PAH degradation. SIP, multi-contaminated soils, Illumina sequencing, ecology INTRODUCTION Polycyclic aromatic hydrocarbons (PAHs) are organic contaminants that represent a major source of environmental pollution, especially in soils. Anthropogenic activities are the main source of these contaminants due to the incomplete combustion of organic materials such as coke and petroleum products during industrial processing. In industrial brownfields, although pollution stopped several decades ago, soil contamination can still reach very high levels (Biache et al. 2017). PAHs can accumulate in soils and be a serious threat for human health and ecosystem functioning with regard to their carcinogenicity and toxicity (Eom et al. 2007; Abdel-Shafy and Mansour 2016). However, the stress from high PAH concentrations leads to adaptation and selection of soil microbial communities (Bourceret et al. 2016). In these soils, a selection of microorganisms able to tolerate PAHs or degrade them has been shown (Cébron et al. 2008; Sawulski, Clipson and Doyle 2014), and involves detoxification (Sutherland 1992) or metabolism (PAHs used as a carbon source; Ghosal et al. 2016) mechanisms. Therefore, high levels of PAH contamination could impair soil functioning and lead to loss of functions (i.e. functions involved in C and N cycling; Liang et al. 2009). However, it also leads to gain of functions (i.e. functions involved in PAH biodegradation; Cébron et al. 2008). Nevertheless, communities from weakly PAH-contaminated soils (soil background level) can also harbor the PAH degradation function (Crampon et al. 2017). Therefore, we can wonder if the effectiveness (i.e. the rate) of PAH biodegradation is similar between non- or low-contaminated soils on the one hand and soils with high and historical PAH contamination on the other hand. This aspect needs to be further explored to better understand the factors controlling the soil PAH biodegradation efficiency, which sometimes limits soil bioremediation (Chauhan et al. 2008; Ghosal et al.2016). In order to study the PAH biodegradation function, most studies have sought to detect only the presence of certain functional genes. PAH-RHDα (RHD: ring hydroxylating dioxygenase) genes, encoding the alpha subunit of PAH-RHD enzymes involved in the first step of the bacterial biodegradation pathway, are choice targets indicating the functional potential of the microbial community (Cébron et al. 2008; Sawulski, Clipson and Doyle 2014), but this approach does not identify the actors of biodegradation. Stable isotope probing (SIP) is a culture-independent technique that links microbial identity with functions in complex systems such as soils (Dumont and Murrell 2005). This technique uses 13C-labeled compounds and allows for the monitoring of 13C during biodegradation and the labeling of cell components (e.g. DNA) of microbes using the 13C-labeled substrate as a carbon source for their growth. SIP has until now been extensively used to study the degradation of single carbon compounds (Dumont et al. 2006; Mosbaek et al. 2016; He et al. 2019), but more and more studies use SIP to decipher the pathways and actors involved in degradation of natural and anthropic complex compounds, such as cellulose/lignin (Wilhelm et al. 2019 and various organic pollutants (Li et al. 2015; Khawand et al. 2016), among which are the PAHs, i.e. naphthalene, phenanthrene and pyrene (Cébron et al. 2011; Wald et al. 2015; Song et al. 2016; Guo et al. 2017; Thomas et al. 2019), in order to identify microbial degraders. However, only few investigations have been carried out on PAH degraders in weakly contaminated soils (Song et al. 2016; Chen et al. 2018), and we do not know whether their PAH-degrading bacterial diversity is similar to historically and highly PAH-contaminated soils. Most studies have used SIP to compare the PAH degradation yield and degrader identity in one soil under various conditions. For example, certain authors sought to understand the impact of surfactants (Crampon et al. 2017; Guo et al. 2017), root exudates (Cébron et al. 2011; Lv et al. 2018; Thomas, Corre and Cébron 2019) or temperature (Wald et al. 2015). On the other hand, SIP was never used to study a wide range of soils and compare their communities of PAH degraders in order to see (i) whether some taxa are more efficient at PAH degradation than others and (ii) whether historically contaminated soils harbor a different diversity of active microorganisms than weakly contaminated soils. In this context, the aim of this study was to compare a range of 10 soils displaying a gradient of anthropization, determine their PAH-degradation efficiency and identify the active PAH degraders. Our hypothesis was to test whether (i) the intensity of PAH degradation was correlated to the level of PAH contamination of the soils, (ii) the microbial communities from historically and highly PAH-contaminated soils were more efficient than those from weakly polluted soils and (iii) PAH degrader identity could explain the differences in PAH degradation rates, involving more or less efficient taxa. Phenanthrene (PHE), which is widely distributed in the environment and in contaminated soils, was used as a model PAH in SIP experiments. PHE-degrading bacteria were identified using high-throughput 16S rRNA gene amplicon sequencing from 13C-labeled DNA. MATERIALS AND METHODS Soil sampling and characteristics The soils used in this study were collected in November 2015 from various woodland ecosystems located on industrial wastelands, natural forest or ancient gravel pits located in the ‘Grand Est’ region (north-east of France); all sites are located within a 50-km radius (Figure S1, Supporting Information). At each of the 10 sites, samples were collected from three independent subsites 1 m apart. The three subsite soil blocks were mixed to get one composite sample per site. Back to the laboratory, the soil samples were air-dried at room temperature for 1 week, and then sieved at 2 mm. All the dried and sieved soils were stored similarly at room temperature in the dark (6 months) before SIP experiment setup. Three soils were considered as weakly anthropized soils: two natural forest soils collected at Hémilly (He; Moselle) and Montiers-sur-Saulx (Mo; Meuse), and one anthropized but unpolluted ancient gravel pit soil collected at Dieulouard (Di; Meurthe-et-Moselle). Seven anthropized soils known to be polluted by metals and/or PAHs were collected from (i) former slag heaps at Homécourt (Ho; Meurthe et-Moselle), Terville (Te; Moselle), Uckange (Uc; Moselle) and Neuves-Maisons (NM; Meurthe-et-Moselle) and (ii) former settling ponds at Pompey (Po; Meurthe-et-Moselle), Mont-St-Martin (MsM; Meurthe-et-Moselle) and Russange-Micheville (RM; Moselle). This soil collection was chosen to cover a wide range of anthropization situations described in Lemmel et al. (2019). Briefly, a gradient of polycyclic aromatic hydrocarbon contamination was shown based on the 16 US-EPA PAHs ranging from 0.03 to 1095.90 mg kg−1 dw soil (Table 1). The soils presented variable 16S rRNA gene abundances ranging from 4.4 to 41.5 × 1010 copies g−1. The lowest and highest 16S rRNA gene abundances were found in the Uc and NM soils and the Po and Te soils, respectively, while the other soils presented intermediate abundance values. The soil bacterial communities were characterized by Illumina MiSeq sequencing, and indices describing alpha-diversity such as the Choa1 richness index were calculated (Table S1, Supporting Information; Lemmel et al. 2019). The soils presented variable Chao1 richness indices ranging from 2541 to 4966, and the lowest and highest values were found in the RM and NM soils and Di and Ho soils, respectively. A gradient of metal contamination, mainly by Zn (60 to 119 000 mg kg−1), Pb (23 to 39 500 mg kg−1) and Cd (0.09 to 152 mg kg−1), was also shown (Table S1, Supporting Information). The pH, known to be one of the main drivers of microbial communities, was similar (from 7.2 to 8.0) among soils except for soil He (pH = 5.4). Other soil characteristics (Laboratoire d'analyse des sols, INRA, Arras, France), namely total organic carbon (24 to 159 g kg−1), total nitrogen (0.56 to 7.19 g kg-1), the C/N ratio (14 to 54) and texture (silty to sandy), varied among soils (Table S1, Supporting Information; Lemmel et al. 2019). Table 1. PAH contents (sum of the 16 regulatory PAHs and of phenanthrene concentration) and PAH-degradation gene (PAH-RHDα-GN and -GP) copy numbers and abundances in the soils and in the sequenced 13C-heavy DNA fraction, allowing to calculate functional gene enrichment in the 13C-heavy DNA fraction consequently to the SIP incubations. Sum of 16 regulatory PAH (mg kg−1) Phenanthrene (mg kg−1) Percentage of functional gene copy relative to 16S rRNA genes Gene enrichment during SIP incubation In soils (before SIP incubation) In soils (after SIP incubation) In sequenced 13C-heavy DNA fractions Total Available Total Available PAH-RHDα GP (× 10−3) PAH-RHDα GN (× 10−5) PAH-RHDα GP (× 10−2) PAH-RHDα GN (× 10−3) PAH-RHDα GP PAH-RHDα GN (× 10−3) PAH-RHDα GP PAH-RHDα GN He 0.03 ± 0.01
97.0%) dissolved in n-hexane and left under a fume hood until complete evaporation of the solvent. The final PHE concentration was 200 mg kg−1. Our experimental design included 60 samples (10 unspiked soils and PHE-spiked soils, in triplicates). The 2.0 g soil samples were placed in 125-ml plasma glass flasks and rewetted using sterilized water to reach 80% water retention capacity (WRC). Then, they were hermetically sealed with rubber-butyl corks and incubated in the dark at 24°C (optimal temperature and humidity conditions) for 17 days for soil samples Di, MsM, Mo, NM, Te and Uc, and incubations were continued until 23 days for soil samples Ho, He, RM and Po for which limited difference of respiration between BR and PHE-SIR could be observed. Respiration levels were monitored by measuring CO2 contents using an infrared spectrophotometer (Binos, absorption at 2325.6 cm−1) on 4-ml fractions of the flask atmosphere sampled after 1, 2, 4, 7, 10, 14, 17 and 23 days of incubation using a syringe. Produced CO2 was expressed as the carbon mass produced by 1 g of dw soil per hour (mg of C g−1 h−1). The flasks were left open under the fume hood for 15 min to renew the atmosphere after each sampling. SIP incubations For SIP experiments, two sets of soil samples were prepared in triplicates by spiking all samples with PHE as described above. The first set was spiked with non-labeled PHE (i.e. 12C-PHE; Fluka, >97.0% purity) and the second set was spiked with uniformly 13C-labeled PHE (Sigma Aldrich, 99 atom % 13C). Both sets were prepared twice: one was harvested at the beginning of the experiment (T0) and the other was incubated and harvested after 12 days (Tf). The SIP incubation experimental design included 120 samples (10 soils in triplicates, spiked with 12C- or 13C-PHE, and analyzed at 2 time points T0 and Tf). Each soil sample was placed in 125-ml plasma glass flasks and rewetted using sterilized water to reach 80% WRC. Then, the T0 samples were directly harvested by freezing at −80°C. A CO2 trap (2 ml of 1 M NaOH in a 5-ml vial) was added in the Tf flasks before being hermetically sealed and incubated for 12 days in the dark at 24°C, to avoid 13C-CO2 fixation (autotrophic and heterotrophic) during incubation. The atmosphere of the flasks was renewed once after 6 days, as described above. Tf incubations were harvested by freezing at −80°C. 13C-Phenanthrene degradation and 13C-dissipation measurements The soil samples (T0 and Tf) from SIP incubation stored at −80°C were freeze-dried and ground to 500 μm with a grinder (Mixer Mill MM 400, Retsch, Eragny-sur-Oise, France) before PAH extraction, PHE measurements and δ13C estimation. Total PHE PAHs were extracted from one aliquot of ∼500 mg of each dw soil with dichloromethane (DCM) at a high temperature (130°C) and a high pressure (100 bars) using accelerated solvent extraction (DIONEX® 200 ASE), as described in Cennerazzo et al. (2017). Solvent extracts were evaporated under a nitrogen flow and dissolved in acetonitrile for PHE analysis. PHE was analyzed by UV detection (254 nm) using a reverse-phase chromatography UHPLC DIONEX® Ultimate 3000 system equipped with a DAD (Diode Array Detector) and a Zorbax Eclipse PAH column (2.1 × 100 mm, 1.8 μm, Agilent). As some soils were already contaminated with aged PHE, the 13C-PHE concentrations (13C-PHEt) at T0 or after 12 days were calculated as follow: 13C-PHEt = PHEt − PHEi, where PHEt is the total PHE concentration measured and PHEi is the initial PHE concentration of the soils before 13C-PHE spiking. Delta13C measurements were performed at the PTEF platform (INRA, Champenoux, France). Briefly, δ13C was measured from one aliquot of ∼5 mg dw of soil using an elemental analyzer (vario ISOTOPE cube, Elementar, Hanau, Germany) interfaced in line with a gas isotope ratio mass spectrometer (IsoPrime 100, Isoprime Ltd, Cheadle, UK; 0.2 ‰ sensitivity). δ13C values were then transformed into R13C/12C using the following formula: R13C/12C = RvPDB × (1 + δ13C/1000), with RvPDB equal to 0.0112375, value of the reference standard (Vienna Pee Dee Belemnite). From these calculated ratios, we determined the 13C-dissipation corresponding to the percentage of 13C content (coming from the 13C-PHE) after the 12 days of incubation relative to the T013C content just after the spiking (representing 100%), taking also into account the natural 13C/12C ratio of the unspiked soils. Available PAH content of unspiked soils was measured using cyclodextrin-based extraction and analyzed by HPLC and UV-detection as described in Lemmel et al. (2019). DNA extraction, isopycnic ultracentrifugation and gradient fractionation Total genomic DNA was extracted from ∼300 mg of freeze-dried and ground dw soil (60 Tf samples from both 12C and 13C incubations) using a Fast DNA Spin Kit for Soil (MP Biomedicals, France), following the manufacturer's instructions. Then, DNA was resuspended in 100 μl of DNase and pyrogen-free water. Concentration and purity (A260/A280 ratio) were measured using a spectrophotometer (UV1800, Shimadzu, Marne-la-Vallée, France) equipped with a TrayCell adapter (Hellma®). Heavy (13C-labeled) DNA was separated from light (12C) DNA according to the protocol described by Neufeld et al. (2007). Briefly, ∼3000 ng of DNA (except NM and Uc, for which 650 ng of DNA was used because lower genomic DNA quantities were recovered from these soils) was added to Quick-Seal polyallomer tubes (13 × 51 mm, 5.1 ml, Beckman Coulter), along with a gradient buffer (0.1 M Tris–HCl, 0.1 M KCl, 1 mM EDTA) mixed with a CsCl solution, to a final buoyant density (BD) of ≈1.725 g ml−1. The tubes were centrifuged at 42 400 rpm (i.e. 176 985 × g; VTI 65.2 rotor, Beckman, Roissy, France) at 20°C for 40 h. Following ultracentrifugation, 13 fractions of ∼400 µl were collected from each tube using a fraction recovery system (Beckman). Then, the BD value of each fraction was measured, and CsCl was removed by glycogen-assisted polyethylene glycol precipitation. DNA fractions were finally resuspended in 30 µl of TE buffer (pH 8.0). Real-time quantitative PCR In order to determine the distribution of heavy and light DNA and compare samples from 12C and 13C incubations, the abundance of 16S rRNA genes in each fraction was quantified using real-time quantitative PCR. The qPCR assay was performed as described in Cébron et al. (2008) by using the primer sets 968F/1401R (Felske, Akkermans and Vos 1998). Briefly, the reaction mixture (20 μl) was composed of 10 μl of iQ SYBR green SuperMix (Bio-Rad), 0.8 μl of each primer (10 μM), 0.4 μl of bovine serum albumin (3%), 0.2 μl of dimethyl sulfoxide, 0.08 μl of T4gp32 (MP Biomedicals, France) and 1 μl of DNA as a template (DNA fraction samples or 10-fold dilution series from 108 to 101 copies μl−1 of the standard plasmid). Quantifications were performed using a CFX96 Real-Time PCR detection system (Bio-Rad, Marne-la-Coquette, France), under the following conditions: an initial denaturation step at 95°C for 5 min, and then 39 cycles of denaturation at 95°C for 30 s, annealing at 56°C for 20 s, extension at 72°C for 30 s and measurement of SYBR Green signal intensities at 82°C for 5 s. Similarly, the abundance of 18S rRNA genes was quantified using the primer sets Fung5F/FF390R (Lueders et al. 2004) as previously described in Thion et al. (2012a). Additionally, PAH-RHDα genes were also quantified in the genomic DNA from unspiked soils (before incubation) and in each DNA fraction (from the SIP assay) using the primer pairs PAH-RHDα GP-F/R for Gram-positive bacteria (Actinobacteria) and PAH-RHDα GN-F/R for Gram-negative bacteria (Proteobacteria) (Cébron et al. 2008). The reaction mixtures and quantification conditions were as described above except annealing temperatures that were 57 and 54°C for PAH-RHDα-GN and -GP, respectively, as described in Cébron et al. (2008). Based on qPCR quantifications, the relative abundances of PAH-RHDα GP and GN genes (related to 16S rRNA genes) were calculated in unspiked soils. Similarly, the abundances of PAH-RHDα GP and GN genes in the soils after incubation and in the sequenced DNA fractions were calculated based on their average relative abundance in the light 12C-DNA fraction and in the heavy 13C-DNA fraction, respectively. All these gene relative abundance values are reported in Table 1. Identification of phenanthrene-degrading bacteria Based on the results of the distribution of heavy and light DNA in the recovered fractions, we observed an enrichment of the 16S rRNA gene in the heavy fraction (numbers 4-5-6; i.e. BD from 1.725 to 1.717 g ml−1) of the soils incubated with 13C-PHE, indicating a significant use of 13C for bacterial growth. In order to identify these phenanthrene-degrading bacteria, 5 µl of the three heavy fractions were pooled from samples of both 12C and 13C incubations. These pooled fractions were used as templates to perform 16S rRNA gene amplicon libraries for Illumina MiSeq sequencing. To do so, the V3/V4 region of bacterial 16S rRNA genes (∼550 bp) was amplified using primers S-D-Bact-0341-a-S-17 and S-D-Bact-0787-b-A-20 (Muyzer, de Waal and Uitterlinden 1993; Caporaso et al. 2011), and following a previously described dual-index strategy (Kozich et al. 2013). PCR reactions were performed on 1 μl of the pooled DNA fractions in a final volume of 50 μl, containing 10 μl of 5× Phusion HF buffer, 1.5 μl of 50 mM MgCl2, 0.25 μl of DMSO, 0.1 μl of T4gp32 (MP Biomedicals, France), 0.1 μl of Phusion high-fidelity polymerase (Thermo Scientific) and 1 μl of each primer at 10 μM. PCR reactions were heated at 94°C for 5 min, followed by 33 cycles of 30 s at 94°C, 30 s at the annealing temperature (18 cycles starting at 63°C, with a subsequent decrease of 0.5°C at each cycle and 15 cycles at 54°C), 30 s at 72°C and a final extension step of 7 min at 72°C. Amplification products were checked by 1% agarose gel electrophoresis and purified using the UltraClean-htp 96-Well PCR Clean-Up kit (Qiagen), following the manufacturer's instructions. After quantification using a Quant-iT Picogreen ds-DNA assay Kit (Invitrogen), an amplicon library was prepared as an equimolar pool of amplicons (10 nM), purified on a QIAquick PCR purification kit column (Qiagen) and sent for sequencing to Genewiz platform (South Plainfield, NJ, USA) using an Illumina MiSeq V2 Kit for 2 × 250 bp paired-end sequencing. Illumina MiSeq paired-end reads have been deposited in the SRA database under BioProject accession number PRJNA507956. Sequence data were analyzed following the MiSeq SOP procedure available in March 2017 and described in Kozich et al. (2013), using Mothur v.1.36.0 (Schloss et al. 2009). Due to the altered sequence quality of soil Uc as compared to the other nine soils, two separate analyses were performed, with the Uc read treated separately from all the others. Paired-end reads were trimmed to a minimum QScore of 20 and joined using the following criteria: 404 bp < length < 480 bp, and a maximum of six ambiguous bases or no ambiguous base for paired-end reads from soil Uc or the other nine soils, respectively. Alignment of unique sequences was performed against the Silva database. Chimeras were detected using Uchime (Edgar et al. 2011) and removed. Taxonomy was assigned using the Silva 132 bacteria database (released in December 2017) using a cutoff of 80. Sequences affiliated to archaea, eukaryota, unknown, mitochondria and chloroplasts were removed for further analysis. Singletons (sequences appearing only once among all samples) were removed. Sequences were clustered in operational taxonomic units (OTUs) at 97% similarity. Finally, datasets were rarefied to the lowest number of sequences per sample (1096 or 20 764 reads/sample for sequences from Uc or the other nine soils, respectively). Using a method describe by Thomas et al. (2019), active PHE-degraders were determined as OTUs present in the pooled heavy fractions of the 13C-PHE incubation at a significantly higher abundance than in the pooled heavy fractions of the 12C-PHE control incubation. We first selected major OTUs (i) represented by at least five sequences in each 13C-PHE triplicate and (ii) with a higher average abundance in 13C-PHE samples than in 12C-PHE controls. After log-transformation, Welch's test followed by Benjamini–Hochberg correction of the P-values allowed determining the active PHE-degrading OTUs among major OTUs, separately for each soil. Statistically selected OTUs with exactly the same taxonomic affiliation were grouped and their sequence numbers were summed. For each soil, active PHE-degraders were reported with (i) their relative abundance related to the whole pool of PHE-degraders, (ii) their enrichment in 13C-PHE heavy fractions compared to 12C-PHE controls and (iii) the corresponding number of OTUs in the groups. Statistical analyses All statistical analyses were performed using RStudio v1.1.442. Significant differences in PHE degradation, 13C dissipation and 16S rRNA gene copy numbers among soils were assessed by using Kruskal–Wallis rank sum test and the multiple comparison test included in the agricolae R package (Mendiburu 2017). Significant differences between PHE degradation and 13C dissipation in a given soil were assessed using Welch's test. Linear correlations between all data were explored using the rcorr function included in the Hmisc package in R (Harell and Dupont 2018). Linear multiple regressions and factor interactions were explored using the lm function. Diversity indices were calculated using the specnumber and diversity functions included in the vegan package in R (Oksanen et al. 2017). RESULTS PAH contamination and PAH-dioxygenase genes of initial soils The initial PAH and phenanthrene contents are presented in Table 1, together with PAH-RHDα gene relative abundances in the 10 initial unspiked soils. Soil collection showed a gradient of total PAH contents ranging from 0.03 (He) to 1095.90 mg kg−1 (NM) with ∼1–10% of available PAHs (Lemmel et al. 2019). The initial unspiked soil PHE content followed a similar trend and ranged from below detection limit (He) to 199.24 mg kg−1 (Ho). We can note that highest initial available PAH and PHE contents were measured in Ho (54.14 and 18.10 mg kg−1, respectively) and NM (82.98 and 15.77 mg kg−1, respectively) soils. Based on the PAH content, we considered three groups of soil displaying low (<20 mg kg−1), medium (20–200 mg kg−1) or high (>200 mg kg−1) PAH contamination. Interestingly, the relative abundance of PAH-RHDα genes in initial unspiked soils showed linear Pearson correlation with total soil PAH contents (P = 2.62 × 10−4,r = 0.91 and P = 3.38 × 10−8, r = 0.99 for the PAH-RHDα-GN and -GP genes, respectively) and available ones (P = 5.77 × 10−5, r = 0.96 and P = 3.70 × 10−6, r = 0.98 for the PAH-RHDα-GN and -GP genes, respectively). Initial soil respiration Depending on the soil activity, BR and respiration induced after phenanthrene addition (PHE-SIR) were monitored for 17 or 23 days (Figure S2, Supporting Information). Soils Mo, Te and Po had the highest BR activity. A globally higher CO2 production was observed in the PHE-SIR flasks than in the BR flasks, except for soils He and Po. Soils Te, Di and Mo had the highest PHE-SIR level as compared to their BR, with the difference in CO2 production between BR and PHE-SIR representing at the end >200 µgC per gram of soil possibly indicating the complete mineralization of the PHE (Figure S2, Supporting Information). The Uc, MSM, NM, RM and Ho soils have lower CO2 production representing between 90% (Uc) and 18% (Ho) of the corresponding C added through PHE spiking. Globally, after 10–12 days of incubation, most of the soils showed a high difference in CO2 production between the two conditions, indicating that most of the spiked phenanthrene had been degraded. Based on these results, we decided to stop SIP incubation after 12 days. 13C-Phenanthrene degradation in the soils We measured PHE concentrations and δ13C levels to estimate 13C-PHE degradation and 13C dissipation in the soils at the end of SIP incubation (Fig. 1). The soils presented variable percentages of 13C-PHE degradation, ranging from 17.9% (RM) to 91.2% (Di) of the initial quantity of spiked 13C-PHE. Soils Di and Po presented the highest PHE degradation, while soils RM and NM presented the lowest (Kruskal–Wallis tests, P = 0.002). There was no correlation (Pearson) between the level of 13C-PHE degradation and the soil characteristics (Table S2, Supporting Information) and no correlation between 13C-PHE degradation and the initial PAH and PHE contents (P-values of 0.21 and 0.59, respectively). Besides, medium PAH-polluted soils (Po, MsM, Te) presented higher 13C-PHE degradation than low-contaminated soils (He, Uc, Mo, RM, except Di) (Kruskal–Wallis test, P = 0.002). Based on the soil δ13C levels, we calculated 13C dissipation in the soils; it ranged from 18.8% (NM) to 59.5% (Di) of the initial 13C quantity, with significant differences (Kruskal–Wallis test, P = 0.002). No correlation was found between 13C dissipation and the soil characteristics, including the initial PAH and PHE contents (P-values of 0.16 and 0.52, respectively). Interestingly, the percentage of degraded PHE showed a positive linear correlation with both initial soil bacterial abundance (16S gene copy number; P = 0.0001, R = 0.65) and initial soil bacterial richness (Chao1 index; P = 0.0002, R = 0.62) (Fig. 2A and B). A strong linear relationship (P = 0.0004; R = 0.90) was also found between the percentage of dissipated 13C and degraded 13C-PHE. The difference between the percentages of dissipated 13C and degraded 13C-PHE (Fig. 1) indicates the amount of 13C still present in the soil but not attributed to the remaining PHE (i.e. mainly 13C-labeled microbial biomass) at the end of the incubations. Five soils (Di, Po, MsM, Te and Ho) presented significantly higher PHE degradation than 13C dissipation (Welch's test, P < 0.05), and this difference was positively correlated to the initial soil bacterial abundance (16S rRNA gene copy number) in the soils (Pearson; P = 0.01; R = 0.76). Figure 1. Open in new tabDownload slide Percentages of dissipated 13C and degraded PHE at the end of SIP incubations (after 12 days). Error bars represent standard errors of the mean (n = 3). Statistical differences (Kruskal–Wallis test, P < 0.05) of dissipated 13C and degraded PHE among the soils are indicated by capital letters (different letters indicate significant differences). Significant difference (Welch's test) between degraded PHE and dissipated 13C for one soil is indicated by asterisks, with *, ** and *** corresponding to P < 0.05, P < 0.01 and P < 0.001, respectively. Soils were ranked from left to right according to their total PAH concentration. Figure 1. Open in new tabDownload slide Percentages of dissipated 13C and degraded PHE at the end of SIP incubations (after 12 days). Error bars represent standard errors of the mean (n = 3). Statistical differences (Kruskal–Wallis test, P < 0.05) of dissipated 13C and degraded PHE among the soils are indicated by capital letters (different letters indicate significant differences). Significant difference (Welch's test) between degraded PHE and dissipated 13C for one soil is indicated by asterisks, with *, ** and *** corresponding to P < 0.05, P < 0.01 and P < 0.001, respectively. Soils were ranked from left to right according to their total PAH concentration. Figure 2. Open in new tabDownload slide Relations between the percentage of degraded PHE and biotic factors for unspiked soils or PHE-degrader characteristics found using SIP. Correlation with total bacterial community Chao1 index (A) and bacterial abundance (B) of unspiked soils, PHE-degrader Shannon diversity index (C) and relative abundance of Mycobacterium taxa among PHE degraders (D). No significant linear correlation was detected with PHE-degraders richness (E). Figure 2. Open in new tabDownload slide Relations between the percentage of degraded PHE and biotic factors for unspiked soils or PHE-degrader characteristics found using SIP. Correlation with total bacterial community Chao1 index (A) and bacterial abundance (B) of unspiked soils, PHE-degrader Shannon diversity index (C) and relative abundance of Mycobacterium taxa among PHE degraders (D). No significant linear correlation was detected with PHE-degraders richness (E). Evidence of 13C-labeled DNA Bacterial 16S rRNA gene abundance was quantified from the DNA fractions obtained after ultracentrifugation of gDNA from the 12C-PHE (control) and 13C-PHE microcosms (Fig. 3). The ‘heavy’ DNA was in the fractions with a buoyant density ranging from 1.717 to 1.725 g ml−1, while most of the ‘light’ DNA corresponded to the fractions with a buoyant density ranging from 1.699 to 1.704 g ml−1. The numbers of 16S rRNA gene copies were higher in the heavy DNA fractions collected from the 13C microcosms than in the 12C controls, highlighting the presence of 13C-labeled DNA belonging to the active 13C-PHE-degrading bacteria. This 13C-DNA enrichment was observed for all soils but Mo, and only a weak signal was observed for soil RM. 18S rRNA gene copies were also quantified in all fractions (Figure S3, Supporting Information), but no fungal 13C-DNA enrichment was detected, indicating that fungi were not significantly involved in PHE degradation or did not use carbon from PHE for their growth, as they did not incorporate 13C into their DNA. Figure 3. Open in new tabDownload slide Quantification of bacterial 16S rRNA gene copy numbers in the 11 fractions (fractions #1 and #13 were discarded) separated by CsCl gradients using real-time quantitative PCR in the 12C (blue) and 13C (orange) microcosms. Data are presented for the 10 soils of the collection. Error bars represent standard errors of the mean (n = 3). Light and dark gray bars represent the light and heavy fractions used for PAH RDHAα gene quantification and enrichment calculations, respectively. Heavy DNAs corresponding to fractions 4 to 6 (highlighted by a dark gray bar) were selected for DNA sequencing. Figure 3. Open in new tabDownload slide Quantification of bacterial 16S rRNA gene copy numbers in the 11 fractions (fractions #1 and #13 were discarded) separated by CsCl gradients using real-time quantitative PCR in the 12C (blue) and 13C (orange) microcosms. Data are presented for the 10 soils of the collection. Error bars represent standard errors of the mean (n = 3). Light and dark gray bars represent the light and heavy fractions used for PAH RDHAα gene quantification and enrichment calculations, respectively. Heavy DNAs corresponding to fractions 4 to 6 (highlighted by a dark gray bar) were selected for DNA sequencing. In addition, PAH-RHDα-GN and -GP genes were also quantified in the DNA recovered from the CsCl gradient fractions to calculate their relative abundance in the soils after incubation and in the sequenced heavy 13C-DNA fractions (Table 1). By comparing PAH-RHDα-GP and -GN abundances in the soils before and after SIP incubation, we calculated a value of enrichment during incubation (Table 1), ranging from 4 (NM) to 221 (Di) and from 3 (NM) to 126 (Te) for PAH-RHDα-GP and -GN, respectively. Interestingly, PAH-RHDα-GP gene abundance and PAH-RHDα-GP gene enrichment in the soils after SIP incubation were positively correlated to both the percentages of degraded PHE and dissipated 13C during SIP incubation (Pearson; P = 0.03, R = 0.68, and P = 0.01, R = 0.75, respectively, for abundance; and P = 0.005, R = 0.81, and P = 0.0006, R = 0.89, respectively, for enrichment). Additionally, PAH-RHDα-GP gene abundance in the heavy 13C-DNA fractions was positively correlated to the percentage of 13C dissipated during SIP incubation (Pearson; P = 0.02, R = 0.72). No correlation was found for PAH-RHDα-GN gene abundance in the soils after SIP incubation or in the 13C-heavy DNA fractions, or for PAH-RHDα-GN gene enrichment. Identity of PHE degraders Based on the results of the quantification of bacterial 16S rRNA genes, the pooled heavy DNA fractions (Fig. 3) were used as a template to prepare a 16S rRNA gene amplicon library for Illumina sequencing. Due to the altered sequence quality of soil Uc as compared to the other nine soils, two separate analyses were performed, with Uc reads treated separately from all the other soils. Totals of 7 150 234 and 405 380 reads were obtained after sequencing, and 3 485 709 and 22 402 reads were kept after the trimming steps for the nine soils and soil Uc (in triplicates and from 12C and 13C incubations), respectively. Finally, data were rarefied to 20 764 and 1096 reads/sample for the nine soils and soil Uc, respectively. Active PHE-degraders were determined as OTU present in the pooled heavy fractions of the 13C-PHE incubation at a significantly higher abundance than in the pooled heavy fractions of the corresponding 12C-PHE controls. Then, OTU corresponding to active PHE-degraders and sharing the same taxonomic affiliation were gathered in groups. Relative abundance (relative to the whole pool of PHE degraders) and enrichment (during 13C-SIP incubation) are reported in Fig. 4 for each PHE-degrader group in the corresponding soil. Only seven soils were reported in Fig. 4 as no OTU was detected with a significantly higher abundance in 13C- than 12C- heavy DNA fractions in He, Mo and RM soils, confirming the fact that we could not detect 16S rRNA gene copies number enrichment in heavy fractions of the 13C-PHE incubations compared to the 12C-PHE controls (Fig. 3). A total of 2 (NM) to 12 (Di and Te) OTU groups (representing 3 to 30 OTUs) were highlighted as active PHE degraders. Most of these OTU groups were detected in 13C-DNA fractions with enrichment higher than 10 times compared to the 12C-controls. Active PHE degraders were relatively diverse and belonged to the Actinobacteria, Proteobacteria (α and β), Firmicutes, Patescibacteria and Cyanobacteria phyla; the last three were in minority compared to other phyla. Depending of the soil, dominant PHE degraders were mainly affiliated to Micrococcaceae, Sphingomonadaceae and Burkholderiaceae family, Mycobacterium, Allorhizobium–Neorhizobium–Pararhizobium–Rhizobium (ANPR) and Massilia genera, and Arthobacter crystallopoietes, Rhizobium rhizogenes species. Altogether, those eight taxonomic groups represented 78% (Te) to 100% (MsM) on average of the whole active PHE degraders in the corresponding soil. Among Actinobacteria, OTUs affiliated to Mycobacterium were the major PHE degraders in soils Di (55%), Po (98%) and MsM (57%), while Arthrobacter crystallopoietes and Micrococcaceae family were the major PHE degraders in soils Te (44%) and NM (>99%), respectively. Besides, Mycobacterium affiliated OTUs were in the minority among the PHE degraders in both soils Te (20%) and Ho (4%). Concerning α-Proteobacteria, the ANPR, Rhizobium rhizogenes and Sphingomonadaceae family represented the second major PHE-degrader group in soils Di (34%), Uc (44%) and MsM (43%), respectively. Finally, among β-Proteobacteria, Burkholderiaceae family members were the major PHE degraders in soil Uc (51%), and more specifically the Massilia genus in soil Ho (87%). Figure 4. Open in new tabDownload slide Active 13C-labeled PHE-degrading taxa identified in each soil, based on differential OTU abundance in heavy-DNA fractions between 13C and 12C-PHE incubations. OTUs having exactly the same taxonomic affiliation (best classification) were summed as taxa groups. Their lowest possible classifications are given with the prefix corresponding to taxonomic rank (c = class, o = order, f = family, g = genus, no prefix = species). For each active PHE-degrading taxon in one corresponding soil, Ra indicates the abundance relative to the whole pool of active PHE degraders identified through color gradient, E indicates the enrichment factor of each taxon in 13C-PHE compared to 12C-PHE control through circle size (with the infinity symbol meaning that OTUs were detected only in 13C- and not in 12C-heavy DNA fractions) and N indicates the corresponding number of OTUs in the taxa groups. OTU groups originally classified as the Betaproteobacteriales order in the γ-Proteobacteria class (in Silva V132 database) are indicated here as β-Proteobacteria. Soils were ranked from left to right according to their total initial PAH concentration. Figure 4. Open in new tabDownload slide Active 13C-labeled PHE-degrading taxa identified in each soil, based on differential OTU abundance in heavy-DNA fractions between 13C and 12C-PHE incubations. OTUs having exactly the same taxonomic affiliation (best classification) were summed as taxa groups. Their lowest possible classifications are given with the prefix corresponding to taxonomic rank (c = class, o = order, f = family, g = genus, no prefix = species). For each active PHE-degrading taxon in one corresponding soil, Ra indicates the abundance relative to the whole pool of active PHE degraders identified through color gradient, E indicates the enrichment factor of each taxon in 13C-PHE compared to 12C-PHE control through circle size (with the infinity symbol meaning that OTUs were detected only in 13C- and not in 12C-heavy DNA fractions) and N indicates the corresponding number of OTUs in the taxa groups. OTU groups originally classified as the Betaproteobacteriales order in the γ-Proteobacteria class (in Silva V132 database) are indicated here as β-Proteobacteria. Soils were ranked from left to right according to their total initial PAH concentration. The percentage of PHE degraded was positively correlated to the relative abundance of the globally most abundant PHE degrader, namely Mycobacterium (P = 6.1 × 10−5, R = 0.76; Fig. 2D). For each soil replicate, we calculated diversity indices (i.e. species richness, Shannon diversity and Pielou's evenness; Table S3, Supporting Information) to describe the PHE-degrader community. Shannon diversity index of PHE-degraders was positively correlated to the percentage of degraded PHE (P = 0.02, R = 0.48; Fig. 2C). The percentage of degraded PHE was not linearly correlated with the specific richness of PHE-degraders (Fig. 2E). However, this relationship could be modeled by a saturation curve, suggesting a functional redundancy in the PHE-degrader community (beyond a certain species richness, several species have the same role in PHE degradation). DISCUSSION We evaluated the level of phenanthrene degradation and tried to identify the phenanthrene degraders in the 10 soils presenting a PAH pollution gradient. We hypothesized that the microbial communities of soils with a high and historical PAH contamination would be better adapted to PAH degradation than the ones from low- or non-contaminated soils. Consequently, the phenanthrene degradation rate would be positively correlated to the level of soil PAH contamination. Previous works indeed showed that greater PHE or PAH mineralization appeared in soils pre-exposed to PAHs or with the highest PAH contamination level as compared to low- or non-contaminated soils (Johnsen and Karlson 2005; Carmichael and Pfaender 2009). In the present study, we did not show any correlation between phenanthrene degradation rates and initial PAH or PHE contamination levels. Pre-exposure to PAH does not seem to be essential for efficient phenanthrene degradation. Nevertheless, this hypothesis was partly confirmed by the positive linear correlation between PAH-dioxygenase (both PAH-RHDα-GP and -GN) gene abundances and the total PAH contamination levels of the 10 soils. Our finding suggests that the selective pressure exerted by PAH pollution led to an enrichment in bacteria capable of PAH degradation. PAH-RHDα gene abundance is indeed a good indicator of the PAH contamination level in environmental samples (Cébron et al. 2008). However, although PHE degradation was surprisingly detected in all the tested soils, the PHE degradation rates were highly contrasted among soils and did not correlate with the PAH concentration or the PAH-dioxygenase gene abundance of the unspiked soils. Yet, although the moderately contaminated soil Di displayed higher PHE degradation than the non-contaminated ones, the highly contaminated soil NM did not display the highest PHE degradation rate. These results show that detecting the potential function (i.e. PAH-RHDα gene abundance) is not the sole predictor of the actual measured activity (i.e. PHE degradation) in the 10 soils. These differences in PHE biodegradation activity among soils could be explained by a myriad of abiotic and biotic factors specific to each soil, such as texture (Haghollahi, Fazaelipoor and Schaffie 2016), nutrient content (Breedveld and Sparrevik 2000), pH (Kästner, Breuer-Jammali and Mahro 1998), bioavailability of pollutants (Biache et al. 2017), multi-contamination (i.e. metal pollution; Sandrin and Maier 2003) and microbial diversity (Cébron et al. 2008). Surprisingly, no correlation was found between the percentage of degraded PHE and any of the soil characteristics. No impact of the nitrogen or phosphorus contents or of the C:N ratio on the PHE degradation level was highlighted, even if it is well established that low nutrient levels limit microbial growth and activity (Wardle 1992) and hydrocarbon degradation (Leahy and Colwell 1990). C:N:P ratios could indeed drive PHE degradation (Smith, Graham and Cleland 1998) and the selection of active microorganisms in SIP experiments (Cébron et al. 2007); besides, adding N or P can increase the soil respiration (Breedveld and Sparrevik 2000) and improve the biodegradation of crude oil or gasoline (Leahy and Colwell 1990) and of PAHs (Joner et al. 2002; Jones et al. 2008) in soils. Moreover, one of the main drivers of soil PAH degradation is PAH availability, mostly impacted by the soil texture (Carmichael and Pfaender 2009), organic matter type and content (Dictor et al. 2003) and the presence of aged pollution (Biache et al. 2008). Dictor et al. (2003) showed that even after a short period (i.e. few hours), a significant part of the spiked PHE became unavailable for soil microorganisms, resulting in reduced degradation. We can thus hypothesize that spiked PHE availability differed depending on the soils. The PHE degradation rate was well predicted from the mineralization data of soil spiked with PHE compared to the basal mineralization, except for the soil Mo where we overestimated it from mineralization pre-experiments. For Ho and NM soil, where the initial PAH contamination was the highest, the difference in mineralization was low compared to the PHE degradation rates. This phenomenon could be due to the adaptation of bacterial communities in these aged PAH contaminated soils, leading to a shift in mineralization from the aged carbon (mostly PAHs) to the fresh added PHE by the same bacteria. We also hypothesized that differences in PHE degradation rates between soils may be explained by the diversity of PHE degraders and/or by the presence of specific taxa with a higher PHE-degrading activity. In addition to the analysis of the diversity of the whole bacterial community in unspiked soils (Lemmel et al. 2019), the use of SIP allowed us to directly link the 13C-PHE degradation function with the diversity of 13C-labeled PHE degraders. We get a very original result showing that the PHE degradation level was positively correlated to both bacterial abundance and bacterial richness of the initial soils, previously measured on unspiked soils (Lemmel et al. 2019). This relationship could be explained by (i) the fact that soils harboring a more abundant and diversified microbial community might contain more PHE-degrading bacteria (the sampling effect hypothesis, e.g. Hector et al. 2002) resulting in higher PHE degradation rates or (ii) facilitation between species improving PHE degradation. As broadly described in plant community ecology (Hooper and Vitousek 1998; Caldeira et al. 2001), facilitation between species enhances an ecosystem function and may be the mechanism responsible for the positive relationship between taxonomic diversity and ecosystem processes. We can consider that some bacterial species indirectly contributed to increase PHE degradation by facilitating the PHE-degrader activity in the soil (for example, by increasing their access to essential elements or by reducing the toxicity due to a high soil metal content). Our finding is in agreement with previous finding in microbial ecology. Jung, Philippot and Park (2016) showed that a large proportion of functional gene categories were significantly altered by a reduction in microbial biodiversity, even if these authors also found that the efficiency of diesel biodegradation was increased in the low-diversity community suggesting that the relationship between microbial diversity and ecological function involves trade-offs among ecological processes. PHE degradation activity was also positively correlated with the active PHE-degrader Shannon diversity index calculated from SIP data, but no correlation was found with PHE-degrader richness. Complementarity between species for the use of PHE should be considered here. When the diversity of PHE-degrading species increases, the use of PHE as a carbon substrate is optimized and a greater amount of PHE is degraded. For example, PHE degradation results from a sum of reactions (Ghosal et al. 2016), and when the diversity of PHE degraders increases, the joint functioning of the different species ensures a greater number of its reactions and thus maximizes PHE degradation. Thomas, Corre and Cébron (2019) recently demonstrated, using SIP combined to metagenomic analysis, that active PHE degraders act in a consortium, whereby complete PHE mineralization is achieved through the combined activity of taxonomically diverse co-occurring bacteria performing successive metabolic steps. Similarly, positive relationships between species diversity and soil respiration (Bell et al. 2005) or the efficiency of hydrocarbon biodegradation in marine sediment (Dell'Anno et al. 2012) have been observed. All these observations refer to well-known ecological concept where increased microbial diversity corresponds to increased catabolic potential and, hence, to better removal of metabolites and pollutants as explained by Dejonghe et al. (2001). The positive correlation found with Shannon diversity index, and not with PHE-degrader species richness, tends to show that PHE degradation efficiency depends of the presence of keystone species, that if they are absent, greatly decreases the biodegradation rate. Indeed, below 10 active PHE-degrading OTUs, the PHE degradation rate is low, but when at least 10 OTUs are active in even proportion the PHE degradation is higher, leading to a more efficient PHE removal. This phenomenon was previously observed for other organic pollutants such as atrazine (Monard et al. 2011). Through labeling of microbes metabolizing PHE and their degradation by-products, DNA-SIP allowed us identifying the active PHE-degraders in the studied soils. For three soils (He, Mo and RM), apparently, the 13C-PHE degradation was too low and PHE-degraders growth too weak to generate a sufficient 13C-labeling of the microbial biomass. Indeed, we could not detect significant enrichment of 13C-DNA in heavy fractions of the CsCl gradients and, even after sequencing of the heavy DNA fractions, no OTUs could be detected as significantly enriched in the 13C-PHE condition compared to the 12C controls. When looking at the seven other soils, a large number of bacterial OTUs were identified as PHE degraders (from 3 to 30 OTUs per soil representing 31 taxonomic groups). We found that the presence of some taxa seems to maximize the PHE degradation rate. Among them, the Mycobacterium, Massilia, Arthrobacter, ANPR and Rhizobium genera, and unclassified OTUs belonging to Sphingomonadaceae, Burkholderiaceae and Micrococcaceae, seemed to be the main PHE degraders best explaining the PHE degradation rates in the seven soils. PHE-degrading Massilia were isolated from soils after PHE exposure (Bodour et al. 2003; Zhang et al. 2010), and from PAH-polluted soils (Baquiran et al. 2012; Wang et al. 2016), or were detected in an oilfield by metagenomics (Zhou et al. 2017). Although no detailed information is available about the role and potential of Massilia sp. in PAH degradation in environmental samples, one study showed that Massilia sp. strain WF1 had high degradation ability and tolerance to PHE (Gu et al. 2016). Additionally, Massilia seems to also tolerate metal contamination (Zhang et al. 2016). Micrococcaceae and among which Arthrobacter genus has been shown for its PAH degradation capacity both in pure culture (Aryal and Liakopoulou-Kyriakides 2013) and in the environment (Thion et al. 2012b). Similarly to our result, a recent study of Storey et al. (2018) also showed a strong increase of one unclassified OTU affiliated to Micrococcaceae family, being dominant in a soil after PHE amendment. ANPR genera, which relate to N-fixing bacteria through symbiosis with legumes, contain various Rhizobium species described for their capacity to tolerate and degrade PAHs (Poonthrigpun et al. 2006; González-Paredes et al. 2013), and have been used to enhance PAH phytoremediation (Johnson, Anderson and McGrath 2005; Teng et al. 2011). Members of the Sphingomonadaceae family are also well-known PAH degraders (Thomas, Corre and Cébron 2019) and many isolates belonging to Sphingomonadaceae were frequently isolated from PAH-contaminated environments (Johnsen, Wick and Harms 2005). Similarly, Burkholderiaceae family is known to contain various bacteria able to grow on PAHs or having PAH degradation genes, such as members of Burkholderia and Ralstonia genera (Fuenmayor et al. 1998; Cébron et al. 2008; Andreolli et al. 2011), and recently members of Cupriavidus genus (Kuppusamy et al. 2016; Oyehan and Al-Thukair 2017). OTUs affiliated to the Mycobacterium (Actinobacteria) genus were the dominant PHE degraders in the soils displaying the higher PHE degradation rates. The PHE degradation rate was positively correlated to the relative abundance of OTUs affiliated to Mycobacterium. This relationship confirmed that Mycobacterium strains were essential drivers of PHE degradation in our soils and could be defined as a key stone species for efficient PHE-degradation. Due to their prevalence in many PAH-contaminated soils, Mycobacterium species have been suggested to potentially play a major role in the natural attenuation of PAHs (Cheung and Kinkle 2001; Chen, Peng and Duan 2016; Chen et al. 2018). Numerous Mycobacterium species have indeed been described for their ability to degrade various low- and high-molecular-weight PAHs, in PAH-contaminated soils (Chen et al. 2018; Li et al. 2018). Previous studies suggest that PAH-degrading Mycobacterium species are well adapted to oligotrophic and low PAH availability conditions, usually found in PAH-contaminated soils. Mycobacterium can degrade PAHs under optimal conditions (C:N:P ratio of 100:10:1) as well as under N and P excess or deficiency (Leys et al. 2005). Additionally, Miyata et al. (2004) suggested that both passive diffusion and high-affinity transport systems contributed to the PHE uptake by Mycobacterium strain RJGII-135, and enabled the bacterium to use aqueous-phase PHE in high and low concentrations. Moreover, Mycobacterium strains can attach on the PAH source and form biofilms that increase PAH bioavailability (Bastiaens et al. 2000; Wick et al. 2002). Given that Mycobacterium strains were highlighted as main active PHE degraders in our soils presenting high PHE degradation rate, and that numerous studies describe Mycobacterium strains with PAH degradation capacity, their presence could constitute a relevant indicator of PAH degradation. Further, they might be of interest for bioaugmentation of PAH degradation in contaminated environments. CONCLUSION For the first time, we studied in parallel and through SIP the degradation of 13C-phenanthrene in a collection of 10 anthropized soils presenting a PAH pollution gradient and identified major active phenanthrene degraders for 7 soils of the collection. Surprisingly, the phenanthrene degradation rate was neither correlated to the initial soil PAH contamination level nor to the functional PAH-degradation potential (occurrence of PAH-RHDα genes) but was best explained by the initial soil bacterial community richness and abundance. No specific taxa, representative of aged and highly contaminated soils, were identified. In the studied soils, the phenanthrene degradation efficiency depended on two parameters: (i) to have a wide diversity of active phenanthrene degraders (correlation between PHE degradation and Shannon diversity index of active taxa) and (ii) to have representatives of the Mycobacterium genus as dominant phenanthrene degrading taxa (correlation between PHE degradation and Mycobacterium occurrence). To conclude, our study shows that not only microbial community characteristics and the presence of specific keystone species (Mycobacterium) but also species complementarity are more important for PAH degradation efficiency than initial levels of soil PAH contamination and the initial PAH-degradation potential. ACKNOWLEDGEMENTS This work was supported by the French national program EC2CO (Ecobios project) and the OSU-OteLo (TraitMic project). We thank Dr S. Uroz (Labex Arbre, INRA Champenoux) for giving us access to the ultracentrifuge equipment. We would like to thank ArcelorMittal, EPFL, GISFI, ONF and LTO of Montiers (ANDRA/INRA, M.P. Turpault) for giving us access to the different sampling sites. We would like to thank C. Friry, G. Kitzinger and D. Billet (LIEC, Nancy, France) for technical assistance. Conflicts of interest. None declared. REFERENCES Abdel-Shafy HI , Mansour MSM. A review on polycyclic aromatic hydrocarbons: source, environmental impact, effect on human health and remediation . Egypt J Pet . 2016 ; 25 : 107 – 23 . Google Scholar Crossref Search ADS WorldCat Andreolli M , Lampis S , Zenaro E et al. . Burkholderia fungorum DBT1: a promising bacterial strain for bioremediation of PAH-contaminated soils . FEMS Microbiol Lett . 2011 ; 319 : 11 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Aryal M , Liakopoulou-Kyriakides M. Biodegradation and kinetics of phenanthrene and pyrene in the presence of nonionic surfactants by Arthrobacter strain Sphe3 . Water Air Soil Poll . 2013 ; 224 : 1426 . Google Scholar Crossref Search ADS WorldCat Baquiran JP , Thater B , Songco K et al. . Characterization of culturable PAH and BTEX degrading bacteria from heavy oil of the Rancho La Brea tarpits . Polycycl Aromat Comp . 2012 ; 32 : 600 – 14 . Google Scholar Crossref Search ADS WorldCat Bastiaens L , Springael D , Wattiau P et al. . Isolation of adherent polycyclic aromatic hydrocarbon (PAH)-degrading bacteria using PAH-sorbing carriers . Appl Environ Microb . 2000 ; 66 : 1834 – 43 . Google Scholar Crossref Search ADS WorldCat Bell T , Newman JA , Silverman BW et al. . The contribution of species richness and composition to bacterial services . Nature . 2005 ; 436 : 1157 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat Biache C , Mansuy-Huault L , Faure P et al. . Effects of thermal desorption on the composition of two coking plant soils: impact on solvent extractable organic compounds and metal bioavailability . Environ Pollut . 2008 ; 156 : 671 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Biache C , Ouali S , Cébron A et al. . Bioremediation of PAH-contaminated soils: consequences on formation and degradation of polar-polycyclic aromatic compounds and microbial community abundance . J Hazard Mater . 2017 ; 329 : 1 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat Bodour AA , Wang JM , Brusseau ML et al. . Temporal change in culturable phenanthrene degraders in response to long-term exposure to phenanthrene in a soil column system . Environ Microbiol . 2003 ; 5 : 888 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat Bourceret A , Cébron A , Tisserant E et al. . The bacterial and fungal diversity of an aged PAH- and heavy metal-contaminated soil is affected by plant cover and edaphic parameters . Microb Ecol . 2016 ; 71 : 711 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat Breedveld GD , Sparrevik M. Nutrient-limited biodegradation of PAH in various soil strata at a creosote-contaminated site . Biodegradation . 2000 ; 11 : 391 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Caldeira MC , Ryel RJ , Lawton JH et al. . Mechanisms of positive biodiversity–production relationships: insights provided by δ13C analysis in experimental Mediterranean grassland plots . Ecol Lett . 2001 ; 4 : 439 – 43 . Google Scholar Crossref Search ADS WorldCat Caporaso JG , Lauber CL , Walters WA et al. . Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample . Proc Natl Acad Sci USA . 2011 ; 108 : 4516 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat Carmichael LM , Pfaender FK . Polynuclear aromatic hydrocarbon metabolism in soils: relationship to soil characteristics and preexposure . Environ Toxicol Chem . 2009 ; 16 : 666 – 75 . Google Scholar Crossref Search ADS WorldCat Cennerazzo J , de Junet A , Audinot J-N et al. . Dynamics of PAHs and derived organic compounds in a soil-plant mesocosm spiked with 13C-phenanthrene . Chemosphere . 2017 ; 168 : 1619 – 27 . Google Scholar Crossref Search ADS PubMed WorldCat Chauhan A , Fazlurrahman , Oakeshott JG et al. . Bacterial metabolism of polycyclic aromatic hydrocarbons: strategies for bioremediation . Indian J Microbiol . 2008 ; 48 : 95 – 113 . Google Scholar Crossref Search ADS PubMed WorldCat Chen S , Peng J , Duan G . Enrichment of functional microbes and genes during pyrene degradation in two different soils . J Soils Sediments . 2016 ; 16 : 417 – 26 . Google Scholar Crossref Search ADS WorldCat Chen S-C , Duan G-L , Ding K et al. . DNA stable-isotope probing identifies uncultivated members of Pseudonocardia associated with biodegradation of pyrene in agricultural soil . Fems Microbiol Ecol . 2018 ; 94 : fiy026 . Google Scholar Crossref Search ADS WorldCat Cheung P-Y , Kinkle BK . Mycobacterium diversity and pyrene mineralization in petroleum-contaminated soils . Appl Environ Microb . 2001 ; 67 : 2222 – 9 . Google Scholar Crossref Search ADS WorldCat Crampon M , Cebron A , Portet-Koltalo F et al. . Low effect of phenanthrene bioaccessibility on its biodegradation in diffusely contaminated soil . Environ Pollut . 2017 ; 225 : 663 – 73 . Google Scholar Crossref Search ADS PubMed WorldCat Cébron A , Bodrossy L , Stralis-Pavese N et al. . Nutrient amendments in soil DNA stable isotope probing experiments reduce the observed methanotroph diversity . Appl Environ Microb . 2007 ; 73 : 798 – 807 . Google Scholar Crossref Search ADS WorldCat Cébron A , Louvel B , Faure P et al. . Root exudates modify bacterial diversity of phenanthrene degraders in PAH-polluted soil but not phenanthrene degradation rates . Environ Microbiol . 2011 ; 13 : 722 – 36 . Google Scholar Crossref Search ADS PubMed WorldCat Cébron A , Norini M-P , Beguiristain T et al. . Real-time PCR quantification of PAH-ring hydroxylating dioxygenase (PAH-RHDα) genes from Gram-positive and Gram-negative bacteria in soil and sediment samples . J Microbiol Meth . 2008 ; 73 : 148 – 59 . Google Scholar Crossref Search ADS WorldCat Dejonghe W , Boon N , Seghers D et al. . Bioaugmentation of soils by increasing microbial richness: missing links . Environ Microbiol . 2001 ; 3 : 649 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat Dell'Anno A , Beolchini F , Rocchetti L et al. . High bacterial biodiversity increases degradation performance of hydrocarbons during bioremediation of contaminated harbor marine sediments . Environ Pollut . 2012 ; 167 : 85 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat Dictor MC , Berne N , Mathieu O et al. . Influence of ageing of polluted soils on bioavailability of phenanthrene . Oil Gas Sci Technol . 2003 ; 58 : 481 – 8 . Google Scholar Crossref Search ADS WorldCat Dumont MG , Murrell JC . Stable isotope probing—linking microbial identity to function . Nat Rev Microbiol . 2005 ; 3 : 499 – 504 . Google Scholar Crossref Search ADS PubMed WorldCat Dumont MG , Radajewski SM , Miguez CB et al. . Identification of a complete methane monooxygenase operon from soil by combining stable isotope probing and metagenomic analysis . Environ Microbiol . 2006 ; 8 : 1240 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat Edgar RC , Haas BJ , Clemente JC et al. . UCHIME improves sensitivity and speed of chimera detection . Bioinforma Oxf Engl . 2011 ; 27 : 2194 – 200 . Google Scholar Crossref Search ADS WorldCat Eom IC , Rast C , Veber AM et al. . Ecotoxicity of a polycyclic aromatic hydrocarbon (PAH)-contaminated soil . Ecotox Environ Safe . 2007 ; 67 : 190 – 205 . Google Scholar Crossref Search ADS WorldCat Felske A , Akkermans ADL , Vos WMD . Quantification of 16S rRNAs in complex bacterial communities by multiple competitive reverse transcription-PCR in temperature gradient gel electrophoresis fingerprints . Appl Environ Microb . 1998 ; 64 : 4581 – 7 . WorldCat Fuenmayor SL , Wild M , Boyes AL et al. . A gene cluster encoding steps in conversion of naphthalene to gentisate in Pseudomonas sp. strain U2 . J Bacteriol . 1998 ; 180 : 2522 – 30 . Google Scholar PubMed WorldCat Ghosal D , Ghosh S , Dutta TK et al. . Current state of knowledge in microbial degradation of polycyclic aromatic hydrocarbons (PAHs): a review . Front Microbiol . 2016 ; 7 : 1369 . Google Scholar PubMed WorldCat González-Paredes Y , Alarcón A , Almaraz Juan J et al. . Tolerance, growth and degradation of phenanthrene and benzo[a]pyrene by Rhizobium tropiciCIAT 899 in liquid culture medium . Appl Soil Ecol . 2013 ; 63 : 105 – 11 . Google Scholar Crossref Search ADS WorldCat Gu H , Lou J , Wang H et al. . Biodegradation, biosorption of phenanthrene and its trans-membrane transport by M assilia sp. WF1 and Phanerochaete chrysosporium . Front Microbiol . 2016 ; 7 : 38 . Google Scholar PubMed WorldCat Guo G , Tian F , Ding K et al. . Bacterial communities predominant in the degradation of C-13-labeled pyrene in red soil . Soil Sediment Contam . 2017 ; 26 : 709 – 21 . Google Scholar Crossref Search ADS WorldCat Haghollahi A , Fazaelipoor MH , Schaffie M . The effect of soil type on the bioremediation of petroleum contaminated soils . J Environ Manage . 2016 ; 180 : 197 – 201 . Google Scholar Crossref Search ADS PubMed WorldCat Harell FE , Dupont C . Hmisc: Harrell Miscellaneous . 2018 . Hector A , Bazeley‐White E , Loreau M et al. . Overyielding in grassland communities: testing the sampling effect hypothesis with replicated biodiversity experiments . Ecol Lett . 2002 ; 5 : 502 – 11 . Google Scholar Crossref Search ADS WorldCat He D , Zhang L , Dumont MG et al. . The response of methanotrophs to additions of either ammonium, nitrate or urea in alpine swamp meadow soil as revealed by stable isotope probing . FEMS Microbiol Ecol . 2019 ; 95 : fiz077 . Google Scholar Crossref Search ADS PubMed WorldCat Hooper DU , Vitousek PM . Effects of plant composition and diversity on nutrient cycling . Ecol Monogr . 1998 ; 68 : 121 – 49 . Google Scholar Crossref Search ADS WorldCat Johnsen AR , Karlson U . PAH degradation capacity of soil microbial communities: does it depend on PAH exposure? Microb Ecol . 2005 ; 50 : 488 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat Johnsen AR , Wick LY , Harms H . Principles of microbial PAH-degradation in soil . Environ Pollut . 2005 ; 133 : 71 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat Johnson DL , Anderson DR , McGrath SP . Soil microbial response during the phytoremediation of a PAH contaminated soil . Soil Biol Biochem . 2005 ; 37 : 2334 – 6 . Google Scholar Crossref Search ADS WorldCat Joner EJ , Corgié SC , Amellal N et al. . Nutritional constraints to degradation of polycyclic aromatic hydrocarbons in a simulated rhizosphere . Soil Biol Biochem . 2002 ; 34 : 859 – 64 . Google Scholar Crossref Search ADS WorldCat Jones MD , Singleton DR , Carstensen DP et al. . Effect of incubation conditions on the enrichment of pyrene-degrading bacteria identified by stable-isotope probing in an aged, PAH-contaminated soil . Microb Ecol . 2008 ; 56 : 341 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Jung J , Philippot L , Park W . Metagenomic and functional analyses of the consequences of reduction of bacterial diversity on soil functions and bioremediation in diesel-contaminated microcosms . Sci Rep . 2016 ; 6 : 23012 . Google Scholar Crossref Search ADS PubMed WorldCat Khawand ME , Crombie AT , Johnston A et al. . Isolation of isoprene degrading bacteria from soils,development of isoA gene probes and identification of the active isoprene degrading soil community using DNA-stable isotope probing . Environ Microbiol . 2016 ; 18 : 2743 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat Kozich JJ , Westcott SL , Baxter NT et al. . Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform . Appl Environ Microb . 2013 ; 79 : 5112 – 20 . Google Scholar Crossref Search ADS WorldCat Kuppusamy S , Thavamani P , Megharaj M et al. . Polyaromatic hydrocarbon (PAH) degradation potential of a new acid tolerant, diazotrophic P-solubilizing and heavy metal resistant bacterium Cupriavidus sp. MTS-7 isolated from long-term mixed contaminated soil . Chemosphere . 2016 ; 162 : 31 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Kästner M , Breuer-Jammali M , Mahro B . Impact of inoculation protocols, salinity, and pH on the degradation of polycyclic aromatic hydrocarbons (PAHs) and survival of PAH-degrading bacteria introduced into soil . Appl Environ Microb . 1998 ; 64 : 359 – 62 . WorldCat Leahy JG , Colwell RR . Microbial degradation of hydrocarbons in the environment . Microbiol Rev . 1990 ; 54 : 305 – 15 . Google Scholar PubMed WorldCat Lemmel F , Maunoury-Danger F , Fanesi A et al. . Soil properties and multi-pollution affect taxonomic and functional bacterial diversity in a range of french soils displaying an anthropisation gradient . Microb Ecol . 2019 ; 77 : 993 – 1013 . Google Scholar Crossref Search ADS PubMed WorldCat Leys NM , Bastiaens L , Verstraete W et al. . Influence of the carbon/nitrogen/phosphorus ratio on polycyclic aromatic hydrocarbon degradation by Mycobacterium and Sphingomonas in soil . Appl Microbiol Biotechnol . 2005 ; 66 : 726 – 36 . Google Scholar Crossref Search ADS PubMed WorldCat Liang Y , Li G , Nostrand JDV et al. . Microarray-based analysis of microbial functional diversity along an oil contamination gradient in oil field . FEMS Microbiol Ecol . 2009 ; 70 : 324 – 33 . Google Scholar Crossref Search ADS PubMed WorldCat Li J , Luo C , Zhang D et al. . Autochthonous bioaugmentation-modified bacterial diversity of phenanthrene degraders in PAH-contaminated wastewater as revealed by DNA-stable isotope probing . Environ Sci Technol . 2018 ; 52 : 2934 – 44 . Google Scholar Crossref Search ADS PubMed WorldCat Li X , Lin Z , Luo C et al. . Enhanced microbial degradation of pentachlorophenol from soil in the presence of earthworms: evidence of functional bacteria using DNA-stable isotope probing . Soil Biol Biochem . 2015 ; 81 : 168 – 77 . Google Scholar Crossref Search ADS WorldCat Lueders T , Wagner B , Claus P et al. . Stable isotope probing of rRNA and DNA reveals a dynamic methylotroph community and trophic interactions with fungi and protozoa in oxic rice field soil . Environ Microbiol . 2004 ; 6 : 60 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat Lv X , Kankan Z , Li H et al. . Niche partition of phenanthrene-degrading bacteria along a Phragmites australis rhizosphere gradient . Biol Fert Soils . 2018 ; 54 : 607 – 16 . Google Scholar Crossref Search ADS WorldCat Mendiburu F de . Agricolae: Statistical Procedures for Agricultural Research . 2017 . Google Preview WorldCat COPAC Miyata N , Iwahori K , Foght JM et al. . Saturable, energy-dependent uptake of phenanthrene in aqueous phase by Mycobacterium sp. strain RJGII-135 . Appl Environ Microb . 2004 ; 70 : 363 – 9 . Google Scholar Crossref Search ADS WorldCat Monard C , Vandenkoornhuyse P , Le Bot B et al. . Relationship between bacterial diversity and function under biotic control: the soil pesticide degraders as a case study . Isme J . 2011 ; 5 : 1048 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat Mosbæk F , Kjeldal H , Mulat DG et al. . Identification of syntrophic acetate-oxidizing bacteria in anaerobic digesters by combined protein-based stable isotope probing and metagenomics . ISME J . 2016 ; 10 : 2405 . Google Scholar Crossref Search ADS PubMed WorldCat Muyzer G , de Waal EC , Uitterlinden AG . Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA . Appl Environ Microb . 1993 ; 59 : 695 – 700 . WorldCat Neufeld JD , Vohra J , Dumont MG et al. . DNA stable-isotope probing . Nat Protoc . 2007 ; 2 : 860 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Oksanen J , Blanchet FG , Friendly M et al. . Vegan: Community Ecology Package . 2017 . Google Preview WorldCat COPAC Oyehan TA , Al-Thukair AA. Isolation and characterization of PAH-degrading bacteria from the Eastern Province, Saudi Arabia . Mar Pollut Bull . 2017 ; 115 : 39 – 46 . Google Scholar Crossref Search ADS PubMed WorldCat Poonthrigpun S , Pattaragulwanit K , Paengthai S et al. . Novel intermediates of acenaphthylene degradation by Rhizobium sp. strain CU-A1: evidence for naphthalene-1,8-dicarboxylic acid metabolism . Appl Environ Microb . 2006 ; 72 : 6034 – 9 . Google Scholar Crossref Search ADS WorldCat Sandrin TR , Maier RM . Impact of metals on the biodegradation of organic pollutants . Environ Health Perspect . 2003 ; 111 : 1093 – 101 . Google Scholar Crossref Search ADS PubMed WorldCat Sawulski P , Clipson N , Doyle E . Effects of polycyclic aromatic hydrocarbons on microbial community structure and PAH ring hydroxylating dioxygenase gene abundance in soil . Biodegradation . 2014 ; 25 : 835 – 47 . Google Scholar Crossref Search ADS PubMed WorldCat Schloss PD , Westcott SL , Ryabin T et al. . Introducing Mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities . Appl Environ Microb . 2009 ; 75 : 7537 – 41 . Google Scholar Crossref Search ADS WorldCat Smith VH , Graham DW , Cleland DD . Application of resource-ratio theory to hydrocarbon biodegradation . Environ Sci Technol . 1998 ; 32 : 3386 – 95 . Google Scholar Crossref Search ADS WorldCat Song M , Jiang L , Zhang D et al. . Bacteria capable of degrading anthracene, phenanthrene, and fluoranthene as revealed by DNA based stable-isotope probing in a forest soil . J Hazard Mater . 2016 ; 308 : 50 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Storey S , Ashaari MM , Clipson N et al. . Opportunistic bacteria dominate the soil microbiome response to phenanthrene in a microcosm-based study . Front Microbiol . 2018 ; 9 : 2815 . Google Scholar Crossref Search ADS PubMed WorldCat Sutherland JB . Detoxification of polycyclic aromatic hydrocarbons by fungi . J Ind Microbiol . 1992 ; 9 : 53 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat Teng Y , Shen Y , Luo Y et al. . Influence of Rhizobium melilotion phytoremediation of polycyclic aromatic hydrocarbons by alfalfa in an aged contaminated soil . J Hazard Mater . 2011 ; 186 : 1271 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Thion C , Cebron A , Beguiristain T et al. . Long-term in situ dynamics of the fungal communities in a multi-contaminated soil are mainly driven by plants . FEMS Microbiol Ecol . 2012a ; 82 : 169 – 81 . Google Scholar Crossref Search ADS WorldCat Thion C , Cébron A , Beguiristain T et al. . PAH biotransformation and sorption by Fusarium solani and Arthrobacter oxydans isolated from a polluted soil in axenic cultures and mixed co-cultures . Int Biodeter Biodegr . 2012b ; 68 : 28 – 35 . Google Scholar Crossref Search ADS WorldCat Thomas F , Corre E , Cébron A . Stable isotope probing and metagenomics highlight the effect of plants on uncultured phenanthrene-degrading bacterial consortium in polluted soil . ISME J . 2019 ; 13 : 1814 – 30 . Google Scholar Crossref Search ADS PubMed WorldCat Wald J , Hroudova M , Jansa J et al. . Pseudomonads rule degradation of polyaromatic hydrocarbons in aerated sediment . Front Microbiol . 2015 ; 6 : 1268 . Google Scholar Crossref Search ADS PubMed WorldCat Wang H , Lou J , Gu H et al. . Efficient biodegradation of phenanthrene by a novel strain Massilia sp. WF1 isolated from a PAH-contaminated soil . Environ Sci Pollut Res . 2016 ; 23 : 13378 – 88 . Google Scholar Crossref Search ADS WorldCat Wardle DA. A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil . Biol Rev . 1992 ; 67 : 321 – 58 . Google Scholar Crossref Search ADS WorldCat Wick LY , Ruiz de Munain A , Springael D et al. . Responses of Mycobacterium sp. LB501T to the low bioavailability of solid anthracene . Appl Microbiol Biotechnol . 2002 ; 58 : 378 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat Wilhelm RC , Singh R , Eltis LD et al. . Bacterial contributions to delignification and lignocellulose degradation in forest soils with metagenomic and quantitative stable isotope probing . ISME J . 2019 ; 13 : 413 . Google Scholar Crossref Search ADS PubMed WorldCat Zhang W , Chen L , Zhang R et al. . High-throughput sequencing analysis of the joint effects of BDE209-Pb on soil bacterial community structure . J Hazard Mater . 2016 ; 301 : 1 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Zhang W , Wang H , Zhang R et al. . Bacterial communities in PAH-contaminated soils at an electronic-waste processing center in China . Ecotoxicology . 2010 ; 19 : 96 – 104 . Google Scholar Crossref Search ADS PubMed WorldCat Zhou Z-F , Wang M-X , Zuo X-H et al. . Comparative investigation of bacterial, fungal, and archaeal community structures in soils in a typical oilfield in Jianghan, China . Arch Environ Con Tox . 2017 ; 72 : 65 – 77 . Google Scholar Crossref Search ADS WorldCat © FEMS 2019. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) TI - DNA stable isotope probing reveals contrasted activity and phenanthrene-degrading bacteria identity in a gradient of anthropized soils JF - FEMS Microbiology Ecology DO - 10.1093/femsec/fiz181 DA - 2019-12-01 UR - https://www.deepdyve.com/lp/oxford-university-press/dna-stable-isotope-probing-reveals-contrasted-activity-and-hsFQsViVHo VL - 95 IS - 12 DP - DeepDyve ER -