Illumina-Based Analysis of Bulk and Rhizosphere Soil Bacterial Communities in Paddy Fields Under Mixed Heavy Metal Contamination

Pedosphere 27(3): 569–578, 2017 doi:10.1016/S1002-0160(17)60352-7 ISSN 1002-0160/CN 32-1315/P c 2017 Soil Science Society of China ⃝ Published by Elsevier B.V. and Science Press

Illumina-Based Analysis of Bulk and Rhizosphere Soil Bacterial Communities in Paddy Fields Under Mixed Heavy Metal Contamination HE Huaidong1 , LI Waichin2,∗ , Riqing YU3 and YE Zhihong1 1 State

Key Laboratory of Biocontrol and Guangdong Key Laboratory of Plant Resources, School of Life Sciences, Sun Yat-sen University, Guangzhou 510006 (China) 2 Department of Science and Environmental Studies, The Education University of Hong Kong, Hong Kong SAR (China) 3 Department of Biology, The University of Texas at Tyler, Tyler TX 75799 (USA) (Received February 16, 2017; revised March 27, 2017)

ABSTRACT There is an increasing concern about rice (Oryza sativa L.) soil microbiomes under the influence of mixed heavy metal contamination. We used the high-throughput Illumina MiSeq sequencing approach to explore the bacterial diversity and community composition of soils in four paddy fields, exhibiting four degrees of mixed heavy metal (Cd, Pb and Zn) pollution, and examined the effects of these metals on the bacterial communities. Our results showed that up to 2 104 to 4 359 bacterial operational taxonomic units (OTUs) were found in the bulk and rhizosphere soils of the paddy fields, with the dominant bacterial phyla (greater than 1% of the overall community) including Proteobacteria, Actinobacteria, Firmicutes, Acidobacteria, Gemmatimonadetes, Chloroflexi, Bacteroidetes and Nitrospirae. A number of rare and candidate bacterial groups were also detected, and Saprospirales, HOC36, SC-I-84 and Anaerospora were rarely detected in rice paddy soils. Venn diagram analysis showed that 174 bacterial OTUs were shared among the bulk soils with four pollution degrees. Rice rhizosphere soils displayed higher bacterial diversity indices (ACE and Chao 1) and more unique OTUs than bulk soils. Total Cd and Zn in the soils were significantly negatively correlated with ACE and Chao 1, respectively, and the Mantel test suggested that total Pb, total Zn, pH, total nitrogen and total phosphorus significantly affected the community structure. Overall, these results provided baseline data for the bacterial communities in bulk and rhizosphere soils of paddy fields contaminated with mixed heavy metals. Key Words:

bacterial diversity, community structure, Illumina MiSeq sequencing approach, long-term contamination, paddy soil

Citation: He H D, Li W C, Yu R, Ye Z H. 2017. Illumina-based analysis of bulk and rhizosphere soil bacterial communities in paddy fields under mixed heavy metal contamination. Pedosphere. 27(3): 569–578.

INTRODUCTION Heavy metal contamination in agricultural soils due to anthropogenic mining activities can result in adverse environmental effects, including soil quality degeneration, inhibition of crop growth and potential risks to human health by food chain transfer (Li Z et al., 2014). Paddy fields, a unique agro-ecosystem, are of vital importance for cereal production in Asia, specifically in China (K¨ogel-Knabner et al., 2010). Thus, heavy metal contamination of paddy soils in the vicinity of mining areas in China has become a growing concern in recent years (Li et al., 2012; Liu et al., 2012; Chen et al., 2014). Paddy fields consist of diverse habitats for microorganisms, which play significant roles in the maintenance of soil quality and rice (Oryza sativa L.) fitness (Liesack et al., 2000). In particular, as the most abun∗ Corresponding

author. E-mail: [email protected].

dant group of soil microorganisms, bacteria are actively involved in various biogeochemical processes of bulk and rhizosphere soils (Bu´ee et al., 2009). Compared to bulk soil bacteria, rhizosphere bacteria can be more directly beneficial to the rooting patterns and supply of nutrient elements to plants (Bu´ee et al., 2009). Given the confirmed importance of soil microbes, it is crucial to understand whether heavy metal contamination can affect soil microbial diversity and community structure. Khan et al. (2010) and Pan and Yu (2011) reported that the addition of Cd or Pb had a significant effect on microbial community structure in greenhouse experiments. However, short-term studies under greenhouse or laboratory conditions using soils spiked with heavy metals cannot be used to infer the long-term effects of heavy metals on soil microorganisms under field conditions (Giller et al., 2009). Tang et al. (2014) found no apparent correlations between

H. D. HE et al.

570

heavy metals and the microbial community in contaminated soils. Moreover, among the environmental factors, soil pH (Chodak et al., 2013) and organic matter concentrations (Hirsch et al., 2010) are considered more important factors than the presence of toxic metals with regard to the effects on soil bacterial communities. Thus, the current findings are inconsistent with reference to heavy metal effects on soil bacterial communities, and such studies have often focused on a single toxic metal. To date, few studies have explored the effects of long-term multi-metal contamination on paddy soil bacterial diversity and community structure under field conditions. In addition, because the diversity and composition of rice rhizosphere bacterial communities are highly complex and likely region-specific, the differences in bacterial communities between rice bulk and rhizosphere soils with different pollution degrees have not been well studied. Culture-independent molecular approaches based on extraction, amplification and sequencing of 16S rRNA genes provide feasibility and a new perspective to characterize soil microbial communities (Hirsch et al., 2010). Recent high-throughput sequencing (e.g., Illumina) has been described as a promising and rapid approach to analyse soil microbial diversity and community structure (Rincon-Florez et al., 2013; Myrold et al., 2014). Thus, the bacterial communities in soils were analysed using the Illumina MiSeq sequencing approach in the present work. Soil samples were collected from four selected paddy fields in the vicinity of several traditional mines in southern China. The aims of this study were: 1) to characterize the bacterial diversity and community composition of bulk and rhizosphere soils in paddy fields with different degrees of mixed heavy metal (Cd, Pb and Zn) contamination and 2) to determine the role of environmental factors, specifically heavy metals, on the bacterial diversity and community structure in paddy soils.

MATERIALS AND METHODS Study sites and soil sampling In southern China, many paddy fields have been subjected to the contamination of mixed heavy metals due to mining activities (Liu et al., 2012). Our study was performed at different paddy fields around traditional mining areas in Guangdong Province, southern China (Table I). The selected paddy field sites included an unpolluted site, LT, from Liantang Village, Qingyuan City and three sites with long-term heavy metal pollution, SB, from Shangba Village around Dabaoshan Mine, Shaoguan City, TX, from Tongxi Village around Tongxi Mine, Qingyuan City, and FK, from Fankou Town around Fankou Mine, Shaoguan City. Bulk soils, LB, SB, TB and FB, from the four sites, LT, SB, TX and FK, respectively, were sampled from the top 10 cm of the profile, and rhizosphere soils, LR and FR, from the two sites, LT and FK, respectively, were separated from the roots of the rice (Oryza sativa L.) cultivar Tianyou 122 (pure line seeds from the Rice Research Institute of Guangdong Academy of Agricultural Sciences, China) (Table I). The soil that remained adhered to the root hairs after gentle shaking was sampled as rhizosphere soils according to an operational definition (Lynch, 1990; Chen et al., 2006). For each plot (bulk or rhizosphere soil), multiple random soil cores were collected and immediately mixed thoroughly in June 2013 immediately after drainage. The soil samples were then transported to the laboratory on dry ice. A fraction of the soil sample was immediately stored at −20 ◦ C for molecular analyses. Another fraction was air-dried for soil chemical characterization. Chemical analysis Soil pH was measured using a portable pH meter (pH 510, Eutech Instruments, Singapore) in 1:2.5 (weight:volume) soil-deionized water ratio, and total

TABLE I Description of soil samples from four paddy fields in Guangdong Province, southern China Site

Location

Pollution source and history

Sample(s)

LT

Liantang Village, Qingyuan City (23◦ 52′ 27′′ N, 113◦ 35′ 59′′ E) Shangba Village, Shaoguan City (24◦ 27′ 83′′ N, 113◦ 48′ 16′′ E) Tongxi Village, Qingyuan City (23◦ 51′ 49′′ N, 113◦ 39′ 54′′ E) Fankou Town, Shaoguan City (25◦ 07′ 06′′ N, 113◦ 39′ 12′′ E)

No direct heavy metal pollution

Bulk soil (LB), rhizosphere soil (LR) Bulk soil (SB)

SB TX FK

Waste water irrigation from a lead and zinc mine since the 1960s Waste water irrigation from a limonite mine since the 1960s Waste water irrigation from a lead and zinc mine since the 1950s

Bulk soil (TB) Bulk soil (FB), rhizosphere soil (FR)

BACTERIAL COMMUNITY UNDER HEAVY METAL CONTAMINATION

organic carbon (TOC) concentration was determined with a TOC analyser (TOC-VCPH, Japan). Total nitrogen (TN) and total phosphorus (TP) were analysed with a Smartchem discrete auto analyser (Smartchem 200, Westco Scientific Instruments, Italy). For metal analyses, soil samples were first digested with an HCl/HNO3 mixture (3:1, weight/volume) at 180 ◦ C for 40 min in a microwave oven (MARS-X, CEM Corporation, USA). Total Cd concentrations in the digests were measured using flame atomic absorption spectrometry (Hitachi Z-2300, Hitachi, Japan). The concentrations of total Zn and total Pb in the digests were analysed using inductively coupled plasma optical emission spectrometry (ICP-OES, Optima 2000 DV, Perkin Elmer, USA). The Nemerow composite index, a widely used pollution indicator, was then calculated based on total heavy metal concentrations to define the different levels of heavy metal contamination according to Hong et al. (2015). Blanks and standard soil material (GBW-07435) from China Standard Materials Research Center, Beijing, China, were used for quality control. The elemental recovery rates for the standard material ranged from 90% to 100% in all cases (Li et al., 2012). DNA extraction, PCR amplification and Illumina MiSeq sequencing Total genomic DNA was extracted from each homogenized soil sample using a FastDNAr Spin Kit for Soil (MP Biomedicals, USA) according to the manufacturer’s instructions. The concentration of the extracted DNA was confirmed using a NanoDrop device (ND-2000, Thermo Fisher, USA), and its integrity and size were confirmed using 1.0% agarose gel electrophoresis. The variable V3–V4 regions of bacterial 16S rRNA genes were utilized as targets for the analyses of Illumina 16S rDNA sequencing. The 16S rRNA gene universal bacterial primers 343F (5′ TACGGRAGGCAGCAG) and 798R (5′ -AGGGTATCTAATCCT) modified with Illumina transposase sequences were used for PCR amplification. The PCR mixtures (25 µL) contained 12.5 µL of 2 × KAPA HiFi Hot Start Readymix (KAPA Biosystems, USA), 5 µL of each primer (1 µmol L−1 ) and 2.5 µL of target DNA (5 ng µL−1 ). The PCR cycling conditions consisted of an initial denaturation step at 95 ◦ C (3 min), followed by 25 cycles of 95 ◦ C (30 s), 60 ◦ C (30 s) and 72 ◦ C (30 s) and a final elongation at 72 ◦ C (10 min). The PCR products were purified using AMPure XP beads (Agencourt Bioscience, USA). A secondary PCR amplification was performed for Illumina MiSeq sequencing using bar-coded primers for the V3–V4 regions of 16S

571

rRNA genes with Illumina’s linker P5/P7 sequences and 8-nt unique index sequences. The secondary PCR mixtures (50 µL) contained 25 µL of 2 × KAPA HiFi Hot Start Readymix, 5 µL of each primer (10 µmol L−1 ) and 5 µL of DNA and PCR grade water (10 µmol L−1 ), and the number of cycles was reduced to 10. Subsequently, the secondary PCR products from different samples were purified using AMPure XP beads. The pure amplicons were quantified with a Quant-iT dsDNA HS Assay Kit and Qubit 2.0 fluorometer (Invitrogen Life Technologies, USA) and pooled in equimolar amounts. Finally, the mixture of PCR products was subjected to 2 × 300-nt paired-end sequencing using a MiSeq Reagent Kit v3 on the Illumina MiSeq Sequencing platform at the National Human Genome Centre of China, Shanghai, China. Bioinformatics and statistical analysis The diversity and composition of bacterial communities of the paddy soil samples were analysed based on the raw sequencing data obtained from the Illumina MiSeq platform. Raw sequences through pairedend sequencing were merged based on the method described by Magoˇc and Salzberg (2011). Sequencing reads were sorted into each sample according to the corresponding unique barcode. Sequences were processed using the QIIME software package (Caporaso et al., 2010b). Operational taxonomic units (OTUs) at 97% similarity were selected using UPARSE (Edgar, 2013). PyNAST alignment (Caporaso et al., 2010a) and ribosomal database project (RDP) assignment (Wang et al., 2007) were performed based on the Greengenes database. The rarefaction analysis based on the method of Mothur (Schloss et al., 2009) was performed to reveal the richness (diversity) indices (ACE and Chao 1), and rank-abundance curves were produced. Venn diagram analysis was performed using the Mothur suite of programs (Schloss et al., 2009). To examine the relationships between soil environmental factors and bacterial community structure, the Mantel test was implemented in the R statistical environment (Bonnet and Van de Peer, 2002). The sequences obtained in this study were deposited at MG-RAST under the accession numbers 4628895.3–4628900.3. Other statistical analyses were performed using SPSS 18.0, and tables and figures were generated using Excel 2010 and Origin 8.5 software packages. RESULTS Environmental parameters The physiochemical properties of the soil samples

H. D. HE et al.

572

from the four paddy fields, including an unpolluted site (LT) and three heavy metal-polluted sites (SB, TX and FK), are shown in Table II. Soil pH, TOC, TN and TP differed to some extent among the four sites. The concentrations of total Cd, Zn and Pb in the soils at the sites SB, TX and FX were clearly higher than those of the unpolluted site LT and exceeded the environmental quality evaluation standards for farmland of edible agricultural products in China: for pH < 6.5, Cd = 0.3 mg kg−1 , Pb = 80 mg kg−1 and Zn = 200 mg kg−1 ; for pH 6.5–7.5, Cd = 0.3 mg kg−1 , Pb = 80 mg kg−1 and Zn = 250 mg kg−1 (Ministry of Environmental Protection of the People’s Republic of China, 2006). The Nemerow composite index showed that the soil samples (LB and LR) obtained from the site LT was unpolluted, while those from the other three sites (SB, TX and FK) were contaminated with heavy metals. The soil samples (FB and FR) from the site FK had the highest index values, indicating that this site was the most heavily polluted. The general metal contamination trend was: LT < SB < TX < FK. The rhizosphere soils exhibited lower pH values and total Cd, Pb and Zn concentrations, but higher contents of TOC and TN, compared to the bulk soils at both the sites LT and FK to some extent. Sequencing reads and bacterial diversity A total of 75 578 high-quality reads were obtained from all paddy soil samples using Illumina MiSeq sequencing analysis. Highly diverse bacterial communities with up to 2 104 to 4 359 OTUs were revealed based on high-quality sequencing reads at a 3% sequence

dissimilarity level in the soil samples (Table I). Rarefaction curve analysis, based on bacterial OTUs at a sequence dissimilarity level of 3%, indicated that the soil bacterial libraries occurred in all the detected soil samples, representing the bacterial communities after deep sequencing with an average of 12 596 sequence reads per sample (Fig. 1a). Rank-abundance curves indicated that a majority of the OTUs belonged to relatively low-abundance groups of bacteria, and all soil samples contained relatively low proportions of highly abundant bacteria (Fig. 1b). The bacterial OTU numbers and diversity indices indicated the same trends: LR > LB > SB > TB > FR > FB (Table II). In general, they were gradually reduced in soils with increasing levels of heavy metal contamination. The rhizosphere soils exhibited higher bacterial diversity indices than bulk soils at both unpolluted (LT) and polluted (FK) sites (Table II). To measure the overlap between the bacterial communities, the Venn diagram analysis based on bacterial OTUs is shown in Fig. 2. The analysis revealed the shared and unique bacterial OTUs among all bulk soil samples (Fig. 2a). The number of OTUs shared was 174 for the intersection among all bulk soils, 1 164 for that between LB and SB, 899 for that between LB and TB and 552 for that between LB and FB. The trend of unique OTU numbers was shown as: LB (1 650) > SB (1 355) > TB (1 185) > FB (1 173). The rhizosphere soils displayed more shared and unique OTUs than the bulk soils (Fig. 2b). The number of OTUs shared was 2 025 for the intersection between LB and LR, 1 154 for that between FB and FR, 721 for that between LR and

TABLE II Bacterial diversity indices defined at a 3% sequence dissimilarity level and selected propertiesa) of the soils sampled from four paddy fields in Guangdong Province, southern China and their correlations with the overall bacterial community structure by the Mantel test (n = 4) Soilb)

Total Cd

LB LR SB TB FB FR

0.1 0.1 2.5 5.5 13.4 12.6

rd)

NSe)

Total Zn

Total Pb

mg kg−1 DWc) 103 55 96 53 329 252 729 109 4 787 3 459 4 564 3 329 0.58*

0.62**

NCI

0.6 0.6 6.8 13.9 40.0 38.2

pH

TOC

TN

TP

5.15 5.11 4.74 4.67 7.24 7.00

20.4 22.5 14.3 34.5 43.1 47.4

g kg−1 DW 1.52 1.67 1.14 1.63 2.45 2.50

0.50 0.51 0.54 0.65 1.40 1.43

0.54*

NS

0.55*

0.57*

OTU number

3 314 4 359 3 177 2 588 2 104 2 375

Bacterial diversity index ACE

Chao 1

8 273 12 695 7 885 7 468 5 782 6 494

5 986 8 885 5 728 5 307 4 228 4 662

*, **Significant at P < 0.05 and P < 0.01, respectively. = Nemerow composite index; TOC = total organic C; TN = total N; TP = total P; OTU = bacterial operational taxonomic unit. b) See detailed description of the soil samples in Table I. c) DW = dry weight. d) Pearson’s correlation coefficient using Mantel test. e) Not significant (P > 0.05). a) NCI

BACTERIAL COMMUNITY UNDER HEAVY METAL CONTAMINATION

573

Fig. 1 Rarefaction (a) and rank-abundance (b) curves of bacterial communities clustered at a dissimilarity level of 3% in the soil samples from four paddy fields in Guangdong Province, southern China. See detailed description of the soil samples LB, LR, SB, TB, FB and FR in Table I.

FR and 552 for that between LB and FB. The trend of unique OTU numbers was ordered as LR (1 991) > LB (1 077) > FR (897) > FB (717). Taxonomic assignment of sequencing reads and bacterial community composition

Fig. 2 Venn diagrams showing bacterial operational taxonomic unit distribution at a 3% sequence dissimilarity level among four bulk soils (LB, TB, FB and SB) (a) and two bulk (LB and FB) and the two corresponding rhizosphere soils (LR and FR) (b) from four paddy fields in Guangdong Province, southern China. See detailed description of the soil samples in Table I.

The sequencing reads recovered from all bulk and rhizosphere soils classified at the phylum level were affiliated with 15 bacterial phyla (Fig. 3). The dominant phyla accounting for more than 1% of the overall community in each bulk and rhizosphere soil were Proteobacteria (25.5%–38.9%), Actinobacteria (22.9%–38.5%), Firmicutes (12.0%–19.4%), Acidobacteria (4.5%–10.7%), Gemmatimonadetes (2.3%–6.5%), Chloroflexi (2.1%–4.8%), Bacteroidetes (1.2%–3.4%) and Nitrospirae (1.3%–1.8%). Bacterial phyla with less abundance in all soil samples included Chlorobi, Verrucomicrobia, Spirochaetes, Elusimicrobia, Cyanobacteria and the candidate phyla OT1 and TM7. Nevertheless, the relative abundances of dominant phyla differed among different soil samples (Fig. 3). For example, compared to other soils, bulk (LB) and rhizosphere (LR) soils at the site LT revealed a higher rela-

tive abundance of Actinobacteria, whereas the bulk (FB) and rhizosphere (FR) soils at the site FK both exhibited higher relative abundances of Proteobacteria, Firmicutes, Gemmatimonadetes and Bacteroidetes. Rhizosphere soils (LR and FR) displayed higher relative abundances of Chloroflexi, Chlorobi and Spirochaetes than the corresponding bulk soils (LB and LR), respectively. At the order level, a total of 25 orders were assigned in the soil samples (Table III). The dominant orders accounting for more than 1% of the overall community in each bulk and rhizosphere soil were Actinomycetales, Bacillales, Clostridiales, Gaiellales, Rhizobiales, Myxococcales, Solirubrobacterales, Acidimicrobiales, Solibacterales, Syntrophobacterales, Rhodospirillales and Nitrospirales. Other orders were present at low abundance. At the genus level, the domi-

H. D. HE et al.

574

Fig. 3 Taxonomic classification of bacterial reads at the phylum level of the bulk (LB, TB, FB and SB) and rhizosphere (LR and FR) soils from four paddy fields in Guangdong Province, southern China. See detailed description of the soil samples in Table I.

nant genera accounting for more than 1% of the overall community were Bacillus, Clostridium, Rhodoplanes, Thiobacillus, Anaeromyxobacter and Candidatus Solibacter in at least three paddy fields (Table IV). Other genera, such as Mycobacterium, Kaistobacter, Geobacter, Streptomyces, Tepidibacter, Phycicoccus, Nitrospira, Bradyrhizobium, Terracoccus, Anaerospora and Desulfosporosinus, were detected in all soils and were dominant in several samples. The orders Saprospirales, HOC36 and SC-I-84 (Table III) and the genus Anaerospora (Table IV) were rarely detected in paddy soils. Relationships between bacterial communities and environmental parameters Non-linear regression analyses of bacterial diversity indices with environmental parameters for all paddy soil samples revealed that total concentrations of the heavy metals Cd and Zn were significantly (P < 0.05) negatively correlated with the diversity indices (ACE and Chao 1) (Fig. 4). However, the bacterial diversity indices were not significantly correlated with pH, TOC, TN, TP and total Pb in soils (data not shown). The Pearson’s correlation coefficients (r) between soil environmental parameters and bacterial community structure using Mantel test are shown in Table II. The overall bacterial community structure was signifi-

cantly correlated with total Pb (r = 0.62, P < 0.01), total Zn (r = 0.58, P < 0.05), pH (r = 0.54, P < 0.05), TN (r = 0.55, P < 0.05) and TP (r = 0.57, P < 0.05) in soils, respectively, but not significantly related to soil TOC and total Cd. Soil total Pb appeared to be the most important environmental parameter linked to the bacterial community structure from the analyses. DISCUSSION Paddy soil microorganisms are of vital importance for soil quality, root patterns and rice growth (Liesack et al., 2000). Due to the rapid response, analysis of soil microbial biodiversity can be used to monitor the potential detrimental effects of anthropogenic pollution on microbial communities (Gol¸ebiewski et al., 2014). Compared to previous studies using traditional methods to explore soil microbes, high-throughput Illumina sequencing offers more DNA sequencing reads and can retrieve large numbers of OTUs (Degnan and Ochman, 2012). In the present study, highly diverse bacterial communities with up to 2 104 to 4 359 OTUs were detected in the paddy soils (Table II). Soil bacterial diversity was evaluated in the samples by accurately estimating OTUs at a 3% sequence dissimilarity level and diversity indices (ACE and Chao 1) as suggested by Acosta-Mart´ınez et al. (2008). In this study, the diversity indices (ACE and Chao 1) of paddy soil ba-

BACTERIAL COMMUNITY UNDER HEAVY METAL CONTAMINATION

575

TABLE III

TABLE IV

Taxonomic classification of bacterial reads at the order level of the bulk (LB, TB, FB and SB) and rhizosphere (LR and FR) soilsa) from four paddy fields in Guangdong Province, southern China

Taxonomic classification of bacterial reads at the genus level of the bulk (LB, TB, FB and SB) and rhizosphere (LR and FR) soilsa) from four paddy fields in Guangdong Province, southern China

Bacterial order

Bacterial genus

Relative abundance LB

LR

Actinomycetales 25.19 21.65 Bacillales 7.53 7.01 Clostridiales 5.68 6.11 Gaiellales 6.18 7.40 Rhizobiales 5.04 5.19 Myxococcales 5.15 4.50 Solirubrobacterales 3.43 3.37 Acidimicrobiales 2.51 2.51 Solibacterales 2.67 3.11 Syntrophobacterales 1.52 1.41 SJA-15 2.62 3.32 Rhodospirillales 1.69 2.12 Hydrogenophilales 0.09 0.15 Acidobacteriales 2.03 2.12 Nitrospirales 1.69 1.47 Sphingomonadales 2.00 2.37 Saprospirales 0.66 0.70 Xanthomonadales 0.87 1.02 Burkholderiales 1.61 1.12 Desulfuromonadales 0.85 0.94 MND1 0.33 0.27 Bacteroidales 0.82 0.79 Rhodobacterales 1.67 0.05 HOC36 1.58 1.94 SC-I-84 0.98 0.74 a) See

SB

TB

FB

FR

% 12.78 16.21 9.45 12.29 2.84 6.06 10.96 12.64 8.99 8.86 8.24 4.19 4.08 5.93 5.07 4.02 4.72 3.41 4.82 4.22 5.27 1.62 3.99 4.19 1.89 7.58 2.32 1.92 2.99 3.34 3.87 3.70 4.13 2.52 2.70 1.62 2.38 1.65 2.87 2.46 2.59 2.15 0.66 0.86 1.65 2.43 1.90 1.70 3.01 3.25 1.87 2.84 3.18 2.42 0.12 0.11 1.35 1.42 1.61 1.80 0.62 0.10 1.53 1.83 0.40 0.18 2.37 1.94 0.99 0.94 1.19 1.08 0.71 0.32 0.81 1.49 2.32 0.28 0.54 0.95 0.78 1.50 1.69 1.18 1.48 0.72 0.36 0.95 0.15 1.02 0.64 0.66 0.00 0.11 0.01 0.01 0.50 0.45 0.20 0.27

detailed description of the soil samples in Table I.

cteria were significantly negatively correlated with total Cd and total Zn (Fig. 4), indicating that mixed heavy metal contamination could play a more important role in shaping bacterial diversity. Chodak et al. (2013) found that the bacterial diversity index Chao 1 was significantly correlated with heavy metals in forest soils. Heavy metal pollution most likely affects microbial diversity by inhibiting the metalsensitive species, which lack sufficient tolerance to the imposed stress of heavy metals, whilst stimulating metal-resistant species (Vig et al., 2003). The abundance of potential metal-resistant species might increase to maintain ecological stability in long-term metal-contaminated soil (Liu et al., 2014). In addition, some bacterial OTUs were shared among the bulk soils with four pollution degrees (Fig. 2a), indicating that they may potentially delineate the core microbiome of the sample set. Thus, the observation may be related to bacterial adaption in soils with different levels of mixed heavy metal contamination. The present results indicated that rhizosphere soils showed higher bacterial diversity and more unique

Relative abundance LB

Bacillus Clostridium Rhodoplanes Thiobacillus Mycobacterium Anaeromyxobacter Candidatus Solibacter Kaistobacter Geobacter Streptomyces Tepidibacter Phycicoccus Nitrospira Bradyrhizobium Terracoccus Anaerospora Desulfosporosinus a) See

5.29 2.94 2.54 0.08 3.01 1.36 1.51 1.78 0.82 1.87 0.05 0.55 0.36 0.60 1.34 1.66 0.14

LR 4.92 2.93 2.75 0.15 3.00 1.32 1.69 1.99 0.86 1.79 0.15 0.39 0.28 0.46 1.04 0.00 0.11

SB

TB

FB

FR

1.75 4.86 1.86 2.99 0.56 1.75 1.72 0.27 2.27 0.24 0.16 0.15 0.57 1.22 0.27 0.02 1.06

% 4.32 4.83 1.09 2.79 3.31 0.18 1.00 0.04 0.28 0.11 0.06 0.08 0.28 0.31 0.33 0.95 0.59

8.01 2.26 1.72 1.86 0.40 0.96 0.33 0.83 0.52 0.62 3.31 1.07 1.29 0.33 0.02 0.00 0.13

11.39 1.17 1.39 2.84 0.32 1.13 0.23 0.90 0.95 0.50 0.95 1.59 1.04 0.39 0.00 0.00 0.02

detailed description of the soil samples in Table I.

OTUs in contrast to bulk soils at both unpolluted and polluted paddy fields (Table II, Fig. 2b). Shentu et al. (2014) found that microorganisms were more abundant in the rhizosphere than bulk soil in a Cd-polluted zone. Nevertheless, rhizobacteria of plants were affected by heavy metals, but generally, the effects could be associated with an indirect stimulation of the roots (Bennisse et al., 2004). The higher diversity of bacteria in rhizosphere soils could be due to the root activities (e.g., high level of organic exudates), which provide suitable ecological niches for bacterial growth and lead to an increased number of microbes (Bu´ee et al., 2009). In our study, rhizosphere soils showed higher organic matter and lower concentrations of heavy metals compared to bulk soils to some extent (Table II). A positive effect on bacterial diversity might also be expected due to the decreased concentrations of heavy metals in the rhizosphere (Gremion et al., 2003). In addition to the bacterial diversity, the community composition of bacteria in soils should be a concern under long-term heavy metal stress (Pan and Yu, 2011). In the present study, the dominant bacterial phyla (greater than 1% of the overall community) were Proteobacteria, Actinobacteria, Firmicutes, Acidobacteria, Gemmatimonadetes, Chloroflexi, Bacteroidetes and Nitrospirae in rice bulk and rhizosphere soils (Fig. 3). These phyla have been described as common bacterial groups in other soils, although their re-

576

H. D. HE et al.

Fig. 4 Correlations between bacterial diversity indices (ACE and Chao 1) and the total concentrations of heavy metals (Cd and Zn) in the bulk (LB, TB, FB and SB) and rhizosphere (LR and FR) soils from four paddy fields in Guangdong Province, southern China. See detailed description of the soil samples in Table I.

lative abundances vary with soils (Chodak et al., 2013). Proteobacteria and Chloroflexi are detected as the most abundant bacterial phyla in paddy soils (Liu et al., 2014). However, the top 2 dominant phyla of bacteria in this study were Proteobacteria and Actinobacteria (Fig. 3). These two phyla comprise the active bacterial fraction in the heavy metal-polluted soils (Margesin et al., 2011). Our results also indicated that the bulk and rhizosphere soils displayed higher relative abundances of Proteobacteria at the heavily polluted site FK, and a higher diversity of Actinobacteria was observed at the unpolluted site LT (Fig. 3). Proteobacteria may be the most metal-tolerant microorganisms discovered in heavily contaminated soils, but it is unclear whether the loss of Actionbacteria is due to their sensitivity to heavy metals (Sheik et al., 2012; Gol¸ebiewski et al., 2014). Expectedly, r-selected organisms (rapidly reproducing), such as Proteobacteria, are favoured after an ecosystem is exposed to a stressor, which is a potential reason for their dominance (Sheik et al., 2012). In this study, high-throughput sequencing provided the opportunity to perform an indepth study on the bacterial community composition in paddy soils, and the microbial community analysis indicated the dominant genera (Table IV). These organisms could be related to dynamically biogeochemical cycling in paddy fields. For example, Clostridium,

Thiobacillus, Anaeromyxobacter, Geobacter and Desulfosporosinus play important roles in iron or sulfate cycling in paddy soils, and the dominance of Rhodoplanes may have a beneficial ecological function in improving soil fertility (Sun et al., 2015). Some bacterial groups, Saprospirales, HOC36, SC-I-84 and Anaerospora (Table IV), are rarely detected in paddy soils, indicating that the paddy field ecosystem may harbour functional microorganisms more than expected. The ecological role of these microorganisms in paddy fields requires further investigation. Analysis of the correlations between environmental variables and microbial community structure in soils will reveal the changes in microbial communities influenced by these parameters. Our results showed that the overall community structure of bacteria was significantly correlated with total Zn, pH, TN, TP and, in particular, total Pb, but was not significantly correlated with total Cd and TOC in paddy soils (Table II). De la Iglesia et al. (2006) reported that elevated concentrations of heavy metals in the tailing soil had a strong effect on the bacterial community composition, but other soil factors, such as soil pH, also played important roles in the bacterial community structure. Li H et al. (2014) also reported that the overall bacterial community composition was significantly correlated with soil pH, TN and TP, but was not signi-

BACTERIAL COMMUNITY UNDER HEAVY METAL CONTAMINATION

ficantly associated with TOC in forest soil. Chodak et al. (2013) suggested that all the dominant bacterial phyla studied were affected by the contamination of heavy metals in the same way. We also found that both the bulk and rhizosphere soils at the heavily polluted site FK exhibited higher relative abundances of the predominant phyla (Proteobacteria, Firmicutes, Gemmatimonadetes and Bacteroidetes) and lower relative abundances of less abundant phyla (Chloroflexi, Chlorobi, Verrucomicrobia, Spirochaetes, Elusimicrobia and Cyanobacteria) compared to other paddy soils in this study (Fig. 3). Thus, if the long-term presence of high heavy metal concentrations in soils affects microbial life, then metal-resistant microorganisms will become the dominant populations (De la Iglesia et al., 2006). Soil heavy metals and other factors such as pH, TN and TP could affect the structure of soil bacterial communities in paddy fields. The effects most likely represent complicated evolution processes integrating bacterial metal resistance, adaption to site-specific soil physicochemical features and rhizosphere alleviation or interference from plants. CONCLUSIONS We characterized the diversity and community composition of bacteria in rice bulk and rhizosphere soils by applying the high-throughput Illumina MiSeq sequencing approach and assessed the microbial effects of long-term contamination of mixed heavy metals in different paddy fields near mining areas. A number of dominant and rare bacterial phyla (groups) were detected. Our results indicated that total Cd and Zn in paddy soils have significantly negative effects on soil bacterial diversity, while total Pb, total Zn, pH, TN and TP had significant effects on the structure of bacterial communities. The rice rhizosphere habitats displayed greater bacterial diversity than bulk soils. Future investigations related to the functional analysis of the microbial communities in rice rhizospheres will likely reveal a more in-depth understanding of their environmental and ecological roles. ACKNOWLEDGEMENTS This research was supported by the Research Grants Council of the Hong Kong Special Administrative Region, China (No. 28100014), the Research and Development Office of the Education University of Hong Kong, China (No. RG84/2012-2013R) and the NSFC-Guangdong United Foundation, China (No. U1501232). We sincerely thank Prof. A J M Baker from the University of Melbourne, Australia, for the help in

577

the initial preparation and improvement of this paper. REFERENCES Acosta-Mart´ınez V, Dowd S, Sun Y, Allen V. 2008. Tag-encoded pyrosequencing analysis of bacterial diversity in a single soil type as affected by management and land use. Soil Biol Biochem. 40: 2762–2770. Bennisse R, Labat M, Elasli A, Brhada F, Chandad F, Lorquin J, Liegbott P P, Hibti M, Qatibi A I. 2004. Rhizosphere bacterial populations of metallophyte plants in heavy metalcontaminated soils from mining areas in semiarid climate. World J Microb Biot. 20: 759–766. Bonnet E, Van de Peer Y. 2002. zt: a software tool for simple and partial Mantel tests. J Stat Softw. 7: 1–12. Bu´ ee M, De Boer W, Martin F, Van Overbeek L, Jurkevitch E. 2009. The rhizosphere zoo: An overview of plant-associated communities of microorganisms, including phages, bacteria, archaea, and fungi, and of some of their structuring factors. Plant Soil. 321: 189–212. Caporaso J G, Bittinger K, Bushman F D, DeSantis T Z, Andersen G L, Knight R. 2010a. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics. 26: 266–267. Caporaso J G, Kuczynski J, Stombaugh J, Bittinger K, Bushman F D, Costello E K, Fierer N, Pena A G, Goodrich J K, Gordon J I. 2010b. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 7: 335–336. Chen J, He F, Zhang X, Sun X, Zheng J, Zheng J. 2014. Heavy metal pollution decreases microbial abundance, diversity and activity within particle-size fractions of a paddy soil. FEMS Microbiol Ecol. 87: 164–181. Chen Y M, Wang M K, Zhuang Y S, Chiang P N. 2006. Chemical and physical properties of rhizosphere and bulk soils of three tea plants cultivated in Ultisols. Geoderma. 136: 378–387. Chodak M, Gol¸ebiewski M, Morawska-Ploskonka J, Kuduk K, Nikli´ nska M. 2013. Diversity of microorganisms from forest soils differently polluted with heavy metals. Appl Soil Ecol. 64: 7–14. Degnan P H, Ochman H. 2012. Illumina-based analysis of microbial community diversity. ISME J. 6: 183–194. De la Iglesia R, Castro D, Ginocchio R, van der Lelie D, Gonz´ alez B. 2006. Factors influencing the composition of bacterial communities found at abandoned copper-tailings dumps. J Appl Microbiol. 100: 537–544. Edgar R C. 2013. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods. 10: 996–998. Giller K E, Witter E, McGrath S P. 2009. Heavy metals and soil microbes. Soil Biol Biochem. 41: 2031–2037. Gol¸ebiewski M, Deja-Sikora E, Cichosz M, Tretyn A, Wr´ obel B. 2014. 16S rDNA pyrosequencing analysis of bacterial community in heavy metals polluted soils. Microb Ecol. 67: 635–647. Gremion F, Chatzinotas A, Harms H. 2003. Comparative 16S rDNA and 16S rRNA sequence analysis indicates that Actinobacteria might be a dominant part of the metabolically active bacteria in heavy metal-contaminated bulk and rhizosphere soil. Environ Microbiol. 5: 896–907. Hirsch P R, Mauchline T H, Clark I M. 2010. Culture-independent molecular techniques for soil microbial ecology. Soil Biol Biochem. 42: 878–887. Hong C, Si Y X, Xing Y, Li Y. 2015. Illumina MiSeq sequencing investigation on the contrasting soil bacterial community structures in different iron mining areas. Environ Sci Pollut Res. 22: 10788–10799.

578

Khan S, Hesham A E L, Qiao M, Rehman S, He J Z. 2010. Effects of Cd and Pb on soil microbial community structure and activities. Environ Sci Pollut R. 17: 288–296. K¨ ogel-Knabner I, Amelung W, Cao Z, Fiedler S, Frenzel P, Jahn R, Kalbitz K, K¨ olbl A, Schloter M. 2010. Biogeochemistry of paddy soils. Geoderma. 157: 1–14. Li B, Wang X, Qi X, Huang L, Ye Z. 2012. Identification of rice cultivars with low brown rice mixed cadmium and lead concentrations and their interactions with the micronutrients iron, zinc, nickel and manganese. J Environ Sci. 24: 1790– 1798. Li H, Ye D, Wang X, Settles M, Wang J, Hao Z, Zhou L, Dong P, Jiang Y, Ma Z. 2014. Soil bacterial communities of different natural forest types in Northeast China. Plant Soil. 383: 203–216. Li Z, Ma Z, van der Kuijp T J, Yuan Z, Huang L. 2014. A review of soil heavy metal pollution from mines in China: pollution and health risk assessment. Sci Total Environ. 468-469: 843–853. Liesack W, Schnell S, Revsbech N P. 2000. Microbiology of flooded rice paddies. FEMS Microbiol Rev. 24: 625–645. Liu Y, Wang J, Zheng Y, Zhang L, He J. 2014. Patterns of bacterial diversity along a long-term mercury-contaminated gradient in the paddy soils. Microbial Ecol. 68: 575–583. Liu Y, Zhou T, Crowley D, Li L, Liu D, Zheng J, Yu, X, Pan G X, Hussain Q, Zhang X, Zheng J. 2012. Decline in topsoil microbial quotient, fungal abundance and C utilization efficiency of rice paddies under heavy metal pollution across South China. PLoS ONE. 7: e38858. Lynch J M. 1990. Soil Rhizosphere. John Wiley and Sons, New York. Magoˇ c T, Salzberg S L. 2011. FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics. 27: 2957–2963. Margesin R, Plaza G A, Kasenbacher S. 2011. Characterization of bacterial communities at heavy-metal-contaminated sites. Chemosphere. 82: 1583–1588. Ministry of Environmental Protection of the People’s Republic of China, 2006. Environmental Quality Evaluation Standards for Farmland of Edible Agricultural Products in China (in Chinese). China Environmental Science Press, Beijing.

H. D. HE et al.

Myrold D D, Zeglin L H, Jansson J K. 2014. The potential of metagenomic approaches for understanding soil microbial processes. Soil Sci Soc Am J. 78: 3–10. Pan J, Yu L. 2011. Effects of Cd or/and Pb on soil enzyme activities and microbial community structure. Ecol Eng. 37: 1889–1894. Rincon-Florez V A, Carvalhais L C, Schenk P M. 2013. Cultureindependent molecular tools for soil and rhizosphere microbiology. Diversity. 5: 581–612. Schloss P D, Westcot S L, Ryabin T, Hall J R, Hartmann M, Hollister E B, Lesniewski R A, Oakley B B, Parks D H, Robinson C J, Sahl J W, Stres B, Thallinger G G, Van Horn D J, Weber C F. 2009. Introducing mothur: open-source, platformindependent, community-supported software for describing and comparing microbial communities. Appl Environ Microb. 75: 7537–7541. Sheik C S, Mitchell T W, Rizvi F Z, Rehman Y, Faisal M, Hasnain S, McInerney M J, Krumholz L R. 2012. Exposure of soil microbial communities to chromium and arsenic alters their diversity and structure. PLoS ONE. 7: e40059. Shentu J L, He Z Y, Zeng Y Y, He S Y, Du S T, Shen D S. 2014. Microbial biomass and PLFA profile changes in rhizosphere of pakchoi (Brassica chinensis L.) as affected by external cadmium loading. Pedosphere. 24: 553–562. Sun M, Xiao T, Ning Z, Xiao E, Sun, W. 2015. Microbial community analysis in rice paddy soils irrigated by acid mine drainage contaminated water. Appl Microbiol Biotechnol. 99: 2911–2922. Tang X, Hashmi M Z, Long D, Chen L, Khan M I, Shen C. 2014. Influence of heavy metals and PCBs pollution on the enzyme activity and microbial community of paddy soils around an E-Waste Recycling Workshop. Int J Env Res Pub Heal. 11: 3118–3131. Vig K, Megharaj M, Sethunathan N, Naidu R. 2003. Bioavailability and toxicity of cadmium to microorganisms and their activities in soil: a review. Adv Environ Res. 8: 121–135. Wang Q, Garrity G M, Tiedje J M, Cole J R. 2007. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microb. 73: 5261–5267.