Advertisement

Environmental Science and Pollution Research

, Volume 26, Issue 3, pp 2531–2546 | Cite as

The chirality of imazethapyr herbicide selectively affects the bacterial community in soybean field soil

  • Hao Wu
  • Hongshan Chen
  • Chongwei Jin
  • Caixian Tang
  • Yongsong ZhangEmail author
Research Article

Abstract

The chiral herbicide imazethapyr (IM) is frequently used to control weeds in soybean fields in northeast China. However, the impact of IM enantiomers on microbial communities in soil is still unknown. Genetic markers (16S rRNA V3-V4 regions) were used to characterize and evaluate the variation of the bacterial communities potentially effected by IM enantiomers. Globally, the bacterial community structure based on the OTU profiles in (−)-R-IM-treated soils was significantly different from those in (+)-S-IM-treated soils, and the differences were enlarged with the treatment dose increasing. Interestingly, the Rhizobiaceae family and several other beneficial bacteria, including Bradyrhizobium, Methylobacterium, and Paenibacillus, were strongly enriched in (−)-R-IM treatment compared to (+)-S-IM treatment. In contrast, the pathogenic bacteria, including Erwinia, Pseudomonas, Burkholderia, Streptomyces, and Agrobacterium, were suppressed in the presence of (−)-R-IM compared to (+)-S-IM. Furthermore, we also observed that the bacterial community structure in (−)-R-IM-treated soils was more quickly restored to its original state compared with those in (+)-S-IM-treated soils. These findings unveil a new role of chiral herbicide in the development of soil microbial ecology and provide theoretical support for the application of low-persistence, high-efficiency, and eco-friendly optical rotatory (−)-R-IM.

Keywords

Imazethapyr Enantiomer 16s rRNA Bacterial community structure Phytopathogen Beneficial bacteria 

Introduction

Chirality is a common phenomenon in agricultural chemical productions. Approximately 40% of agrochemicals currently used in China are chiral pesticides, such as synthetic pyrethroids, organophosphorus insecticides, imidazolinones, and metolachlor herbicides (Zhou et al. 2009b). Imazethapyr (IM), with the highest market share of the six imidazolinones, has two enantiomers ((−)-R-IM and (+)-S-IM) due to a chiral imidazole moiety and a pyridine ring in the molecular structure, which inhibit the biosynthesis of branched chain amino acids by blocking acetolactate synthase (ALS) (Zabalza et al. 2007). Because of the relatively low toxicity and high efficiency, 3000–6000 tons of imazethapyr is used each year to control weeds in soybean fields in northeast China (Zabalza et al. 2007). Northeast China is a prime agricultural area, where the soil is under intensive agricultural use with one crop per annum, which necessitates the use of agrochemicals (Li 2005). Intensive use of IM has caused serious physiological injury to crops and the accumulation of residual IM in the environment, especially in soil with the higher clay and organic matter contents (Goetz et al. 1990). Interestingly, recent studies have shown that the inhibitory effects of the (−)-R-IM on many physiological and growth processes in plants are substantially higher than those of (+)-S-IM (Qian et al. 2009). This finding suggests that selecting (−)-R-IM may be a promising strategy to simultaneously improve the weeding ratio and reduce the import of hazardous IM into field soils. However, until now, little information has been available on the effects of IM enantiomers on microbe communities in soils. Microbes, especially bacteria, have fundamental roles in the functioning of most ecosystems (Falkowski et al. 2008), particularly in vast soil ecosystem (Boot et al. 2016). They dominated the cycling of nutrient elements and played a major role in maintaining soil quality (Murugan et al. 2014). The soil microbes also have great impact on plant growth. Many soil bacteria, such as rhizobia and plant growth–promoting bacteria (PGPB), are beneficial for plant growth (Maheshwari 2012), whereas there are also harmful bacteria, such as pathogenic bacteria, suppressing plant growth (Islam et al. 2016). Therefore, it is necessary to investigate the impacts of IM enantiomers on soil microbe communities, particularly the communities of beneficial bacteria and harmful bacteria, to provide more information for clarifying whether the application of (−)-R-IM in agricultural management is more suitable than (+)-S-IM.

In this study, we used a barcoded Illumina paired-end sequencing method targeting the V3 and V4 hypervariable regions of the 16S ribosomal RNA (rRNA) gene to investigate (I) whether the IM enantiomers can affect the diversity patterns of bacterial communities in soybean fields with a history of soybean cultivation, (II) whether the bacterial community structures were different in soils treated with the two enantiomers, and (III) which enantiomer was more friendly to beneficial bacteria in soybean field soils.

Material and method

Preparation and absolute configuration determination of IM enantiomers

Analytical standards of racemic imazethapyr (99% purity) (IUPAC name, 2-[4,5-dihydro-4-methyl-4-(1-methylethyl)-5-oxo-1Himidazol-2-yl]-5-ethyl-3-pyridinecarboxylic acid) were purchased from the Shenyang Research Institute of Chemical Industry (Shenyang, China). The stereo configuration of IM was given in Fig. 1(Zhou et al. 2009a). Other solvents or chemicals used in this study were of analytical or HPLC grade.
Fig. 1

Chemical structures of imazethapyr (IM)

Enantiomers were separated with the method developed in previous studies (Jenkins and Hedgepeth 2005; Lin et al. 2007; Zhou et al. 2009b). Briefly, the IM enantiomers extracted from the soil samples were detected using chiral separation on a cellulose tri (4-methylbenzoate (OJ-H) column-HPLC-CD detector system. The Agilent 1200 HPLC system (Agilent Technologies Company, Germany) was equipped with a vacuum degasser, a quaternary pump (QP-G1311A), an auto-sampler (TA-G1329A), a thermostatic column compartment (CT-G1316A), a multiple wavelength detector (VWD-G1314B), and a fluorescence detector (FLD-G1321A) for normal-phase analysis. The separation of (+)-S-IM and (−)-R-IM was performed using an OJ-H column with a normal mobile phase of n-hexane/ethanol/acetic acid (75/25/0.1, v/v/v) at a flow rate of 1 ml/min, an injection volume of 20 μl, and a UV-detector at a wavelength of 250 nm. The optimum column temperature was 25 °C. The circular dichroism (CD) detector (J-1100, Jasco, Tokyo, Japan) was operated at 250 nm for detection. Chromatographic data was acquired and processed with ChromPass software (Jasco, Tokyo, Japan). The first peak of the graph, the (+)-enantiomer, was eluted at 6 min, and the (−)-enantiomer peak was eluted at 9.5 min (Supporting Information Fig. S1a). The resolved enantiomers were manually collected in separate glass vials at 5 min and 8.5 min, respectively, and then dried under a stream of nitrogen and redissolved in ethanol. The purity and concentration of the recovered enantiomers were verified by HPLC under the same separation conditions (Supporting Information Fig. S1b). IM enantiomers were stable in low light conditions, and there were no signs of enantiomer conversion or degradation during the experiment.

The CD detector, based on an absorption difference between the right and left circularly polarized light, was coupled with HPLC to successfully characterize the absolute configuration of the enantiomers and the elution order of several chiral compounds (Jenkins and Hedgepeth 2005). The octant rule and CahnlngoldPrelog rule were used to establish the absolute configuration of the IM enantiomers. The CD spectra showed the stable configuration of the positive and negative Cotton IM enantiomers based on the MMFF94 force field calculation (Halgren 1996; Jenkins and Hedgepeth 2005). Thus, the absolute configuration of S-IM is (+)-IM, or (+)-S-IM. The absolute configuration of the other enantiomer R-IM is (−)-IM, or (−)-R-IM (Zhou et al. 2009b).

Soil sampling and characterization

The site chosen for sampling is an agricultural field located at the experimental farm of JiLin Agriculture University in northeast China (43°48′49.22″ N and 125°25′18.20″ E), which has been planted soybean for 5 years from 2008 to 2013 and had not been tilled or planted with crops in the following 2 years. The recommended concentration of IM was applied in this field from 2008 to 2011. The top 20 cm of the soils was collected from nine random subsites in June 2015. The field soils were air dried at room temperature by natural wind and then passed through a 2-mm sieve to remove the plant matter and small stones. The samples were then well-mixed in a portable agitator and stored at 4 °C until use. Their physicochemical characteristics are shown in Supporting Information Table S5.

Soil incubation experiment

Initially, the well-mixed soils were wetted to 50% maximum water-holding capacity (MWHC) and incubated at 25 °C for 15 days in the dark to avoid the impact of the explosive rise of microbial activity before pre-incubation. The soils were divided into five groups that each group included eight portions; Group-1 spiked with deionized water was set as negative controls (treatment: CK); Group-2 and Group-3 were respectively spiked with (+)-S-IM at a recommended field application concentration of 0.1 μg/g (200 g of dry weight) (treatment: S0.1) and 1 μg/g (200 g of dry weight; tenfold of the recommended concentration) (treatment: S1); Group-4 and Group-5 were respectively spiked with (−)-R-IM at 0.1 μg/g (treatment: R0.1) and 1 μg/g (treatment: R1) (Perucci and Scarponi 1994) (Supporting Information Table S1). A brown glass container containing 200.0 g dry soil from each portion and a small uncovered glass tube containing 15 ml of deionized water above the soil served as an incubating sample unit (Supporting Information Fig. S2). Each glass container, loosely capped to guarantee aerobic conditions, was incubated at 25 °C in a dark thermostatic chamber. Deionized water was added to each uncovered glass tube to maintain the 50% MWHC level if necessary.

Samples of each treatment were withdrawn at 0 days, 10 days, 30 days, 50 days, 70 days, and 90 days. Three grams of the incubated soil from each sample was collected and well-mixed every two repetitions to reduce repetitions from eight to four on 10-day, 50-day, and 90-day time nodes. Furthermore, eight repetitions from 0-day, 30-day, and 70-day time nodes were well-mixed into one testing sample. All standby soil testing samples were stored at − 80 °C for no more than 1 month before extracting the DNA. The total number of testing samples was 75, representing 5 treatment levels and 6 sampling times (10 days, 50 days, and 90 days in quadruplicate; 0 day, 30 days, and 70 days in one testing sample) (Supporting Information Table S2).

DNA isolation, PCR, and sequencing

For the MiSeq library preparation, soil microbial DNA were extracted by the PowerSoil DNA kit (Mo Bio Laboratories Inc., USA) following the manufacturer’s instructions. The DNA concentration was measured by Thermo Scientific NanoDrop 2000c using an UV-Vis spectrophotometer (Germany). The barcode Forward primer and barcode Reverse primer were used to amplify the bacterial 16S rRNA V3-V4 fragments (Klingenberg et al. 2013; Grahn et al. 2010) (Supporting Information Table S3). The 50-μl PCR reaction mix consisted of 25 μl of Phusion High-Fidelity PCR Master Mix (M0531, New England Biolabs Inc., USA), 3 μl of dimethyl sulfoxide (DMSO), 10 μl of 5 ng/μl template DNA, 3 μl of 1 μM barcode F primer, 3 μl of 1 μM barcode R primer, and 6 μl of nuclease free water. After denaturing at 98 °C for 30 s, the amplification was carried out with 30 cycles of 15 s at 98 °C, 15 s at 58 °C, 15 s at 72 °C, and a final extension step at 72 °C for 1 min. PCR products were purified using 0.85 × of AMPure XP beads (Agencourt), eluted in low Tris-EDTA (TE) buffer, and quantified using a Qubit dsDNA HS (high sensitivity) kit (Life Technologies). Equimolar amplicon suspensions were combined adequately for different samples. The mixed suspensions were then subjected to paired-end (PE, 2 × 300 bp) sequencing by an Illumina MiSeq Analyzer. Raw sequences were submitted to the NCBI database under the accession number PRJNA419867. After sequencing was complete, image analysis, base calling, and error estimation were performed using an Illumina Analysis Pipeline (version 2.6).

DNA sequence data preparation

The barcoded Illumina paired-end sequencing (BIPES) method (Zhou et al. 2010) was used to process the raw sequences and generate overlapped V3-V4 tags from Illumina Pair-end reads. Briefly, we used QIIME (Caporaso et al. 2010) to sort the sequences. The PE reads were overlapped by using the Needleman–Wunsch (NW) algorithm (Needleman and Wunsch 1970). We removed all the sequences that contained one or more ambiguous reads, those that contained any errors in the forward or reverse primers, and those with more than one mismatch within the 42 to 70-bp region during the overlap step. The variable tags that were shorter than 88 bp or longer than 117 bp were also removed. USEARCH61 was used to remove the chimera sequences by the UCHIME method (Edgar 2010; Edgar et al. 2011). The clean tags were taxonomically annotated with the RDP classifier 2.2 (Wang et al. 2007) based on the Greengenes taxonomy and a Greengenes reference database (currently version gg_13_5). Then, any sequences belonging to undesired taxa (Archaea, Eukaryota, Chloroplast and Mitochondria) were discarded. The resulting taxonomy files were clustered into operational taxonomic units (OTUs) by UCLUST (Edgar 2010, Edgar et al. 2011) at 97% identity. Bacterial groups in the OTUs table containing a total of < 5 reads were excluded.

OTU data diversity analysis

The alpha diversity analysis based on the Quantitative Insights into Microbial Ecology (QIIME) (Caporaso et al. 2010) was conducted to reveal the rarefaction curve plots and the diversity indices. The beta diversity analysis was carried out using UniFrac (Lozupone and Knight 2005; Lozupone et al. 2011) to compare the results of the PCoA (Bray–Curtis distance based) at the OTU level with the community ecology package. PCA was processed using R software (version 3.3.2) with the Prcomp function (vegan 2.0 package) at different microflora levels among the groups and time nodes. Student’s t test was processed to statistically compare differences between any two samples. Scatter Matrix Plots , bar and line charts were generated by OriginPro 2016. The hierarchical clustering analysis was performed using R software (version 3.3.2) with the Hclust function (stats package). All statistical analyses were conducted at P = 0.05.

Taxonomic data analysis

The relative proportions were calculated for each sample by dividing the total number of sequences for a specific bacterial group at a specific taxonomic level by the total number of sequences for that sample (White et al. 2017). The relative proportions were used to calculate the average proportion of a particular bacterial group at a specific taxonomic level for each treatment type (CK, R0.1, R1, S0.1, and S1), and then were used to statistically compare any two of the five samples at different incubation times. The calculated averages were also used to create stacked bar charts with Origin Pro 2016 to examine the bacterial community structure at the phylum level. The calculated averages were further processed to generate heat maps of the bacterial families in each treatment type (CK, R0.1, R1, S0.1 and S1) by using R software (version 3.3.2; R Foundation for Statistical Computing) with the Heatmap.2 function of the gplots package (version 2.12.1) (Warnes et al. 2005).

Results

High-throughput sequencing of 16S rRNA bacteria genes

Here, we amplified and sequenced the 16S V3–V4 variable regions from (I) control group soils (CK), (II) (−)-R-IM soils at an application rate of 0.1 μg/g (R0.1), (III) (−)-R-IM soils at 1 μg/g (R1), (IV) (+)-S-IM soils at 0.1 μg/g (S0.1), and (V) (+)-S-IM soils at 1 μg/g (S1) (Supporting Information Table S1). The 16S rRNA sequences (V3–V4) were processed through QIIME analysis pipeline (Supporting Information Table S2 and Fig. S3) involving R software (version 3.3.2) to obtain operational taxonomic units (OTUs) and taxonomic profiles (Caporaso et al. 2010). A total of 6.7 million sequence reads were obtained from 75 test samples collected from five different treatments at different incubation time nodes. After barcoded Illumina paired-end (PE) sequencing (BIPES) (Zhou et al. 2011) and chimera pruning (Edgar et al. 2011), 5,792,190 clean tags were gained with an average of 77,229.2 ± 21,577.3 tags per test sample. After the operational taxonomic units (OTUs) were clustered, we eliminated the very-low-abundance OTUs by removing those that had fewer than 5 reads in all 75 samples combined.

α- and β-diversity indices based on OTU profiles were affected by IM enantiomers in soil

The α-diversity indices based on the OTUs directly showed that the bacterial community structure was affected by IM enantiomers and was different in the (−)-R-IM samples compared to the (+)-S-IM samples (Table 1). Briefly, we found that the OTU diversity indices, the most direct indicator, were notably (P < 0.05) lower in S1 samples than other treatment samples, as well as the Shannon and Simpson diversity indices at three treatment times (Supporting Information Fig. S4). Furthermore, except for the OTU richness in R0.1 vs S0.1 at 10-day time node, the significant difference in these diversity indices between the (−)-R-IM and (+)-S-IM samples indeed existed irrespective of the application rate. These results clearly showed that IM enantiomers affected bacteria community diversity.
Table 1

Results (t test) comparing the diversity indices and relative abundance of the dominant bacterial phyla among the different treatments. Data from 10, 50, and 90 days of time were tested

 

CK

R0.1

R1

S0.1

S1

R0.1 vs R1 P value

S0.1 vs S1 P value

R0.1 vs S0.1 P value

R1 vs S1 P value

Bacteria diversity indices

10 days

OTUs richness

453.25 (23.795)

326** (41.725)

325.75** (33.229)

339** (31.804)

92.25*** (8.258)

0.994

0.000

0.683

0.000

Shann diversity

8.214 (0.051)

6.773** (0.533)

6.332*** (0.076)

4.721*** (0.350)

2.824*** (0.128)

0.206

0.000

0.001

0.000

Simpson evenness

0.681 (0.024)

0.572** (0.027)

0.395*** (0.028)

0.333*** (0.027)

0.585* (0.043)

0.000

0.000

0.000

0.001

50 days

OTUs richness

391.75 (14.202)

370.75 (36.072)

389.25 (37.499)

225*** (36.311)

91.5*** (30.451)

0.561

0.003

0.003

0.000

Shann diversity

7.913 (0.393)

7.553 (0.297)

6.936* (0.354)

6.033*** (0.323)

4.553*** (0.152)

0.054

0.000

0.001

0.000

Simpson evenness

0.460 (0.031)

0.287*** (0.024)

0.380* (0.022)

0.366** (0.030)

0.403** (0.037)

0.003

0.229

0.012

0.393

90 days

OTUs richness

327.5 (43.038)

297.25 (7.293)

248.75* (47.981)

148** (34.183)

77.25*** (5.356)

0.134

0.012

0.000

0.001

Shann diversity

7.674 (0.400)

7.564 (0.237)

6.989 (0.415)

5.369*** (0.227)

3.238*** (0.188)

0.051

0.000

0.000

0.000

Simpson evenness

0.523 (0.062)

0.532 (0.005)

0.513 (0.050)

0.550 (0.054)

0.492 (0.014)

0.542

0.049

0.111

0.346

Dominant bacteria phyla

10 days

Actinobacteria

0.166 (0.011)

0.161 (0.009)

0.155 (0.006)

0.227** (0.018)

0.270*** (0.013)

0.389

0.016

0.007

0.000

Proteobacteria

0.188 (0.035)

0.258* (0.011)

0.302** (0.004)

0.224 (0.023)

0.232 (0.012)

0.003

0.714

0.100

0.001

Acidobacteria

0.262 (0.026)

0.208* (0.017)

0.151** (0.006)

0.188* (0.029)

0.127*** (0.006)

0.006

0.023

0.371

0.009

50 days

Actinobacteria

0.158 (0.008)

0.211*** (0.005)

0.153 (0.010)

0.265** (0.006)

0.309*** (0.010)

0.000

0.003

0.010

0.005

Proteobacteria

0.223 (0.014)

0.297 (0.027)

0.403*** (0.017)

0.278* (0.018)

0.250 (0.017)

0.001

0.007

0.943

0.000

Acidobacteria

0.252 (0.012)

0.171*** (0.007)

0.144*** (0.019)

0.107*** (0.009)

0.110** (0.021)

0.041

0.809

0.003

0.053

90 days

Actinobacteria

0.100 (0.007)

0.142* (0.017)

0.138* (0.015)

0.247* (0.027)

0.308* (0.018)

0.798

0.116

0.005

0.000

Proteobacteria

0.230 (0.023)

0.308* (0.025)

0.362*** (0.017)

0.242 (0.027)

0.283** (0.016)

0.036

0.017

0.028

0.016

Acidobacteria

0.337 (0.029)

0.207** (0.025)

0.138*** (0.015)

0.170** (0.030)

0.170** (0.027)

0.016

0.989

0.177

0.151

Asterisk indicates a significant difference (* P < 0.05, ** P < 0.01, ***, P < 0.001) compared with the control check (CK)

Values in boldface indicate the dominant bacteria phylum

Values in bold italics indicate a significant difference (P < 0.05)

Stand errors are shown in parentheses

For the β-diversity index analysis, we achieved the ultimate goal of determining the influence of IM enantiomers on the soil bacterial community by Detrended Correspondence Analysis (DCA) (Fig. 2a, c, e) and Hierarchical Cluster Analysis (Fig. 2b, d, f). The first two axes for the DCA plots explained approximately 70–80% variance caused by the variation in the IM enantiomers and their applied concentration. Our first objective was to ascertain the differences in the bacterial community structure between the CK and enantiomer-treated groups. Both the DCA and hierarchical clustering analyses displayed clear separations between the CK and IM enantiomer–treated groups at the 10-, 50-, and 90-day time nodes (Fig. 2). Although the DCA plot showed a very close OTU profile between CK and R0.1 at the 90-day time node, the differences between the treated samples and CK samples were still noted as statistically significant (P < 0.05) based on permutational multivariate analysis of variance using Bray–Curtis distance matrices (Adonis; R software, package vegan) (Supporting Information Table S6). This indicated that the IM enantiomers affected the bacterial community structure at the OTU level, which was supported by principal component analysis (PCA), principal coordinate analysis (PCoA), constrained correspondence analysis (CCA), and constrained analysis of principal coordinates (Capscale) (Supporting Information Fig. S5–8).
Fig. 2

DCA and hierarchical clustering analyses based on OTU profiles. DCA and hierarchical clustering analyses of OTU profiles indicate the extent of dissimilarities between the control ( Open image in new window = CK) group soil without application of IM enantiomers and (−)-R-IM at 0.1 μg/g and 1 μg/g, and of (+)-S-IM at 0.1 μg/g and 1 μg/g ( Open image in new window = R0.1, Open image in new window = R1, Open image in new window = S0.1, Open image in new window = S1) applied to the enantiomer groups at 10-day (a, b), 50-day (c, d), and 90-day (e, f) time nodes. DCA1 and DCA2 represent the major axes of dissimilarity. Data points of the same sample type from different experiments are displayed by the same colors

Our second objective was to determine the differences in the bacterial community structure between an application concentration of 0.1 and 1 μg/g at different time nodes. Both DCA and hierarchical clustering analysis displayed complete separation in R0.1 vs. R1 and S0.1 vs. S1 (Fig. 2). Adonis analysis showed statistically significant (P < 0.05; Supporting Information Table S6) differences in the R0.1 vs R1 and S0.1 vs S1 samples at different time nodes. These indicated the variation in the applied concentration for IM enantiomers impacted the bacterial community structure. Further, the OTU variation (Bray-Distance; data not showed) in S0.1 vs. S1 was larger than that in R0.1 vs R1 at different time nodes, especially at 90-day time node. These observations were also verified via PCA, PCoA, CCA, and Capscale (Supporting Information Figs. S5–8).

Our third and most important objective was to determine the differences in bacterial community structure between the (−)-R-IM and (+)-S-IM samples. We observed complete separation for the (−)-R-IM samples and (+)-S-IM samples (Fig. 2; Compare R0.1 vs. S01 and R1 vs. S1), which were also identified as statistically significant based on Adonis (P < 0.05) at 10-, 50-, and 90-day time nodes (Supporting Information Table S6). Intriguingly, the OTU profiles in the CK samples were more similar (Euclidean-Distance P < 0.05) to that in the R-IM samples than that in the S-IM samples. These indicated that the S-IM had a more intense impact on the bacterial community structure and was not just significantly different with respect to R-IM. Additionally, the differences based on the OTUs (Bray-Distance; data not shown) in R1 vs S1 were significantly larger (P < 0.05) than that in R0.1 vs S0.1, suggesting the application concentration variation enlarged the differences between the R-IM and S-IM samples in the soils with a history of soybean cultivation.

IM enantiomers change bacteria taxa in soybean soil

After detecting the bacterial community structure based on the OTU profiles, we focused on the variation at specific taxonomic levels. First, we took a panoramic view of the situation on the bacterial taxa at the phylum level to find the distribution pattern of each treatment sample, which were similar among the different time nodes (Fig. 3). This indicated that the bacterial taxa at the phylum level remained relatively stable rather than exhibiting variation during the treatment period. Then, in contrast with CK, we observed that Proteobacteria (~ 26–40%) replaced Acidobacteria (~ 14–21%) to become the most dominant phylum in the R-IM samples while Actinobacteria (~ 23–30%) dominated the bacterial community in the S-IM samples (Table 1). Furthermore, these dominant phyla in the specific treatment samples were significantly (P < 0.05) higher than that of same phylum in other treatment samples, except that in R0.1 vs S0.1 at the 10- and 50-day time nodes. Despite this, the tendency toward variation of the most dominant phylum in the specific treatment was noticeable, especially in R1 vs S1. As far as the entire bacterial taxa structure at phylum level was concerned, the differences among the different treatment samples were still statistically (P < 0.05) significant based on Adonis, which was consistent with the results of the α- and β-diversity index analysis.
Fig. 3

The distribution of bacterial phyla in whole treatment period. Stacked bar graphs comparing the proportions of bacterial phyla from the control check (CK) samples and enantiomer applied groups (R0.1, R1, S0.1, and S1) at different times. “Other (< 1%)” includes the phyla whose average proportions account for < 1% of the bacterial community. The “Other < 1%” includes Gemmatimonadetes, Bacteroidetes, Firmicutes, Verrucomicrobia, Nitrospirae, Planctomycetes, WS3*, Cyanobacteria, OD1*, Armatimonadetes, Elusimicrobia, Chlorobi, WPS-2*, Tenericutes, AD3*, BRC1*, WS2*, Chlamydiae, Fibrobacteres, FBP*, TM6*, and WS4*. * indicate candidate phylum on phylogeny and physiology

A total of 170 families were detected for all the samples in the V3–V4 library, except “unassigned” or “others.” Heat maps were identified by calculating the variation from the average abundance in the individual treatment samples. Every grid in the column indicates different bacterial families; the green (or red) color indicates that the relative percentage of one family in one treatment was more (or less) than the average relative percentage of this family in all treatments in the row; the darker (or brighter) the color of one grid, the closer (or further away from) it gets to the average value of this family in the row (Fig. 4). Based on the heat maps of the specific bacterial families, we used Adonis method to calculate the differences of each treatments at bacterial family levels and observed a clear distinction (Adonis; P < 0.05) for every treatments at 10-, 50-, and 90-day time nodes. Furthermore, only 5 (3%, 10 days), 3 (2%, 50 days), and 5 (4%, 90 days) bacterial families showed little or no difference in the abundance across all treatments. Approximately 24% (38) bacterial families in the (−)-R-IM samples (R0.1 and R1) were higher than average abundance of each these families in all treatment samples, and 73% (115) bacterial families were at or below their average levels in the (−)-R-IM samples (R0.1 and R1) compare to CK and (+)-S-IM samples (S0.1 and S1) at 10-day time node (Fig. 4a). Then, these ratios changed to 38% (52) and 60% (83) at 50-day time node (Fig. 4b). By the end of the treatment period, only approximately 12% (16) bacterial families in (+)-S-IM samples were higher than the average abundance, while about 55% (75) bacterial families in the (−)-R-IM samples were higher than the average abundance (Fig. 4c). These indicated that (−)-R-IM enriched more bacterial families compared to (+)-S-IM in the soybean soils.
Fig. 4

The variation of bacterial families based on cluster analysis. Heat maps (ac) displaying enrichment (green) or reduction (red) from the average abundance (black) for the control check samples (CK), 0.1 μg/g (−)-R-IM samples (R0.1), 1 μg/g (−)-R-IM samples (R1), 0.1 μg/g (+)-S-IM samples (S0.1), and 1 μg/g (+)-S-IM samples (S1) at 10, 50, and 90-day time nodes separately. The heat map (a) consists of 158 families, heat map (b) consists of 138 families, and heat map (c) consists of 137 families

Then, we overlapped the significantly different bacterial families from R0.1 vs R1, S0.1 vs S1, and R-IM vs S-IM to determine the impact of IM enantiomers and their application concentration on the bacterial community structure at the family level. In contrast to R-IM vs S-IM, 41(30%) bacterial families were detected in both R0.1 vs S0.1 and R1 vs S1. However, we focused on 22 of these bacterial families and ignored the remaining 19 families (Supporting Information Table S4, green filled for 22 families and brownness filled for 19 families). For this reason, the 22 bacteria families showed the same variation tendency whereas 19 families showed the opposite variation tendency in R0.1 vs S0.1 and R1 vs S1. Seven of 22 bacterial families were significantly (P < 0.05) higher in their relative abundance in the S-IM than the R-IM samples, and the remaining 15 bacterial families showed the opposite result in their abundance (Fig. 5b, marked by blue arrow). This indicated that S-IM facilitated the colonization of seven bacterial families and suppressed the presence of 15 bacterial families in soils compared to R-IM. Intriguingly, Rhizobiaceae in relative abundance were significantly (P > 0.05) higher in (−)-R-IM and (+)-S-IM samples compared to CK. Furthermore, it was strongly enriched in (−)-R-IM samples relative to (+)-S-IM samples, especially in the whole treatment period (Fig. 5b and Supporting Information S13f). This indicated that (−)-R-IM provoked Rhizobiaceae to settle in soils, whereas (+)-S-IM weakened this group of bacteria. On the other hand, although the proportions of significantly different families in R0.1 vs R1 and S0.1 vs S1 were lower than that in R-IM vs S-IM, but they were close to 50% (Supporting Information Table S4). To our surprise, these families in the R0.1 vs R1 and S0.1 vs S1 shared only 10 (8%) families with that in the R-IM vs S-IM, although the 10 families occupied half of the population in R-IM vs S-IM (Fig. 5a; Supporting Information Table S4, purple filled). Moreover, no less than half of the families were unique regardless of the R0.1 vs R1 or S0.1 vs S1 (Supporting Information Table S4, yellow filled). These results indicate that the effects and the influencing mechanism of the variation in the application concentration of the IM enantiomers on the bacterial community structure differed from those caused by the change in enantiomers.
Fig. 5

The differences between (−)-R-IM and (+)-S-IM at family level. a Venn diagram showing the number of shared and unique bacterial families with comparisons of the R0.1 vs R1, S0.1 vs S1, and R vs S at the end of the treatment period. R0.1 vs R1 and S0.1 vs S1 indicate the number of families with significant difference between the 0.1 and 1 μg/g (−)-R-IM ((+)-S-IM) and 0.1 and 1 μg/g (+)-S-IM, respectively. The R vs S indicates the number of significantly different families with uniform changes in the trends with the comparison of R0.1 vs S0.1 and R1 vs S1. b Bar graphs comparing the relative abundance of the selected bacterial families with uniform changes in the trends from samples of (−)-R-IM and (+)-S-IM samples at 0.1 and 1 μg/g by the end of the treatment period. All families in the plots had statistically (P < 0.05) significant differences in comparison with the (−)-R-IM versus (+)-S-IM samples. Error bars indicate the standard deviation values. Blue arrows showed that these families were enriched in the (+)-S-IM samples compared to the (−)-R-IM samples

(−)-R-IM provoked beneficial bacteria while (+)-S-IM suppressed phytopathogenic bacteria

The variation of the beneficial bacteria and phytopathogens was emphatically investigated at the genus level. By the end of the treatment period, statistically different and significant genera were identified and categorized based on similar functions (Table 2; Supporting Information Table S7). The presence of (+)-S-IM compared to (−)-R-IM many folds increased the proportions of Pseudomonas (Molloy 2005), Erwinia (Toth et al. 2006), Burkholderia (Nguyen et al. 2017), Agrobacterium (Goodner et al. 2001), and Streptomyces (Bignell et al. 2010) which were identified as plant pathogenic bacteria, suggesting that (−)-R-IM inhibited the presence of these pathogenic bacteria in relation to (+)-S-IM. In contrast, Bradyrhizobium (Schneijderberg et al. 2018), Methylobacterium (Madhaiyan et al. 2015), Paenibacillus (Seldin 2011), and Rhizobium (Koskey et al. 2018), identified as plant beneficial bacteria in relative abundance, were multiplied by the presence of (−)-R-IM, suggesting that (−)-R-IM had high bio-availability rather than (+)-S-IM. In the degradative bacteria category, (−)-R-IM markedly promoted seven bacteria that were able to degrade xenobiotics to settle in soils relative to (+)-S-IM. However, Methylibium was much more sensitive to (+)-S-IM than (−)-R-IM. These indicated that more degradative bacteria were affected by (−)-R-IM compared to (+)-S-IM, which also provided evidence to our previous result that (−)-R-IM was preferred for degradation by soil microorganisms than (+)-S-IM (Wu et al. 2017).
Table 2

Results of the t tests to compare the proportions of functional bacterial genera between the (−)-R-IM and (+)-S-IM samples at the end of treatment (90 days)

Bacterial genus

R0.1

S0.1

R0.1 vs S0.1 P value

R1

S1

R1 vs S1 P value

Pathogenic bacteria

 Pseudomonas

0.00358 (0.00033)

0.05257 (0.00399)

0.000

0.00164 (0.00016)

0.04758 (0.00512)

0.000

 Erwinia

0.00039 (0.00049)

0.00211 (0.00141)

0.049

0

0.0024 (0.00171)

0.031

 Burkholderia

0.00072 (0.00056)

0.00281 (0.00009)

0.000

0.001189 (0.00032)

0.00416 (0.00061)

0.000

 Agrobacterium

0.00046 (0.00059)

0.00281 (0.00009)

0.000

0

0.00313 (0.00025)

0.000

 Streptomyces

0.00448 (0.0003)

0.01068 (0.0008)

0.000

0.00341 (0.00017)

0.03229 (0.00107)

0.000

Beneficial bacteria

 Bradyrhizobium

0.00115 (0.00014)

0

0.000

0.00157 (0.00036)

0

0.000

 Paenibacillus

0.02378 (0.00239)

0.00397 (0.00265)

0.000

0.03914 (0.00815)

0.00334 (0.00224)

0.001

 Rhizobium

0.00138 (0.00004)

0

0.000

0.00236 (0.00057)

0

0.000

Degradative bacteria

 Bacillus

0.10248 (0.00413)

0.03737 (0.00194)

0.000

0.02521 (0.00235)

0.00249 (0.00188)

0.000

 Phenylobacterium

0.00143 (0.00007)

0

0.000

0.01257 (0.00041)

0.00197 (0.00131)

0.000

 Methylibium

0.00024 (0.00048)

0.00353 (0.00236)

0.034

0.00045 (0.0009)

0.00262 (0)

0.003

 Methylobacterium

0.00143 (0.00007

0

0.000

0.00114 (0.00076)

0

0.024

 Geobacter

0.00969 (0.00022)

0.00225 (0.00171)

0.000

0.00186 (0.00048)

0

0.000

 Arthrobacter

0.01292 (0.00028)

0.00463 (0.00085)

0.000

0.01196 (0.00118)

0.00472 (0)

0.000

 Rhodococcus

0.00098 (0.00003)

0

0.000

0.00194 (0.0001)

0

0.000

 Hyphomicrobium

0.01057 (0.00029)

0.00148 (0.00189)

0.000

0.01319 (0.0006)

0.00197 (0.00131)

0.000

 Rhodoplanes

0.12479 (0.00113)

0.04442 (0.00355)

0.000

0.11572 (0.00659)

0.04966 (0.00218)

0.000

Numbers in boldface, P < 0.05

Stand errors were shown in parentheses

In response to the longer interval of the sampling time, five pathogenic bacteria (Burkholderia, Pseudomonas, Agrobacterium, Streptomyces, and Erwinia) compared to only two beneficial bacteria (Rhizobiaceae and Paenibacillus) showed clear separation trends over time (Supporting Information Fig. S9). Burkholderia showed the same reduced tendency for all the treatments, but different rates of degradation between R-IM and S-IM over time suggested that (+)-S-IM inhibited Burkholderia more strongly than (−)-R-IM during the whole treatment period (Supporting Information Fig. S9-b). The abundance of other pathogenic bacteria increased in the S-IM soil samples, whereas they were reduced or kept relatively stable in R-IM soil samples over time. The reverse was observed for Rhizobiaceae and Paenibacillus. By the end of the treatment period, (−)-R-IM decreased the colonization of beneficial bacteria to a lesser extent than the plant pathogenic bacteria relative to (+)-S-IM. Additionally, they showed clear separations between (−)-R-IM and (+)-S-IM at earlier than 50-day time node no matter if they were affected by (+)-S-IM, (−)-R-IM, or both, which agreed with our previous results about the IM half-lives of 42–64 days (Wu et al. 2017). This also provided evidence on the reliability of our previous results.

Bacterial community structure partly recovered in (−)-R-IM sample soil after treatment period

We observed that the bacterial community structure partially recovered in (−)-R-IM sample soils, especially in the R0.1 sample soil after 90-day treatment. The α-diversity indices in R-IM displayed significant difference from those of CK at 10-day time node. However, these differences lessened with time, and only the bacterial OTU richness in R1 was significantly different from that of CK (Table 1, see bacteria diversity indices). Following the analysis of the β-diversity indices, the differences in R-IM vs CK also displayed the same tendency, reaching a maximum at 50-day time node and lessening with time. Although the differences in R-IM vs CK were still noted as significantly different (Adonis; P < 0.05; Supporting Information Table S6), the tendency of R-IM, especially R0.1, compared to the CK was quite clear (Fig. 2c). Moreover, the heatmaps based on the abundance of the bacterial families displayed similar clustering patterns between 50- and 90-day time nodes, and the pattern at 50-day time node was more chaotic than that of the other time nodes (Fig. 4). We considered that maybe the impact of IM enantiomers on the soil microbial community structure at the family level was relatively strong at the start (10 days) and middle (50 days) of treatment and gradually weakening toward the end of the treatment (90 days).

In contrast, the recovery tendency of the bacterial community structure in the S-IM soil samples was not quite explicit and strong. The α-diversity indices in S-IM displayed significant differences with that in the CK from the start to the end, except for the Simpson evenness index at 90-day time node (Table 1, see Simpson evenness). The β-diversity indices also displayed separations of S-IM and CK that were consistently larger than that of R-IM and CK during the whole treatment period. Furthermore, 37% of the bacterial phyla and 35% of the bacterial families in R-IM displayed significant (P < 0.05) differences with CK at 10- and 50-day time nodes, and then recovered at the end of treatment period, whereas this was the case for only 19% of the bacterial phyla and 22% of the bacterial families in S-IM. The bacterial community structure in S-IM based on the relative proportion of the bacterial taxa was consistently significantly (Adonis; P < 0.05) different compared to CK from the start to the end of treatment. These results indicated that the recovery tendency in S-IM was weak in the whole treatment period and that (+)-S-IM had a more lasting impact on the bacterial community structure compared to (−)-R-IM.

Discussion

The soil bacterial community structure can be altered to display completely different community patterns by exogenous and endogenous chemical compounds. Interactions between exogenous chiral chemical compounds and soil microbes are subject to increasing interest as the need for high-efficiency sustainable agriculture and environmental preservation arises. Discovering changes in the soil microbial community due to chiral herbicides is the first step to achieving such goals. However, to our knowledge, no earlier studies had comprehensively examined the impact of the chiral herbicide imazethapyr enantiomers on the plant-associated bacterial communities. In this study, we performed an in-depth characterization of the compositions of plant and material cycle–associated bacteria in bulk soil with a history of soybean cultivation and gained insights into the effects of IM enantiomers on each of these bacterial communities. The studied bacterial community structure was significantly different among the control check soils, (−)-R-IM, and (+)-S-IM affected soils.

Our α- and β-diversity analysis results, as well as the bacterial taxa analysis, mostly supported the hypothesis that the bacterial community structures in the CK, (−)-R-IM, and (+)-S-IM were different from each other. Similarly, Qian et al. (2009) reported α-diversity indices based on the OTU profiles in (−)-R-IM, (+)-S-IM, and the control samples, which were different from each other in Arabidopsis thaliana rhizosphere (Qian et al. 2009; Zhou et al. 2009b). This suggested that, as exogenous and physiological active substances, the IM enantiomers could change the soil bacterial community structure. The bacterial responses to (−)-R-IM on the soil bacterial communities are different with (+)-S-IM due to the significant differences in the bio-availability of (−)-R-IM and (+)-S-IM (Qian et al. 2009). The (−)-R-IM with higher bio-availability could interact with bacteria more easily and intensely compare to the (+)-S-IM, which may lead to the preferential change on the bacteria community structure and the preferential degradation of (−)-R-IM (Lehmann et al. 2010). One possible reason of the different responses of bacteria to IM enantiomers is that IM enantiomers can play an accelerant or suppressant for some specific bacteria to make these bacteria of enrichment or inhibition in soil. For example, the Proteobacteria phylum and Rhizobiaceae family were enriched in the presence of (−)-R-IM and inhibited in the presence of (+)-S-IM, whereas the Actinobacteria phylum and the Erwinia genus were enriched in the presence of (+)-S-IM and suppressed in the presence of (−)-R-IM in soils. The enrichment of Proteobacteria and Actinobacteria in abundance had a direct relationship with the presence of (−)-R-IM and (+)-S-IM, which was similar in nonylphenol, metal cyanide, glyphosate, and even 2,4,6-trinitrotoluene (TNT)-contaminated soil (Baxter and Cummings 2006b; George et al. 2009; Newman et al. 2016; Wang et al. 2015). In contrast, the addition of bromoxynil inhibited photosynthesis, but the enzyme acetohydroxy acid synthase suppressed the presence of Proteobacteria to make Acidobacteria and Firmicutes the prominent phyla in the soil community (Baxter and Cummings 2006a). Another possible reason for different responses is that some specific bacteria can resist the physiological damage caused by IM enantiomers, which make those resistant bacteria survived and enriched in soil. For example, Streptococcus can resist (+)-QPT-1 but are inhibited and killed by (−)-QPT-1 which is a structurally novel class of bacterial topoisomerase II (TopoII) inhibitors (Miller et al. 2008). These indicated the intense variation in the abundance of these bacteria that are sensitive to particular exogenous chemical compounds ultimately leading to changes in the bacterial community structure. However, how one or multiple chemical compounds interact with the bacteria to shift the dominant bacteria is uncertain on a deeper level, let alone how the presence of a single IM enantiomer changes the soil bacterial community structure.

Even so, the effect of (−)-R-IM and (+)-S-IM on the bacterial community structure as well as the specific families and genera in the soybean soils is still worthy of study to get closer to our goal. Our results revealed the different influences of (−)-R-IM and (+)-S-IM on the beneficial bacteria at the family and genus levels. The Rhizobiaceae family and Bradyrhizobium genus belong to the Proteobacteria phylum, which contain multiple subgroups that enhance plant development (Spaink et al. 1998; Vessey 2003), and most of them are diazotrophs, which are able to fix nitrogen and are symbiotic with plant roots (Antoun et al. 1998; Van et al. 1985). Previous study showed that (R:S/50:50)-IM can inhibit acetolactate synthase and nitrogenase in Rhizobium and Bradyrhizobium, and affect their symbiosis with legume crops (pea, horse bean, yellow lupine, white lupine, soybean) (Parsa et al. 2013; Royuela et al. 2015; Sawicka and Selwet 1998). However, we observed that they significantly increased in the presence of (−)-R-IM, especially in the R1 samples at 50- and 90-day time nodes. These indicated that maybe the inhibition of (+)-S-IM on Rhizobium and Bradyrhizobium was possibly stronger than the acceleration of (−)-R-IM on them, despite the molecular mechanism is unknown. For example, the inhibition of the chiral herbicide enantiomer (R)-diclofop acid on Microcystis aeruginosa was stronger than the promotion of (S)-diclpfop acid on them, because (R)-diclofop acid probably acts as a proton ionophore shuttling protons across the plasmalemma, whereas (S)-diclofop acid did not demonstrate such action (Ye et al. 2013). Another possible reason is that (+)-S-IM suppress Rhizobium and Bradyrhizobium while (−)-R-IM have no effect on them. For the result in our study, it may be that the promotion of (−)-R-IM on the abundance of the Rhizobiaceae family and Bradyrhizobium genus is more intense than that of (+)-S-IM on them. For example, the acceleration of (S)-Metolachlor on Bacillus simplex and Candida xestobii were stronger than that of (R)-Metolachlor to represent oxygen uptake rates more than two times higher upon incubation with the (S)-Metolachlor than upon incubation with the (S)-Metolachlor (Munoz et al. 2011). Similar with Rhizobiaceae, the Enterobacteriaceae and Nostocaceae families also contain nitrogen fixers, such as Klebsiella (Brisse et al. 2006; Cakmakci et al. 1981), Cylindrospermum (Castenholz 1973), Anabaena (Herrero and Flores 2007), and Nostoc (Dodds et al. 2010), which are capable of providing essential sources of nitrogen for soybeans. Furthermore, many species belong to the Paenibacillus genus, which serves as an efficient plant growth–promoting (e.g., phosphate solubilization, nitrogen fixation, pollutants degradation, and phytopathogens control) rhizobacteria (PGPR), that competitively colonizes plant roots and can simultaneously act as a biofertilizer and as an antagonist (biopesticides) of recognized root pathogens, such as bacteria, fungi, and nematodes (Debashan et al. 2012; Glick 2012; Sood 2010; Souza et al. 2015). It is amazing that Paenibacillus multiplied to > 1% in the (−)-R-IM groups which is significantly higher than that in the (+)-S-IM groups in bulk soil. Thus, the colonization of these families in the soils indicated that (−)-R-IM was friendly with plants, especially leguminous plants.

Furthermore, our results also revealed some degradative bacteria that colonized under the effects of the IM enantiomers. Some of them were identified to have the capability to degrade members of the imidazolinones herbicide. For example, Imazaquin was degraded into quinoline-2,3-dicarboxylic anhydride by Arthrobacter crystallopoietes (Wang et al. 2007). Acinetobacter baumannii solely applicated imazamox (Liu et al. 2016). Pseudomonas fluorescenes and Bacillus cereus could efficiently degrade imazapyr as sole carbon resource (Xuedong et al. 2005). Numerous Arthrobacter species have been examined for their potential to degrade xenobiotics and other harmful substances, such as PAHs (Li et al. 2015), triazine herbicides (Bazhanov et al. 2017; Guo et al. 2014; Zhang et al. 2015), quinaldine (Zhang et al. 2017), and substituted phenyl urea herbicides (Guo et al. 2014). Degradative bacteria Bacillus, Phenylobacterium, Methylobacterium, Geobacter, Arthrobacter, Rhodococcus, Hyphomicrobium, and Rhodoplanes in this study were enriched in the (−)-R-IM soil samples compared to the (+)-S-IM soil samples, which were evident in the preferentially degrading of (−)-R-IM in our previous study (Wu et al. 2017). However, Methylibium preferred to colonize (+)-S-IM soils, and Methylibium petroleiphilum, the methyl tert-butyl ether depredator (Schäfer et al. 2011), was conductive to exploring the enantio-selective degrading bacteria for (+)-S-IM. Thus, they will possibly help us to reduce the hazards of IM and increase the quality and safety of the crops in agricultural production.

On the other hand, based on the distinction of (−)-R-IM and (+)-S-IM in bio-availability, we pessimistically hypothesized that higher bio-available (−)-R-IM also promotes pathogenic bacteria to settle in soils. However, to our delight, our results were the opposite of this hypothesis. The abovementioned Rhizobiaceae contain species capable of fixing N2 to ammonia for soybeans, but also contain the phytopathogenic Agrobacterium genus. Although Agrobacterium was famous as an important tool for genetic engineering, it caused crown-gall disease in plants, especially for dicotyledon (Brown et al. 2018). The decrease of Agrobacterium in the presence of (−)-R-IM may indicate that this plant pathogenic genus is normally suppressed by (−)-R-IM. Not only Agrobacterium but also Pseudomonas, Streptomyces, Erwinia, and Burkholderia caused huge negative effects on the growth and development of plants, even animals and humans, as a pathogen, and were reduced in abundance in the (−)-R-IM groups compared to the (+)-S-IM groups during the treatment period. The Erwinia genus, named for the famous plant pathologist, infected woody plants, causing fire blight in apples, pears, and other rosaceae crops (Norelli et al. 2007; Vrancken et al. 2013) as well as bacterial wilt in cucurbits, tomato, and tobacco (Prior et al. 2010; Słomnicka et al. 2015; Tanaka et al. 2009) which was also caused by the Pseudomonas genus. Moreover, Pseudomonas still contained a species of plant pathogen named P. syringae that invaded the plant cell by secreting effector proteins. It was an important model system for the experimental characterization of the molecular dynamics of the plant–pathogen interactions due to their ability to suppress plant defense and select strains that caused disease based on well-characterized host plants (e.g., Arabidopsis thaliana, Nicotiana benthamiana, and tomato) (He et al. 2004; Jamali et al. 2009; Meaden and Koskella 2017). Thus, the possible reason for this phenomenon is that the (−)-R-IM could damage cell structure and retard vital physiological process (e.g., inhibit functional enzyme, protein synthesis, and genetic information transmission) to suppress or kill them, but could not for (+)-S-IM. For example, the effect of chiral herbicide metalaxy on selectively interfering with the synthesis of ribosomal RNA was mainly derived from R-metalaxy (Gu et al. 2018; Xie and Yang 2018), and R-(−)-diniconazole strongly inhibited α-sterol demethylase demethylation (P-450 DM), while S isomer was a valid inhibitor (Dong et al. 2013; Wang et al. 2014). The higher phytocidal bio-availability of (−)-R-IM profited those positive bacteria to colonize and get rid of contamination and promote plant growth, as well as suppress pathogenic bacteria to reduce environmental loss caused by pollution and plant damages as far as possible, suggesting that maybe (−)-R-IM was a more efficient and friendlier rotatory enantiomer for environmental sustainability.

Intriguingly, we also observed that the recovery tendency of the bacterial community structure in the (−)-R-IM samples was stronger than that in the (+)-S-IM samples at the later period. This phenomenon indicated that the effect duration of (−)-R-IM on the bacterial community structure was shorter than that of (+)-S-IM. This could be because (−)-R-IM was preferentially bio-degraded by soil microbial organism, which was confirmed in our earlier study (Wu et al. 2017). In that study, the half-lives of (−)-R-IM (42–53 days) were shorter than that of (+)-S-IM (55–64 days), which means that the impact of (−)-R-IM on the soil bacterial community structure disappeared faster than that of (+)-S-IM. The distinction of the degradation rate of (−)-R-IM and (+)-S-IM could be the most important factor for the recovery capability of the soil bacterial community. Similarly, cis-nitromethylene neonicotinoid insecticide Paichongding (IPP) is a novel chiral herbicide for controlling weeds or killing insects; the degradation ratios of RS-IPP and SR-IPP were > 90%, higher than that of RR-IPP and SS-IPP at 60-day in yellow paddy soil. The recovery tendency of bacterial community structure was more significantly in RS-IPP and SR-IPP applied soils than that in RR-IPP and SS-IPP applied soils (Cai et al. 2015; Cai et al. 2016). Thus, this soil bacterial community recovery characteristic in (−)-R-IM-treated soil may be helpful to reduce the negative environmental effects caused by IM overuse.

In summary, our study unveiled high bacterial differences and the subversion of bacterial taxa within (−)-R-IM and (+)-S-IM soils. These differences were enlarged by the increase of the application concentration of IM enantiomers. Furthermore, (−)-R-IM facilitates increases in the colonization of plant growth–promoting bacteria and hazardous substance–degrading bacteria but suppresses the colonization of phytopathogenic bacteria compared with (+)-S-IM. Furthermore, the soil after (−)-R-IM treatment more quickly was restored its original bacterial community structure compared with the soil after (+)-S-IM treatment. These findings, together with the preferential bio-degradation of (−)-R-IM by soil microbial organism (Wu et al. 2017) and higher phytocidal bio-availability of (−)-R-IM on weed plants (Qian et al. 2009), provide strong encouragement for the exploitation and application of the low-toxicity, high-efficiency, and eco-friendly rotatory imazethapyr enantiomer.

Notes

Funding information

This research was funded by the Natural Science Foundation of China (NSFC, No. 41071314).

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

11356_2018_3736_MOESM1_ESM.doc (1.4 mb)
ESM 1 (DOC 1484 kb)
11356_2018_3736_MOESM2_ESM.xlsx (10 kb)
ESM 2 (XLSX 9 kb)
11356_2018_3736_MOESM3_ESM.xlsx (13 kb)
ESM 3 (XLSX 13 kb)
11356_2018_3736_MOESM4_ESM.xlsx (44 kb)
ESM 4 (XLSX 43 kb)
11356_2018_3736_MOESM5_ESM.xlsx (24 kb)
ESM 5 (XLSX 23 kb)
11356_2018_3736_MOESM6_ESM.xlsx (9 kb)
ESM 6 (XLSX 9 kb)
11356_2018_3736_MOESM7_ESM.xlsx (12 kb)
ESM 7 (XLSX 11 kb)
11356_2018_3736_MOESM8_ESM.xlsx (17 kb)
ESM 8 (XLSX 17 kb)

References

  1. Antoun H, Beauchamp CJ, Goussard N, Chabot R, Lalande R (1998) Potential of Rhizobium and Bradyrhizobium species as plant growth promoting rhizobacteria on non-legumes: effect on radishes (Raphanus sativus L.). In: Hardarson G, Broughton WJ (eds) Molecular microbial ecology of the soil: results from an FAO/IAEA co-ordinated research programme, 1992–1996. Springer, Netherlands, pp 57–67CrossRefGoogle Scholar
  2. Baxter J, Cummings SP (2006a) The application of the herbicide bromoxynil to a model soil-derived bacterial community: impact on degradation and community structure. Lett Appl Microbiol 43:659–665CrossRefGoogle Scholar
  3. Baxter J, Cummings SP (2006b) The impact of bioaugmentation on metal cyanide degradation and soil bacteria community structure. Biodegradation 17:207–217CrossRefGoogle Scholar
  4. Bazhanov DP, Yang K, Li H, Li C, Li J, Chen X, Yang H (2017) Colonization of plant roots and enhanced atrazine degradation by a strain of Arthrobacter ureafaciens. Appl Microbiol Biotechnol 101:6809–6820CrossRefGoogle Scholar
  5. Bignell DRD, Huguet-Tapia JC, Joshi MV, Pettis GS, Loria R (2010) What does it take to be a plant pathogen: genomic insights from Streptomyces species. Antonie Van Leeuwenhoek 98:179–194CrossRefGoogle Scholar
  6. Boot CM, Hall EK, Denef K, Baron JS (2016) Long-term reactive nitrogen loading alters soil carbon and microbial community properties in a subalpine forest ecosystem. Soil Biol Biochem 92:211–220CrossRefGoogle Scholar
  7. Brisse S, Grimont F, Grimont PAD (2006) The genus klebsiella. Prokaryotes 32:159–196CrossRefGoogle Scholar
  8. Brown P, Attai H, Boon M, Phillips K, Noben JP, Lavigne R (2018) Larger than life: isolation and genomic characterization of a jumbo phage that infects the bacterial plant pathogen, Agrobacterium tumefaciens. Front Microbiol 9Google Scholar
  9. Cai Z, Wang J, Ma J, Zhu X, Cai J, Yang G (2015) Anaerobic degradation pathway of the novel chiral insecticide paichongding and its impact on bacterial communities in soils. J Agric Food Chem 63:7151–7160CrossRefGoogle Scholar
  10. Cai Z, Yan R, Chen J, Wang J, Ma J, Zhang W, Zhao X (2016) Effects of the novel cis-nitromethylene neonicotinoid insecticide paichongding on enzyme activities and microorganisms in yellow loam and Huangshi soils. Environ Sci Pollut Res 23:7786–7793CrossRefGoogle Scholar
  11. Cakmakci ML, Evans HJ, Seidler RJ (1981) Characteristics of nitrogen-fixing Klebsiella oxytoca isolated from wheat roots. Plant Soil 61:53–63CrossRefGoogle Scholar
  12. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer N, Peña AG, Goodrich JK, Gordon JI, Huttley GA, Kelley ST, Knights D, Koenig JE, Ley RE, Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder J, Sevinsky JR, Turnbaugh PJ, Walters WA, Widmann J, Yatsunenko T, Zaneveld J, Knight R (2010) QIIME allows analysis of high-throughput community sequencing data. Nat Methods 7:335–336CrossRefGoogle Scholar
  13. Castenholz R (1973) The biology of blue-green algae. Sci Act 9:320–339Google Scholar
  14. Debashan LE, Hernandez JP, Bashan Y (2012) The potential contribution of plant growth-promoting bacteria to reduce environmental degradation—a comprehensive evaluation. Appl Soil Ecol 61:171–189CrossRefGoogle Scholar
  15. Dodds WK, Gudder DA, Mollenhauer D (2010) The ecology of nostoc. J Phycol 31:2–18CrossRefGoogle Scholar
  16. Dong F, Li J, Chankvetadze B, Cheng Y, Xu J, Liu X, Li Y, Chen X, Bertucci C, Tedesco D (2013) Chiral triazole fungicide difenoconazole: absolute stereochemistry, stereoselective bioactivity, aquatic toxicity, and environmental behavior in vegetables and soil. Environ Sci Technol 47:3386–3394CrossRefGoogle Scholar
  17. Edgar RC (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26:2460–2461CrossRefGoogle Scholar
  18. Edgar RC, Haas BJ, Clemente JC, Christopher Q, Rob K (2011) UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 27:2194–2200CrossRefGoogle Scholar
  19. Falkowski PG, Fenchel T, Delong EF (2008) The microbial engines that drive earth’s biogeochemical cycles. Science 320:1034–1039CrossRefGoogle Scholar
  20. George IF, Liles MR, Hartmann M, Ludwig W, Goodman RM, Agathos SN (2009) Changes in soil Acidobacteria communities after 2,4,6-trinitrotoluene contamination. FEMS Microbiol Lett 296:159–166CrossRefGoogle Scholar
  21. Glick BR (2012) Plant growth-promoting bacteria: mechanisms and applications. Scientifica 2012:963401CrossRefGoogle Scholar
  22. Goetz AJ, Lavy TL, Gbur EE (1990) Degradation and field persistence of imazethapyr. Weed Sci 38:421–428CrossRefGoogle Scholar
  23. Goodner B, Hinkle G, Gattung S, Miller N, Blanchard M, Qurollo B, Goldman BS, Cao Y, Askenazi M, Halling C, Mullin L, Houmiel K, Gordon J, Vaudin M, Iartchouk O, Epp A, Liu F, Wollam C, Allinger M, Doughty D, Scott C, Lappas C, Markelz B, Flanagan C, Crowell C, Gurson J, Lomo C, Sear C, Strub G, Cielo C, Slater S (2001) Genome sequence of the plant pathogen and biotechnology agent agrobacterium tumefaciens C58. Science 294:2323–2328CrossRefGoogle Scholar
  24. Grahn N, Olofsson M, Ellnebo-Svedlund K, Monstein HJ, Jonasson J (2010) Identification of mixed bacterial DNA contamination in broad-range PCR amplification of 16S rDNA V1 and V3 variable regions by pyrosequencing of cloned amplicons. FEMS Microbiol Lett 219:87–91CrossRefGoogle Scholar
  25. Gu J, Ji C, Yue S, Shu D, Su F, Zhang Y, Xie Y, Zhang Y, Liu W, Zhao M (2018) Enantioselective effects of metalaxyl enantiomers in adolescent rat metabolic profiles using NMR-based metabolomics. Environ Sci Technol 52Google Scholar
  26. Guo Q, Zhang J, Wan R, Xie S (2014) Impacts of carbon sources on simazine biodegradation by Arthrobacter strain SD3-25 in liquid culture and soil microcosm. Int Biodeterior Biodegrad 89:1–6CrossRefGoogle Scholar
  27. Halgren TA (1996) Merck molecular force field. I. Basis, form, scope, parameterization, and performance of MMFF94. J Comput Chem 17:490–519CrossRefGoogle Scholar
  28. He J, Baldini RL, Déziel E, Saucier M, Zhang Q, Liberati NT, Lee D, Urbach J, Goodman HM, Rahme LG (2004) The broad host range pathogen Pseudomonas aeruginosa strain PA14 carries two pathogenicity islands harboring plant and animal virulence genes. Proc Natl Acad Sci U S A 101:2530–2535CrossRefGoogle Scholar
  29. Herrero A, Flores E (2007) The cyanobacteria: molecular biology, genomics and evolution. Caister Academic PressGoogle Scholar
  30. Islam S, Akanda AM, Prova A, Islam MT, Hossain MM (2016) Isolation and identification of plant growth promoting Rhizobacteria from cucumber rhizosphere and their effect on plant growth promotion and disease suppression. Front Microbiol 6:1360CrossRefGoogle Scholar
  31. Jamali F, Sharifi-Tehrani A, Lutz MP, Maurhofer M (2009) Influence of host plant genotype, presence of a pathogen, and coinoculation with Pseudomonas fluorescens strains on the rhizosphere expression of hydrogen cyanide- and 2,4-diacetylphloroglucinol biosynthetic genes in P. fluorescens biocontrol strain CHA0. Microb Ecol 57:267–275CrossRefGoogle Scholar
  32. Jenkins AL, Hedgepeth WA (2005) Analysis of chiral pharmaceuticals using HPLC with CD detection. Chirality 17:S24–S29CrossRefGoogle Scholar
  33. Klingenberg BK, Michael R, Hestbjerg HL, Susanne S, Thor LS, Johannes SS, Angeliki KK (2013) The murine lung microbiome in relation to the intestinal and vaginal bacterial communities. BMC Microbiol 13:303–303CrossRefGoogle Scholar
  34. Koskey G, Simon MW, Kimiti JM, Ombori O, Maingi JM, Njeru EM (2018) Genetic characterization and diversity of rhizobium isolated from root nodules of mid-altitude climbing bean (Phaseolus vulgaris L.) varieties. Front Microbiol 9Google Scholar
  35. Lehmann K, Crombie A, Singer AC (2010) Reproducibility of a microbial river water community to self-organize upon perturbation with the natural chemical enantiomers, R- and S-carvone. FEMS Microbiol Ecol 66:208–220CrossRefGoogle Scholar
  36. Li J (2005) Issues and some strategies of sustainable development of black soil resources in black soil zone in northeast China. Chin Agric Sci Bull 21:352–352Google Scholar
  37. Li F, Zhu L, Wang L, Zhan Y (2015) Gene expression of an Arthrobacter in surfactant-enhanced biodegradation of a hydrophobic organic compound. Environ Sci Technol 49:3698–3704CrossRefGoogle Scholar
  38. Lin K, Xu C, Zhou S, Liu W, Gan J (2007) Enantiomeric separation of imidazolinone herbicides using chiral high-performance liquid chromatography. Chirality 19:171–178CrossRefGoogle Scholar
  39. Liu CG, Yang X, Lai Y, Hong-Gang LU, Zeng WM, Geng G, Yang FS (2016) Imazamox microbial degradation by common clinical bacteria: acinetobacter baumannii IB5 isolated from black soil in China shows high potency. J Integr Agric 15:1798–1807CrossRefGoogle Scholar
  40. Lozupone C, Knight R (2005) UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol 71:8228–8235CrossRefGoogle Scholar
  41. Lozupone C, Lladser ME, Dan K, Stombaugh J, Knight R (2011) UniFrac: an effective distance metric for microbial community comparison. ISME J 5:169–172CrossRefGoogle Scholar
  42. Madhaiyan M, Tan HHA, Si TN, Prithiviraj B, Ji L (2015) Leaf-residing methylobacterium species fix nitrogen and promote biomass and seed production in Jatropha curcas. Biotechnol Biofuels 8:222CrossRefGoogle Scholar
  43. Maheshwari DK (2012) Bacteria in agrobiology: plant probiotics. Springer, BerlinCrossRefGoogle Scholar
  44. Meaden S, Koskella B (2017) Adaptation of the pathogen, Pseudomonas syringae, during experimental evolution on a native vs. alternative host plant. Mol Ecol 26:1790–1801CrossRefGoogle Scholar
  45. Miller AA, Bundy GL, Mott JE, Skepner JE, Boyle TP, Harris DW, Hromockyj AE, Marotti KR, Zurenko GE, Munzner JB, Sweeney MT, Bammert GF, Hamel JC, Ford CW, Zhong WZ, Graber DR, Martin GE, Han F, Dolak LA, Seest EP, Ruble JC, Kamilar GM, Palmer JR, Banitt LS, Hurd AR, Barbachyn MR (2008) Discovery and characterization of QPT-1, the progenitor of a new class of bacterial topoisomerase inhibitors. Antimicrob Agents Chemother 52:2806–2812CrossRefGoogle Scholar
  46. Molloy S (2005) Signalling complexities for Pseudomonas. Nat Rev Microbiol 3:192CrossRefGoogle Scholar
  47. Munoz A, Koskinen WC, Cox L, Sadowsky MJ (2011) Biodegradation and mineralization of metolachlor and alachlor by Candida xestobii. J Agric Food Chem 59:619–627CrossRefGoogle Scholar
  48. Murugan R, Beggi F, Kumar S (2014) Belowground carbon allocation by trees, understory vegetation and soil type alter microbial community composition and nutrient cycling in tropical Eucalyptus plantations. Soil Biol Biochem 76:257–267CrossRefGoogle Scholar
  49. Needleman SB, Wunsch CD (1970) A general method applicable to the search for similarities in the amino acid sequence of two proteins. J Mol Biol 48:443–453CrossRefGoogle Scholar
  50. Newman MM, Hoilett N, Lorenz N, Dick RP, Liles MR, Ramsier C, Kloepper JW (2016) Glyphosate effects on soil rhizosphere-associated bacterial communities. Sci Total Environ 543:155–160CrossRefGoogle Scholar
  51. Nguyen TT, Lee H-H, Park J, Park I, Seo Y-S (2017) Computational identification and comparative analysis of secreted and transmembrane proteins in six Burkholderia species. Plant Pathol J 33:148–162CrossRefGoogle Scholar
  52. Norelli JL, Jones AL, Aldwinckle HS (2007) Fire blight management in the twenty-first century: using new technologies that enhance host resistance in apple. Plant Dis 87:756–765CrossRefGoogle Scholar
  53. Parsa M, Aliverdi A, Hammami H (2013) Effect of the recommended and optimized doses of haloxyfop-p-methyl or imazethapyr on soybean-bradyrhizobium japonicum symbiosis. Ind Crop Prod 50:197–202CrossRefGoogle Scholar
  54. Perucci P, Scarponi L (1994) Effects of the herbicide imazethapyr on soil microbial biomass and various soil enzyme-activities. Biol Fertil Soils 17:237–240CrossRefGoogle Scholar
  55. Prior P, Bart S, Leclercq S, Darrasse A, Anais G (2010) Resistance to bacterial wilt in tomato as discerned by spread of Pseudomonas (Burholderia) solanacearum in the stem tissues. Plant Pathol 45:720–726CrossRefGoogle Scholar
  56. Qian H, Hu H, Mao Y, Ma J, Zhang A, Liu W, Fu Z (2009) Enantioselective phytotoxicity of the herbicide imazethapyr in rice. Chemosphere 76:885–892CrossRefGoogle Scholar
  57. Royuela M, Gonzalez A, Arrese-Igor C, Aparicio-Tejo PM, Gonzalez-Murua C (2015) Imazethapyr inhibition of acetolactate synthase in rhizobium and its symbiosis with pea. Pest Manag Sci 52:372–380CrossRefGoogle Scholar
  58. Sawicka A, Selwet M (1998) Effect of active ingredients on rhizobium and bradyrhizobium legume dinitrogen fixation. Pol J Environ Stud 7:317–320Google Scholar
  59. Schäfer F, Muzica L, Schuster J, Treuter N, Rosell M, Harms H, Müller RH, Rohwerder T (2011) Formation of alkenes via degradation of tert-alkyl ethers and alcohols by Aquincola tertiaricarbonis L108 and Methylibium spp. Appl Environ Microbiol 77:5981–5987CrossRefGoogle Scholar
  60. Schneijderberg M, Schmitz L, Cheng X, Polman S, Franken C, Geurts R, Bisseling T (2018) A genetically and functionally diverse group of non-diazotrophic Bradyrhizobium spp. colonizes the root endophytic compartment of Arabidopsis thaliana. BMC Plant Biol 18:61CrossRefGoogle Scholar
  61. Seldin L (2011) Paenibacillus, nitrogen fixation and soil fertility. In: Logan NA, Vos P (eds) Endospore-forming soil bacteria. Springer, Berlin, pp 287–307CrossRefGoogle Scholar
  62. Słomnicka R, Olczak-Woltman H, Bartoszewski G, Niemirowicz-Szczytt K (2015) Genetic and pathogenic diversity of Pseudomonas syringae strains isolated from cucurbits. Eur J Plant Pathol 141:1–14CrossRefGoogle Scholar
  63. Sood SG (2010) Chemotactic response of plant-growth-promoting bacteria towards roots of vesicular-arbuscular mycorrhizal tomato plants. FEMS Microbiol Ecol 45:219–227CrossRefGoogle Scholar
  64. Souza RD, Ambrosini A, Passaglia LMP (2015) Plant growth-promoting bacteria as inoculants in agricultural soils. Genet Mol Biol 38:401–419CrossRefGoogle Scholar
  65. Spaink HP, Kondorosi A, Hooykaas PJJ (1998) The rhizobiaceae: molecular biology of model plant-associated bacteria. Springer, BerlinCrossRefGoogle Scholar
  66. Tanaka H, Negishi H, Maeda H (2009) Control of tobacco bacterial wilt by an avirulent strain of Pseudomonas solanacearum M4S and its bacteriophage. Jpn J Physiol 56:243–246Google Scholar
  67. Toth IK, Moleleki L, Pritchard L, Liu H, Humphris S, Hyman L, Axelsen GW, Brurberg MB, Ravensdale M, Gilroy E (2006) Erwiniae: genomics and the secret life of a plant pathogen. Microbial Ecology of Aerial Plant Surfaces, 191–199 ppGoogle Scholar
  68. Van BP, Sloger C, Weber DF, Cregan PB, Keyser HH (1985) Relationship between Ureide N and N2 fixation, aboveground N accumulation, acetylene reduction, and nodule mass in greenhouse and field studies with Glycine max L. (Merr). Plant Physiol 77:53–58CrossRefGoogle Scholar
  69. Vessey JK (2003) Plant growth promoting rhizobacteria as biofertilizers. Plant Soil 255:571–586CrossRefGoogle Scholar
  70. Vrancken K, Holtappels M, Schoofs H, Deckers T, Valcke R (2013) Pathogenicity and infection strategies of the fire blight pathogen Erwinia amylovora in Rosaceae: state of the art. Microbiology 159:823–832CrossRefGoogle Scholar
  71. Wang X, Liu X, Wang H, Dong Q (2007) Utilization and degradation of imazaquin by a naturally occurring isolate of Arthrobacter crystallopoietes. Chemosphere 67:2156–2162CrossRefGoogle Scholar
  72. Wang H, Chen J, Guo B-Y, Li J (2014) Enantioseletive bioaccumulation and metabolization of diniconazole in earthworms (Eiseniafetida) in an artificial soil. Ecotoxicol Environ Saf 99:98–104CrossRefGoogle Scholar
  73. Wang Z, Yang Y, Dai Y, Xie S (2015) Anaerobic biodegradation of nonylphenol in river sediment under nitrate- or sulfate-reducing conditions and associated bacterial community. J Hazard Mater 286:306–314CrossRefGoogle Scholar
  74. Warnes G, Bolker B, Gorjanc G, Grothendieck G, Korosec A, Lumley T, MacQueen D, Magnusson A, Rogers J (2005) gdata: Various R programming tools for data manipulationGoogle Scholar
  75. White LJ, Ge X, Brozel VS, Subramanian S (2017) Root isoflavonoids and hairy root transformation influence key bacterial taxa in the soybean rhizosphere. Environ Microbiol 19:1391–1406CrossRefGoogle Scholar
  76. Wu H, He X, Dong H, Zhou Q, Zhang Y (2017) Impact of microorganisms, humidity, and temperature on the enantioselective degradation of imazethapyr in two soils. Chirality 29:348–357CrossRefGoogle Scholar
  77. Xie W, Yang F (2018) CYP450 enzyme-specific enantioselective species-specific response for metalaxyl in in vitro hepatic cells. Ecotoxicol Environ Saf 149:10CrossRefGoogle Scholar
  78. Xuedong W, Huili W, Defang F (2005) Biodegradation of imazapyr by free cells of Pseudomonas fluorescene biotype II and Bacillus cereus isolated from soil. Bull Environ Contam Toxicol 74:350–355CrossRefGoogle Scholar
  79. Ye J, Wang L, Zhang Z, Liu W (2013) Enantioselective physiological effects of the herbicide diclofop on cyanobacterium Microcystis aeruginosa. Environ Sci Technol 47:3893–3901CrossRefGoogle Scholar
  80. Zabalza A, Gaston S, Sandalio LM, del Río LA, Royuela M (2007) Oxidative stress is not related to the mode of action of herbicides that inhibit acetolactate synthase. Environ Exp Bot 59:150–159CrossRefGoogle Scholar
  81. Zhang C, Li M, Xu X, Liu N (2015) Effects of carbon nanotubes on atrazine biodegradation by Arthrobacter sp. J Hazard Mater 287:1–6CrossRefGoogle Scholar
  82. Zhang YQ, Markiewicz M, Filser J, Stolte S (2017) Toxicity of a quinaldine-based liquid organic hydrogen carrier (LOHC) system toward soil organisms arthrobacter globiformis and Folsomia candida. Environ Sci Technol 52Google Scholar
  83. Zhou Q, Xu C, Zhang Y, Liu W (2009a) Enantioselectivity in the phytotoxicity of herbicide imazethapyr. J Agric Food Chem 57:1624–1631CrossRefGoogle Scholar
  84. Zhou Y, Li L, Lin K, Zhu X, Liu W (2009b) Enantiomer separation of triazole fungicides by high-performance liquid chromatography. Chirality 21:421–427CrossRefGoogle Scholar
  85. Zhou Q, Zhang N, Zhang C, Huang L, Niu Y, Zhang Y, Liu W (2010) Molecular mechanism of enantioselective inhibition of acetolactate synthase by imazethapyr enantiomers. J Agric Food Chem 58:4202–4206CrossRefGoogle Scholar
  86. Zhou HW, Li DF, Tam NF, Jiang XT, Zhang H, Sheng HF, Qin J, Liu X, Zou F (2011) BIPES, a cost-effective high-throughput method for assessing microbial diversity. ISME J 5:741–749CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Ministry of Education Key Laboratory of Environmental Remediation and Ecosystem Health, College of Natural Resources and Environmental ScienceZhejiang UniversityHangzhouChina
  2. 2.Department of Agricultural SciencesLa Trobe UniversityMelbourneAustralia

Personalised recommendations