Potential environmental risks of genetically modified (GM) crops have raised concerns. To better understand the effect of transgenic rice on the bacterial community in paddy soil, a field experiment was carried out using pairs of rice varieties from two subspecies (indica and japonica) containing bar transgene with herbicide resistance and their parental conventional rice. The 16S rRNA gene of soil genomic DNA from different soil layers at the maturity stage was sequenced using high-throughput sequencing on the Illumina MiSeq platform to explore the microbial community diversity among different rice soils. There were no significant differences in diversity indices between transgenic japonica rice and its sister conventional rice (japonica pair) among different soil layers, but, significant differences was observed between transgenic indica rice and its conventional rice (indica pair) in the topsoil layer around concentrated rice roots according to the ace diversity index. Though the japonica rice soil and indica rice soil were shared several key genera, including Rivibacter, Anaeromyxobacter, Roseomonas, Geobacter, Thiobacillus, Clostridium, and Desulfobulbus, the primary bacterial genera in indica rice soil were different from those in japonica rice. Synechococcus and Dechloromonas were present in japonica rice samples, while Chloronema, Flexibacter, and Blastocatella were observed in indica rice soil. Moreover, the abundance of genera between GM and non-GM varieties in japonica rice was significantly different from indica rice, and several bacterial communities influenced these differences. Anaerovorax was more abundant in transgenic japonica rice soil than conventional rice soil, while it was deficient in transgenic indica rice soil compared to conventional rice soil, and opposite responses to Deferrisoma were in that of indica rice. Thus, we concluded that transgenic indica and japonica rice had different effects on soil bacteria compared with their corresponding sister conventional rice. However, these composition and abundance difference only occurred for a few genera but had no effect on the primary genera and soil characteristics were mainly contributed to these differences. Thus, differences in bacterial community structure can be ignored when evaluating the impacts of transgenic rice in the complex soil microenvironment.
Citation: He M, Zhang J, Shen L, Xu L, Luo W, Li D, et al. (2019) High-throughput sequencing analysis of microbial community diversity in response to indica and japonica bar-transgenic rice paddy soils. PLoS ONE 14(9): e0222191. https://doi.org/10.1371/journal.pone.0222191
Editor: Guangyuan He, Huazhong University of Science and Technology, CHINA
Received: November 26, 2018; Accepted: August 25, 2019; Published: September 9, 2019
Copyright: © 2019 He et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Sequencing data have been deposited to the NCBI Sequence Read Archive under accession number SRP188606 (https://www.ncbi.nlm.nih.gov/sra/?term=SRP188606, Run numbers were from SRR8737695 to SRR8737704, SRR8737725 to SRR8737735, SRR8737750 to SRR8737761, SRR8737765 to SRR8737769, SRR8737780 to SRR8737789, and SRR8737833 to SRR8737839). All other relevant data are within the paper and its Supporting Information files.
Funding: This research was supported by the National Natural Science Foundation of China (31260153, 31560188) (https://isisn.nsfc.gov.cn), the Special Fund for Agro-scientific Research in the Public Interest of China (201403075), and the Excellent Graduate Thesis Program of Hainan University of China (01J1N10001003). The funders are respectively Qianhua Yuan, Xinwu Pei, Qianhua Yuan and Meidan He,and they have played specify the roles.
Competing interests: The authors have declared that no competing interests exist.
Since genetically modified organisms (GMOs) developed and GM crops commercialized in 1996, the global area where biotech crops are planted has sharply increased. Up to 17 million farmers in 24 countries planted 189.8 million hectares of biotech crops in 2017 . GM crops are considered to be the most rapidly adopted crop in modern times . Nevertheless, scientists were worried about the potential risks of GMOs due to diversification of exogenous gene sources and the unpredictability of biological development [3–6]. The potential risks of GMOs have also received scrutiny from consumers, making assessments of GMOs contained in food and feed more important [5–8].
Soil microbial community is an important ecological indicator when accessing the safety of transgenic crops . Due to the interposition of an exogenous gene, GM crops may affect the soil microbial community through their exudates and residues. Numerous, recent studies on interactions between microorganisms and plants had found that microbial diversity of the soil ecosystem is primarily confined to the rhizosphere. To determine the main effects on soil microorganisms from GM crops, researchers had used multiple approaches to conduct comprehensive analyses of transgenic potato producing T4 lysozyme for protection against bacterial infections. Heuer et al. found that environmental factors were the primary influence on the rhizosphere communities, but the effects of T4 lysozyme released from transgenic potato roots did not affect these communities . Bacillus thuringiensis (Bt) crops with insect resistance did not have disadvantageous effects on the composition or activity of the microbial community when planted in field conditions [11–14]. With the ribosomal 16S rRNA gene pyrosequencing, a GM maize, containing three insecticidal proteins in the root tissue, which did not affect the rhizosphere bacterial community compared to conventional varieties . Drought-tolerant CaMSRB2-expressing transgenic rice had a similar contribution to soil microorganisms as the parental rice based on pyrosequencing analysis . Salt-tolerant SUV3 overexpressing transgenic rice was healthy and had equivalent functionality (i.e. the activity of bacterial enzymes and plant growth promotion) to the rhizosphere microorganisms as non-transgenic rice . Beta-carotene transgenic rice with four synthetic carotenoid genes had similar effects on bacterial communities and enzyme activities in the rhizosphere to its parent lines . However, cultivation of drought-tolerant and insect-resistant rice had remarkable effect on bacterial community composition .
Rhizosphere bacterial diversity is influenced by both plant and soil type. Studies on rhizoplane microbial communities from genetically modified crops and control plants have shown that major differences occur among non-transgenic cultivars rather than between non-transgenic and transgenic lines . However, Andreote et al.  observed an increase in Erwinia spp. and a decrease in Agrobacterium spp. associated with the transgenic plants. Zhu et al.  showed that the methanogenic archaeal community in the Bt rice rhizosphere was more active than that in the conventional rice varieties. The inconsistent results of the previous studies could be due to differences in materials, sites, and, approaches and technologies. Soil factors, plant root exudates and agricultural management are the major factors that determine community composition and abundance within the rhizosphere . So far, the evaluation of impacts of GM crops on their soil microenvironments have mostly focused on insect resistance and other functional characteristics, while there have been few assessments related to herbicide resistance. The bialaphos resistance (bar) gene which was isolated from bacteria Streptomyces hygroscopicus confers tolerance to the herbicide glufosinate, has been globally used in basic plant research and genetically engineered crops [24, 25]. Bar gene is also considered to be a useful marker for the selection of transgenic plants . Despite benefits like this, it is essential to investigate the potential environmental risk of transgenic rice before commercial release . Most early studies of soil microorganisms were based on a plate colony counting technique, physiological approaches, and uncultivable molecular assays (including PCR-DGGE, T-RFLP, and pyrosequencing) [16, 18, 28, 29]. Few studies have taken advantage of high-throughput sequencing technology on the Illumina MiSeq platform.
Rice (Oryza sativa L.) has a strategic role as a food source for humans. Rice paddies are important ecosystems. Environmental assessments of GM rice are not only helpful for commercialization of transgenic rice, but also for determining safety of subsequent crops, especially root and tuber crops. Cultivated rice is classified into indica rice (O. sativa L. subsp. indica Kato) and japonica rice (O. sativa L. subsp. japonica Kato), but till now, comparisons of the two subspecies or their GM/non-GM plants effect on soil microorganisms have not been conducted. In addition, rice roots are pooled on the surface during the late growth stage, which differs from the other growth stages. As yet, little research was carried out on soil microorganism community in different soil layers according to the root distribution.
In this study, we focused on soil microbes at the maturity stage. We carried out high-throughput sequencing of 16S rRNA on the Illumina MiSeq platform, to determine differences in the structure and population of rhizosphere soil microbial communities in transgenic and conventional rice paddy soil. We were also interested in identifying any differences in the effects of transgenic indica rice and japonica rice on soil bacteria with their parental conventional rice. Results of this study will provide a theoretical basis for evaluating the impact of genetically modified rice on the soil environment.
Materials and methods
Experimental design and sampling
The experiment was carried out in a paddy field at Hainan University in Haikou, China (20° 03' 27.16" N 110° 19' 5.26" E). Two pairs of stably inherited rice varieties were used in this study: herbicide resistant bar-transgenic japonica rice B2 and its conventional sister line Xiushui63 without herbicide resistance; and bar-transgenic indica rice B68-1 with herbicide resistance and its conventional indica sister line D68 without herbicide resistance. Japonica rice B2 containing bar gene was a sibling of japonica line L201 which was produced by japonica cv. Xiushui63 crossed with a homozygous line that with bar gene transferred into japonica rice cv. JingYin119 [30, 31]. B68-1 was developed by transferring bar gene into indica rice variety D68 . B2 and B68-1 were obtained from the China Rice Research Institute and the Institute of Subtropical Agriculture and Ecology, Chinese Academy of Sciences, respectively. The four rice varieties were arranged in a randomized block design with three replicates each for 4 m × 4 m plots. The plant spacing was 20 cm × 20 cm. According to the rice root distribution in the 20 cm of plowed soil, soils around the roots were divided into four parts to distinguish between the horizontal and vertical distances to the basal part of rice stem: the soil layer in topsoil where horizontal distance was 0–10 cm and vertical distance was 0–10 cm from rice stem, which was considered as topsoil of concentrated roots (TC); the soil layer in topsoil, where horizontal distance was 10–20 cm and vertical distance was 0–10 cm from rice stem, was considered as topsoil of dispersed roots (TD); the soil layer in subsoil where the horizontal distance was 0–10 cm and vertical distance was 10–20 cm from rice stem, was defined as subsoil of concentrated roots (SC); and the soil layer in subsoil where horizontal distance and vertical distance were 10–20 cm from rice stem, which was defined as subsoil of dispersed roots (SD) (S1 Fig, S1 Table). In addition, three replicates of blank soils with no plants were sampled randomly at every five positions in the same experimental field (Table 1). Rhizosphere soils in each position were collected using the S-shaped method with a geotome, excluding the bulk soil without roots, and then, collecting the soil that shaking off from roots in five random positions and mixed together as a representative composite soil sample (about 500 g fresh weight) at each plot. All samples were collected at the rice maturity stage in June 2015.
Each soil sample was placed in individual germfree bags, and immediately stored in a freezer and samples were processed as soon as possible to make sure that the soil microorganisms were closer to the in situ environment. Each sample was split into two unequal parts. One part was about two thirds of each sample for detection of soil nutrients, the other part was about one third for DNA extraction. The former was kept at room temperature, and the latter was kept at -80°C after removing the rice roots.
Physical and chemical properties of soil
There were 18 treated sample groups with a total of 54 independent soil samples. Soil pH was measured in 1:2.5 (W/V) suspensions of soil in distilled water using pH meter (with the electric potential method) and chromic acid oxygen titration was used to measure organic matter (OM). Total nitrogen (TN), ammonium nitrogen (AN), and nitrate nitrogen (NN) content of soil were respectively measured by the semi-micro Macro Kjeldahl method , the indophenol blue spectrophotometric method, and UV spectrophotometry after NaCl extraction (1:5 (W/V) suspensions of soil in 2.0 mol/L NaCl solution). Meanwhile, soil available phosphorus (AP) and total phosphorus content (TP) were measured using Bray I extraction—molybdenum antimony spectrophotometric and acid soluble molybdenum antimony colorimetric assay. Simultaneously, total potassium content (TK) of soil and soil available potassium (AK) were extracted using flame photometric determination.
DNA extraction and polymerase chain reaction
A total of 54 paddy soil samples was centrifuged at 10,000 ×g for 30 secs, and the precipitate was collected as rhizosphere soil and used for DNA extraction; each sample was processed separately. Soil genomic DNA was extracted using the Power Soil DNA Isolation Kit (MoBio Laboratories Inc., USA) following the manufacturer’s instructions. The concentration and quality of the DNA were determined using a spectrophotometer (NanoVue Plus, USA) and the DNA quality and integrity was checked by electrophoresis on a 0.8% agarose golden view gel. The DNA extracted from each soil sample served as a template for amplification of bacterial 16S rRNA gene sequences. A set of primers (forward primer, 338F: 5’-ACTCCTACGGGAGGCAGCA-3’ and reverse primer, 806R: 5’-GGACTACHVGGGTWTCTAAT-3’) [34, 35] were used to amplify the V3-V4 region of the bacterial 16S rRNA gene. After amplification, the PCR product was detected by electrophoresis on a 2% agarose golden view gel.
Illumina MiSeq sequencing of the V3-V4 region of bacterial 16S rRNA genes
The PCR products were extracted from 2% agarose gels and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to the manufacturer’s instructions. DNA was quantified by QuantiFluor™-ST (Promega, USA). Purified products from all samples were homogenized and pooled in equimolar concentrations to construct a PE library. Paired-end sequencing (2×300) with the TruSeq Universal Adapter (5’-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT-3’) and the TruSeq Adapter (5’GATCGGAAGAGCACACGTCTGAACTCCAGTCAC ATCACG ATCTCGTATGCCGTCTTCTGCTTG-3’ (the index was shown in boldface)) was conducted on an Illumina MiSeq platform (Majorbio, Shanghai). Sequencing data had been deposited to the NCBI Sequence Read Archive under accession number SRP188606 (https://www.ncbi.nlm.nih.gov/sra/?term=SRP188606).
Processing and analyzing of sequencing data
After obtaining raw sequencing data, optimized data for bioinformatics analysis was obtained by removing the adapter, barcode, and chimera sequences and correcting the sequence direction. Then, operational taxonomic unit (OTU) sequence analysis was performed. The OTUs with 97% similarity cutoff were clustered using UPARSE (version 7.0)  in USEARCH. Taxonomy information based on OTU can be used to carry out statistical analysis of community structure at each level of classification. Representative OTU sequences were analyzed using a taxonomic database (silva 128/16S_bacteria). Rarefaction analysis was conducted in Mothur v.1.30.1  to calculate alpha diversity, including sobs, chao1, ace, Shannon Wiener, and Simpson’s diversity indices. Beta diversity analysis and phylogenetic analysis were based on UniFrac . Based on the above analysis, a series of in-depth statistical and visual analyses of community structure and phylogeny were carried out. Principal co-ordinates analysis (PCoA) were carried out using the community ecology package, the Vegan 2.0 package was used to generate a PCoA figure, and Venn diagrams were generated by Venn Diagram, while Kernel density estimation, redundancy analysis (RDA), and the heat-map figures were generated in Vegan 2.0 in R .
Physical and chemical characteristics of different soil samples
Soil pH value ranged from 6.71 to 7.09 in this study (Table 2). In TC and SD soil layers, bar-japonica rice was similar to conventional rice, this was also the case for the indica pair. At the TD and SC soil layers, bar-japonica rice was also similar to conventional rice, however, the pH of bar-indica soil was significantly different from that of conventional rice. There were clear, significant differences between japonica rice and indica rice soils at each soil layers. Organic matter content ranged from 8.92 g/Kg to 17.98 g/Kg, and the OM, TP, and AP contents were not significantly different among all the compared groups. Total nitrogen at the TC soil layers only showed difference in that of the indica rice pair. The TK content of transgenic rice soils were almost the same as conventional samples. TK content of the bar-indica rice soil was significantly different from that of conventional rice at the TC soil layer, while the AK content showed some significant differences between the transgenic rice soils and their conventional pairs. For example, the AK content of bar-indica rice soil was significantly lower than that of conventional rice soil at the TC soil layer, while it was significantly higher than conventional samples at the SD soil layer. Interestingly, AN and NN content had a similar pattern: their content in bar-japonica rice soils was significantly lower than conventional rice soil at the TC and TD soil layers, while they were significantly higher than conventional rice soil at the SC and SD soil layers. Moreover, the AN content of bar-indica rice soil was significantly lower than in conventional rice soil at the SD soil layers.
Taxonomy and alpha diversity analysis of sequenced data
A total of 2,062,254 raw sequences of 16S rRNA were generated from 54 soil samples after Illumina MiSeq sequencing. There were 1,630,398 high quality sequences following quality control (80% retention), and each sample had around 30,192 sequences for later analysis. The average sequence length was 437 bp (S2 Table). Based on the minimum sample sequence, the rhizosphere bacterial community contained 59 phyla, 141 classes, 266 orders, 497 families, 937 genera, 1969 species, and 5384 OTUs. The most abundant phyla in all samples were 14 unnamed phyla, Proteobacteria (30%), and Chloroflexi (20%).
Each sample contained an average of 2,719 OTUs, according to the Shannon curve analysis. With an increase in the number of sequenced bands, samples of microbial diversity plateaued and increasing the sequencing reads did not significantly increase the diversity of microorganisms (S2 Fig). Diversity indices (S3 Table) were typically not significantly different at the same soil layer, except at the TC soil layer there were significant differences of ace index between bar-transgenic indica rice and its parental conventional rice and shannon index between bar-transgenic japonica rice and its parental conventional rice (Table 3).
Soil microbe distribution between indica rice and japonica rice
Various bacteria were obtained at OTU taxonomic levels. Relative abundance of bacterial communities up to 0.5% was defined as a dominant genus. Bacillus was a primary bacterium in paddy soil. The dominant genus Leptolyngbya, Roseomonas, and Spirochaeta were observed among the blank soils, indica rice soils, and japonica rice soils. Besides, blank soil also contained dominant bacterial genera Blastocatella, Bryobacter, Sphingomonas, Thiobacillus, Marmoricola, Phormidium, Roseiflexus, Lyngbya, Nitrospira, and Flexibacter.
Though the japonica rice soil and indica rice soil were shared several key genera, including Rivibacter, Anaeromyxobacter, Roseomonas, Geobacter, Thiobacillus, Clostridium, and Desulfobulbus, however, dominant genera, Synechococcus and Dechloromonas were present in japonica rice samples, while Chloronema, Flexibacter, and Blastocatella were observed in indica rice soil. Dominant bacterial genera were shown in (Fig 1A, 1B and 1C).
(a), the dominant bacterial genus in japonica rice soil. (b), the dominant bacterial genus in indica rice soil. (c), the dominant bacterial genus in blank rice soil. (d), the shared and unique OTU for blank soil, japonica rice soil, and indica rice soil.
Further analysis based on common microbial taxa showed that three kinds of paddy soil samples contained 4,296 OTUs, up to 71.70% of all OTUs. These shared OTUs mainly belonged to Bacillus, Clostridium, Sphingomonas, Flexibacter, Blastocatella, Thiobacillus, Marmoricola, Leptolyngbya, and Anaeromyxobacter. Almost all OTUs (98.18%) were shared by indica and japonica rice, indicating minimal differences in the bacteria composition (Fig 1D, S3 Fig). About 13 core microorganisms including Rivibacter, Synechococcus, and Bacillus, were observed among the soil samples where rice was planted.
Correlations of these primary microorganisms were calculated based on Spearman rank correlation. Positive correlations were found between Bacillus and Geobacter; Rivibacter and Synechococcus; and Leptolyngbya with Chloronema, Blastocatella and Flexibacter. Leptolyngbya was negatively correlated with Roseomonas and Anaeromyxobacter. Flexibacter was positively correlated with most core bacteria (Fig 2). Positive correlations indicated growth promotion (synergistic) and negative correlations indicated inhibition (antagonistic).
Comparison of bacterial community structure among groups
Samples were divided into different groups based on cultivar, genetic modification status, and difference of soil depth. Principal coordinate analysis (PCoA) was used to identify the community structure differences among different groups. Cluster analysis and the PCoA maps were constructed based on weighted UniFrac (Fig 3) and unweighted UniFrac (S4 Fig).
(a), principal coordinate analysis of group samples categorized by genetic modification status and rice subspecies. (b), principal coordinate analysis for different groups of rice subspecies. (c), principal coordinate analysis of the GM group samples and non-GM group samples. (d), principal coordinate analysis of topsoil samples and subsoil samples.BS, blank soil; IT, transgenic indica rice, IC, indica rice control; JT, transgenic japonica rice and JC for japonica rice control; JR, japonica rice, IR, indica rice; TR, transgenic rice, CR, conventional rice control; D, subsoil soil; S, topsoil.
When all tested samples were divided into five groups according to the cultivar and genetic modification status, the japonica pair were clearly distinct from each other, but indica pair were in a cluster (Fig 3A). While samples were separated into blank, japonica, and indica groups, the japonica and indica groups did not form two distinct clusters based on weighted UniFrac of PCoA analysis (Fig 3B). Similarly, transgenic and conventional groups were not separated (Fig 3C) and blank samples were clustered together but separated from the other groups (Fig 3A, 3B and 3C). Two groups at different soil depths were not distinct, indicating no difference between the two tested groups. Differences in bacterial community structure were only shown in japonica pair (bar-transgenic and its conventional rice soil samples); the other groups were similar.
Dominant factors affecting the distribution of plant root microorganisms
As mentioned above, rice lines and transgenes might affect the distribution of plant root microorganisms. Indica rice and japonica rice were differentiated and developed over long-term evolution in different climates and ecological environments, while transgenic rice was artificially created by inserting a gene fragment that could be stably inherited by offspring. Evolutionary distance of microorganisms was compared to find the main factor differentiating the richness of soil microorganisms. The evolutionary distance of samples from transgenic and non-transgenic groups was farther than that between subspecies rice varieties (Fig 4). This result may be attributed to foreign genes extracted from unrelated organisms.
Analysis of root microorganisms between transgenic and non-transgenic rice
Since transgenes were the dominant factor affecting root microorganisms, it was important to determine where the differences originated. The indica rice pair and the japonica rice pair (GM and non-GM) were analyzed using Welch’s t-test (unequal variances). The result showed that there were around 92 different bacterial genera for the comparison of transgenic japonica rice and non-transgenic japonica rice, and 55 different genera for the comparison of transgenic indica rice and its non-transgenic indica rice. Differences in japonica rice were more obvious than for indica rice. Rivibacter, Synechococcus, Anaerolinea, Dechloromonas, Thiobacillus, and Roseomonas were the dominant bacterial genera in the japonica group, meanwhile, Arthronema, Roseiflexus, Blastocatella, Flexibacter, and Roseomonas were the primary bacterial genera in the indica pair. The japonica pair and the indica pair had 19 shared bacterial genera among the significantly different microorganisms. The relative abundance of bacterial microflora was different. Certain bacterial genera (e.g. Synechococcus, Flexibacter, Alkaliflexus, Leptolinea, Piscinibacter, Sandaracinus, and Cryptanaerobacter) were more abundant in transgenic rice soil samples than conventional rice soil samples. However, Roseomonas, Nitrosomonas, Desulforhabdus, Arthronema, Desulfuromonas, Geoalkalibacter, Mesorhizobium, Pedomicrobium, Sporacetigenium, and Thiobacillus were deficient in transgenic rice soil samples compared with conventional rice samples. Of the common bacterial genera, Anaerovorax was more abundant in transgenic japonica rice soil than soil from conventional rice, while it was deficient in transgenic indica rice soil compared to soil with conventional rice and opposite responses to Deferrisoma in the indica rice pair were observed (Fig 5).
Soil microorganisms from transgenic root systems influenced neighboring soil
To determine the effects of transgenes on neighboring soil environments, microbial community structure in the surrounding soil was analyzed based on evolutionary distance. Both transgenic japonica rice and indica rice had a similar effect on their surrounding soil, in that microbial community structure in the surrounding soils was similar to that of the neighboring plants, but significantly different from the control blank soil (Fig 6). The results showed that the rhizosphere microorganisms of transgenic rice could affect the distribution of microorganisms in the surrounding soil.
A total of 25 bacterial communities had significantly different relative abundance at the genus level (Table 4). For example, Leptolyngbya, Roseomonas, and Tropicimonas from the transgenic rice rhizosphere soil were respectively about tenfold, fivefold, and three-fold less abundant than in blank soil. In contrast, Dechloromonas, Fastidiosipila, and Cryptanaerobacter from transgenic rice soil were more abundant than in blank soil. There were more significantly different bacterial genera between transgenic japonica rice soil and blank soil than that between indica rice soil and blank soil. These results showed clear effects of microorganisms on soil adjacent to the root system of the transgenic plant.
Correlation analysis between root microbes and soil physical and chemical indices
Further analysis of correlations between environmental factors and the top 50 most abundant bacterial communities at the genus level were based on the Spearman rank correlation coefficient. Results revealed correlations between soil physical indices and bacterial communities in rice root systems (Fig 7). About 16 bacterial genera were significantly correlated with soil pH, six with negative correlations and ten with positive correlations (Table 5). Six bacterial genera were significantly correlated with soil organic matter (OM) concentration, five with positive correlations and one with negative correlations. Meanwhile, some genera were significantly correlated with the concentration of available nutrients. The concentrations of available potassium (AK), nitrate nitrogen (NN), and ammonium nitrogen (AN) were significantly correlated to 13, 13, and 8 genera, respectively. No genus was significantly correlated with available phosphorus (AP). Total potassium (TK) concentration was significantly correlated with 12 genera, but only four genera had positive correlations. The total nitrogen (TN) concentration and the total phosphorus (TP) concentration were significantly related to six and four genera, respectively. In addition, the bacterial genus Arthronema was significantly negatively correlated to most environmental factors, while Bacillus and Anaeromyxobacter were positively correlated to most environmental factors in this study. The correlation between other genera and environmental factors were varied. In short, environmental factors affected the bacterial community and created differences in their abundance.
AK, available potassium; AP, available phosphorus; AN, ammonium nitrogen; NN, nitrate nitrogen; OM, organic matter; TK, total potassium; TP, total phosphorus; TN, total nitrogen. The red colors represented positive correlations, while blue colors represented negative correlations. Darker colors represent stronger correlations. Significant differences were represented by: *0.01<p = <0.05, **0.001<p = <0.01, ***p<0.001.
Biological diversity including composition, structure, and function , which was rapidly lost with the development of agriculture. Taking advantage of biodiversity can benefit agricultural production . As a large number of GM crops come out, the effect from cultivation of GM crops to biological diversity had raised concern. However, scientists were considered that GM crops seriously reduce biodiversity and damage the environment . Whether GM crops will affect the soil microbial community, it has become an issue of public concern .
In response to this concern, scientists had carried out related research. Insect resistant Bt maize, for example, which was the most important transgenic crop did not change the microbial populations of the soil or in the activity of the microbial community , the rhizosphere bacterial community  and the soil microarthropod communities . Similarly, Bt cotton cultivation did not affect microbial communities in the rhizosphere soil  and the community characteristics of rhizosphere soil nematodes . And Cry1Ac transgenic sugarcane would not change the diversity of microbial communities or modify enzyme activities in rhizosphere soil within one crop season . Bt rice also had no persistent or adverse effect on the microbial community composition in its rhizosphere [12, 29]. However, planting Bt rice had the potential to modify bacterial abundances and community structures , the methanogenic community composition  in their rhizosphere and a strong influence on active methanogens . And for other functional traits of genetically modified crops, transgenic potato roots with T4 lysozyme released had no deviation in the rhizosphere communities compared to the control lines . T8-ipt (isopentenyl transferase) rice cannot affect the soil bacteria biomass during growth of plants in the field . Beta-carotene transgenic rice with four synthetic genes had no specific effect on rhizosphere enzyme activities and bacterial communities at different growth stages . Salt tolerant SUV3 rice sustained a healthy ecology and usual functions of the rhizospheric organisms . Transgenic Brassica napus containing the antifungal synthetic chitinase (NiC) may not affect the enzyme activities and community structure of microbes in rhizosphere soil . GM eggplants had a short-term impact on soil quality and microbial diversity and the differences disappeared post-harvest . The cultivation of transgenic tobacco can affect rhizosphere/ rhizoplane microbial communities during early plant development, but the original bacterial diversity would restore after one cycle of plant cultivation , while transgenic tobacco plants that express and release extracellular microbial phytases from their roots, could not change the microbial community in the rhizosphere . Altogether, most Bt and other transgenic crops had no/little effect on the soil microbial community.
Herbicide resistance was a dominant trait of GM crops, and it was widely used transgene research. In this way, it was more important to evaluate the effect of GM crops with herbicide resistance on the soil microbial community. EPSPS-transgenic soybean unleashed temporary effects on the taxonomic diversity of rhizosphere bacterial communities at the vegetative and seed-filling stages compared to its recipient cultivar under field conditions, and main symbiotic nitrogen-fixing bacterial genera in the roots evidently changed from the flowering stage to the seed-filling stage . No effect of the bacterial community composition in the rhizosphere was detected between herbicide-resistant maize with the pat-gene and its non-transgenic cultivar . Using a conventional culture technique and culture-independent molecular methods to assess soil microbial community in the rhizosphere soil of herbicide resistant genetically modified Chinese cabbage, the bacterial, fungal and actinomycetes population densities, DGGE banding patterns and bacterial species diversity indices were all similar to those of the non-GM Chinese cabbage soils  with the result of the total counts of bacteria, fungi, and actinomycetes, dominant members and PLFA (phospholipid fatty acid) profiles in the soil microbial community, Sohn. et al. believed that the herbicide-resistant bar-transgenic perilla had insignificant impact on the soil microbial communities when compared with those of conventional perillas . However, despite these findings, there have been few studies of whether the bar gene or transgenic crops with herbicide resistance affected soil microorganisms. Transgenic rice that highly resistant to protoporphyrin oxidase (PPO)-inhibiting herbicides, was used to evaluate the structure of soil microbial communities, and showed no adverse effects in paddy field . Similar results for cultivating transgenic rice MSRB2-Bar-8 were reported by Sohn et al. . These two researches only paid attention to the effects on transgenic japonica rice and their conventional rice varieties, in this study, comparison of two ecological subspecies rice (indica rice and japonica rice) and different soil layers during maturing period were both concern to affect the soil bacterial community.
Plants and soil have a variety of interactions, due to complicated environmental factors, including climate and soil characteristics. In this research, the organic matter, total phosphorus, and available phosphorus content were not significantly different among treated groups, and soil pH values were almost the same, it was agreed with the results of salt tolerant SUV3 overexpressing transgenic rice  and CaMSRB2-expressing transgenic rice , except for difference in indica pair at the TD and SC soil layer. Total nitrogen and total potassium were in a similar way and showed difference at the TC soil layer. Soil available potassium was significant different between bar-indica rice and its conventional rice at TC and SD. Ammonium nitrogen and nitrate nitrogen were quite different between bar-japonica rice and its conventional rice at each treated groups. And the Ammonium nitrogen content of bar-indica rice soil was significantly lower than in conventional rice soil at the SD soil layer. The overall analysis of these chemical characteristics showed some significant differences between the soils of GM and non-GM rice. Moreover, soil pH was clear, significant differences between japonica rice and indica rice soils at each sampling zone. It is well known that pH plays an important role on soil microbial composition and biogeochemical function [55, 56]. For ammonium nitrogen and nitrate nitrogen, significant differences were observed between japonica rice and indica rice soils. It may be the reason that indica and japonica rice plants responded differently to NO3-, but similarly to NH4+ . Some significant differences of chemical characteristics between japonica rice and indica rice soils were also observed. Nitrogen-use efficiency of indica varieties of rice was better than that of japonica varieties. NRT1.1B, a rice nitrate transporter and sensor, which is associated with the recruitment of a large proportion of indica-enriched bacteria, contributes to the variation in the root microbiota of indica and japonica .
Through the Kernel density estimation, we observed that transgenic varieties and non-transgenic varieties had larger evolutionary distances than subspecies rice varieties, suggesting that the effect from transgenes was much stronger than the effect of subspecies. This result may be attributed to foreign genes extracted from unrelated organisms. Similarly, transgenic rice containing cry1Ab had potential risks of modification of the methanogenic community composition in its rhizosphere . Certain bacterial genera (e.g. Synechococcus, Flexibacter, Alkaliflexus, and Cryptanaerobacter) were abundant in transgenic rice but deficient in soil around conventional rice. In contrast, Roseomonas, Nitrosomonas, Desulforhabdus, and Sporacetigenium were deficient in transgenic rice soil samples compared with conventional rice samples. This phenomenon was clear with japonica rice. Thus, several bacterial communities strongly influence the differences, which requires further study of the impact of GM crops. The soil microenvironment was extremely complex, meaning that some small but significant differences were most likely unnoticed.
In this study, Bacillus was positively correlated with Geobacter, and species of Bacillus and Geobacter were known to reduce or sorb hexavalent chromium . Arthronema was significantly negatively correlated to most environmental factors, while Bacillus and Anaeromyxobacter were positively correlated to most environmental factors (Fig 7). Flexibacter is a Gram-negative bacterium that is usually strictly aerobic and F. canadensis isolated from soil environments denitrifies NO3- and NO2- to gaseous forms with tolerance to oxygen ; Flexibacter was not significantly correlated with the nitrate nitrogen (NN) concentration in this study. The activity of microorganisms from the rice rhizosphere varied in different bacterial genera depending on environmental stress . George et al.  confirmed that soil microorganisms were involved in controlling the phosphorus available to plants and that the microbial community in the rhizosphere appeared to be resistant to the impacts of inserting single genes in plants to change rhizosphere biochemistry and nutrient cycling. However, we observed that there was no significant correlation between AP and the top 50 bacterial genera. In a previous study, bacterial abundance was positively associated with soil pH , and pH differences were the main factor influencing soil archaeal diversity and community structure in the tropical zone and chemoautotrophic carbon dioxide fixation in drained paddy soils [62, 63]. Environmental factors clearly affected the bacterial community in this study and were the main reason for their differences in abundance. The results indicated that the differences found in the soil microbial structure between GM and non-GM rice soil, which will not be a result of the cultivation of the GM rice, but soil characteristics.
Microbial community of rice root system influences ecosystem functioning is the key to improving crop health and sustainable productivity of paddy ecosystems, and reducing methane emissions . Most studies on the rhizosphere have paid more attention to the number and diversity of bacterial taxa in independent rhizospheres rather than effects from one rhizosphere to others . Our study focused on rhizosphere and adjacent soil. Further research could focus on rhizosphere competition between transgenic rice and conventional rice to identify key growth stages in different areas. In this study, several bacterial communities were found to influence differences, further research is needed to determine why these differences occur.
This study represented the first ecological risk assessment of the potential effects of herbicide resistance rice cultivation between two rice subspecies with different soil layers on rhizosphere microbial communities. There was no significant difference on the diversity indices between transgenic and non-transgenic rice soils, but the primary bacterial genera in indica rice soil were different from those in japonica rice. Moreover, the abundances of genera between GM and non-GM varieties in japonica rice were significantly different from indica rice, and several bacterial communities influenced these differences. Anaerovorax was more abundant in transgenic japonica rice soil than conventional soil, while it was deficient in transgenic indica rice soil compared to conventional rice, and opposite responses to Deferrisoma were in that of indica rice. Thus, we concluded that transgenic indica and japonica rice had different effects on soil bacteria compared with their corresponding sister conventional rice. However, these effects only occurred for a few genera but had no effect on the primary genera and soil characteristics were the main factor driving these differences. Therefore, differences in bacterial structure can be ignored when evaluating the impacts of transgenic rice on the complex soil microenvironment.
S2 Table. Sequence description of all samples.
S1 Fig. The axonometric drawing and profile map for sampling in the field.
(a), axonometric drawing for soil sampling in the rice field. (b), profile map for soil sampling in the rice field. (c), profile map for sampling blank soil. TC was topsoil near rice basal part of rice stem while TD was topsoil far from rice basal part of rice stem; SC was subsoil near rice basal part of rice stem and SD was subsoil far from rice basal part of rice stem; BTS was blank topsoil and BSS was blank subsoil, neither of which have rice.
S2 Fig. The rarefaction curve and the shannon index curve for all sequenced samples.
(a), the rarefaction curve for the 54 sequenced samples. (b), the shannon index curve for the 54 sequenced samples.
S3 Fig. Venn diagrams.
(a), the shared and unique OTU among indica, japonica, and blank soil. (b), the shared and unique species among indica, japonica, and blank soil. (c), the common and unique OTU among conventional, transgenic, and blank soil. (d), the common and unique species among conventional, transgenic, and blank soil. (e), the identical and unique OTU among blank topsoil, crop topsoil, blank subsoil, and crop subsoil. (f), the identical and unique species among blank topsoil, crop topsoil, blank subsoil, and crop subsoil.
S4 Fig. Principal coordinate analysis of different groups based on unweighted UniFrac distance.
(a), principal coordinate analysis of group samples categorized by genetic modification status and rice subspecies. (b), principal coordinate analysis for different groups of rice subspecies. (c), principal coordinate analysis of the GM group samples and non-GM group samples. (d), principal coordinate analysis of topsoil samples and subsoil samples.
We would like to thank Dr. Qian Qian in the National Rice Research Institute and Dr. Guoying Xiao in Subtropical Agriculture and Ecology, Chinese Academy of Sciences for providing us B2 and B68-1 seeds, and Dr. Jiemeng Xu from International Rice Research Institute (IRRI) for giving some constructive comments for this manuscript.
- 1. James C. Global Status of Commercialized Biotech/GM Crops in 2017:Biotech Crop Adoption Surges as Economic Benefits Accumulate in 22 Years: ISAAA; 2018 [cited 2018 Sep 22]. Available from: http://www.isaaa.org/resources/publications/annualreport/2017/default.asp.
- 2. Brookes G, Barfoot P. Economic impact of GM crops: The global income and production effects 1996–2012. Gm Crops & Food. 2014;5(1):65–75. pmid:24637520
- 3. Lu BR, Yang X, Ellstrand NC. Fitness correlates of crop transgene flow into weedy populations: a case study of weedy rice in China and other examples. Evol Appl. 2016;9(7):857–70. pmid:27468304
- 4. Mao J, Sun X, Cheng JH, Shi YJ, Wang XZ, Qin JJ, et al. A 52-week safety study in cynomolgus macaques for genetically modified rice expressing Cry1Ab/1Ac protein. Food Chem Toxicol. 2016;95:1–11. pmid:27338709
- 5. Domingo JL. Safety assessment of GM plants: An updated review of the scientific literature. Food Chem Toxicol. 2016;95:12–8. pmid:27317828
- 6. Tothova T, Sobekova A, Holovska K, Legath J, Pristas P, Javorsky P. Natural glufosinate resistance of soil microorganisms and GMO safety. Cent Eur J Biol. 2010;5(5):656–63.
- 7. Katie Eastham JS. Genetically modified organisms (GMOs): The significance of gene flow through pollen transfer. Environmental issue report: the European Science Foundation and the European Environment Agency, 2002 28.
- 8. Tsatsakis AM, Nawaz MA, Tutelyan VA, Golokhvast KS, Kalantzi O-I, Chung DH, et al. Impact on environment, ecosystem, diversity and health from culturing and using GMOs as feed and food. Food Chem Toxicol. 2017. pmid:28645870
- 9. Chun YJ, Kim HJ, Park KW, Jeong SC, Lee B, Back K, et al. Two-year field study shows little evidence that PPO-transgenic rice affects the structure of soil microbial communities. Biol Fertil Soils. 2012;48(4):453–61.
- 10. Heuer H, Kroppenstedt RM, Lottmann J, Berg G, Smalla K. Effects of T4 lysozyme release from transgenic potato roots on bacterial rhizosphere communities are negligible relative to natural factors. Appl Environ Microbiol. 2002;68(3):1325–35. pmid:11872484
- 11. Oliveira AP, Pampulha ME, Bennett JP. A two-year field study with transgenic Bacillus thuringiensis maize: effects on soil microorganisms. Sci Total Environ. 2008;405(1–3):351–7. Epub 2008/07/29. pmid:18656246
- 12. Wu WX, Liu W, Lu HH, Chen YX, Medha D, Janice T. Use of 13C labeling to assess carbon partitioning in transgenic and nontransgenic (parental) rice and their rhizosphere soil microbial communities. FEMS Microbiol Ecol. 2009;67(1):93–102. Epub 2008/12/04. pmid:19049503
- 13. Yang B, Chen H, Liu X, Ge F, Chen Q. Bt cotton planting does not affect the community characteristics of rhizosphere soil nematodes. Applied Soil Ecology. 2014;73:156–64.
- 14. Zhang YJ, Xie M, Wu G, Peng DL, Yu WB. A 3-year field investigation of impacts of Monsanto’s transgenic Bt-cotton NC 33B on rhizosphere microbial communities in northern China. Applied Soil Ecology. 2015;89:18–24.
- 15. Dohrmann AB, Küting M, Jünemann S, Jaenicke S, Schlüter A, Tebbe CC. Importance of rare taxa for bacterial diversity in the rhizosphere of Bt- and conventional maize varieties. Isme Journal. 2013;7(1):37. pmid:22791236
- 16. Sohn SI, Oh YJ, Kim BY, Cho HS. Effects of CaMSRB2-Expressing Transgenic Rice Cultivation on Soil Microbial Communities. J Microbiol Biotechnol. 2016;26(7):1303–10. Epub 2016/04/20. pmid:27090184
- 17. Sahoo RK, Ansari MW, Tuteja R, Tuteja N. Salt tolerant SUV3 overexpressing transgenic rice plants conserve physicochemical properties and microbial communities of rhizosphere. Chemosphere. 2015;119:1040–7. pmid:25303666
- 18. Li P, Dong J, Yang S, Bai L, Wang J, Wu G, et al. Impact of β-carotene transgenic rice with four synthetic genes on rhizosphere enzyme activities and bacterial communities at different growth stages. Eur J Soil Biol. 2014;65:40–6.
- 19. Li P, Ye S, Liu H, Pan A, Ming F, Tang X. Cultivation of Drought-Tolerant and Insect-Resistant Rice Affects Soil Bacterial, but Not Fungal, Abundances and Community Structures. Front Microbiol [Internet]. 2018; 9:[1390 p.]. pmid:30008701
- 20. Dunfield KE, Germida JJ. Impact of genetically modified crops on soil- and plant-associated microbial communities. Journal of environmental quality. 2004;33(3):806–15. pmid:15224914
- 21. Andreote FD, Mendes R, Dini-Andreote F, Rossetto PB, Labate CA, Pizzirani-Kleiner AA, et al. Transgenic tobacco revealing altered bacterial diversity in the rhizosphere during early plant development. Antonie Van Leeuwenhoek. 2008;93(4):415–24. pmid:18181027
- 22. Zhu W, Lu H, Hill J, Guo X, Wang H, Wu W. (1)(3)C pulse-chase labeling comparative assessment of the active methanogenic archaeal community composition in the transgenic and nontransgenic parental rice rhizospheres. FEMS Microbiol Ecol. 2014;87(3):746–56. Epub 2013/11/26. pmid:24266498
- 23. Doornbos R, van Loon L, Bakker PHM. Impact of root exudates and plant defense signaling on bacterial communities in the rhizosphere. A review. Agronomy for Sustainable Development. 2012;32(1):227–43.
- 24. Duke SO. Taking stock of herbicide-resistant crops ten years after introduction. Pest Manage Sci. 2010;61(3):211–8. pmid:15660452
- 25. Christ B, Weng JK, Guyer L, Hochstrasser R, Francisco R, Hörtensteiner S, et al. Non-specific activities of the major herbicide-resistance gene BAR. Nature Plants. 2017;3(12). pmid:29180815
- 26. White J, Chang SY, Bibb MJ, Bibb MJ. A cassette containing the bar gene of Streptomyces hygroscopicus: a selectable marker for plant transformation. Nucleic Acids Res. 1990;18(4):1062. pmid:2315036
- 27. Wu JR, Yu MZ, Xu JH, Du J, Ji F, Dong F, et al. Impact of Transgenic Wheat with wheat yellow mosaic virus Resistance on Microbial Community Diversity and Enzyme Activity in Rhizosphere Soil. Plos One. 2014;9(6). pmid:24897124
- 28. Liu W, Hao Lu H, Wu W, Kun Wei Q, Xu Chen Y, Thies JE. Transgenic Bt rice does not affect enzyme activities and microbial composition in the rhizosphere during crop development. Soil Biol Biochem. 2008;40(2):475–86.
- 29. Song YN, Su J, Chen R, Lin Y, Wang F. Diversity of Microbial Community in a Paddy Soil with cry1Ac/cpti Transgenic Rice. Pedosphere. 2014;24(3):349–58.
- 30. Jia S, Wang F, Shi L, Yuan Q, Liu W, Liao Y, et al. Transgene flow to hybrid rice and its male-sterile lines. Transgenic Res. 2007;16(4):491–501. pmid:17443417
- 31. Yuan QH, Shi L, Wang F, Cao B, Qian Q, Lei XM, et al. Investigation of rice transgene flow in compass sectors by using male sterile line as a pollen detector. Theoretical & Applied Genetics. 2007;115(4):549–60. pmid:17622509
- 32. Xiao G, Yuan L, Sun SSM. Strategy and utilization of a herbicide resistance gene in two-line hybrid rice. Mol Breed. 2007;20(3):287–92.
- 33. Edwards A. H. The semi-micro Kjeldahl method for the determination of nitrogen in coal. Journal of Chemical Technology & Biotechnology Biotechnology. 2010;4(6):330–40.
- 34. Dennis KL, Wang Y, Blatner NR, Wang S, Saadalla A, Trudeau E, et al. Adenomatous Polyps Are Driven by Microbe-Instigated Focal Inflammation and Are Controlled by IL-10–Producing T Cells. Cancer Res. 2013;73(19):5905–13. pmid:23955389
- 35. Xu N, Tan G, Wang H, Gai X. Effect of biochar additions to soil on nitrogen leaching, microbial biomass and bacterial community structure. Eur J Soil Biol. 2016;74:1–8.
- 36. Edgar RC. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods. 2013;10(10):996–8. Epub 2013/08/21. pmid:23955772
- 37. Schloss PD, Gevers D, Westcott SL. Reducing the Effects of PCR Amplification and Sequencing Artifacts on 16S rRNA-Based Studies. PLoS One. 2011;6(12):e27310. pmid:22194782
- 38. Lozupone C, Knight R. UniFrac: a New Phylogenetic Method for Comparing Microbial Communities. Appl Environ Microbiol.2006;71(12): 8228–35.
- 39. Hong C, Si Y, Xing Y, Li Y. Illumina MiSeq sequencing investigation on the contrasting soil bacterial community structures in different iron mining areas. Environ Sci Pollut Res Int. 2015;22(14):10788–99. pmid:25761991
- 40. Clergue B, Amiaud B, Pervanchon F, Lasserrejoulin F, Plantureux S. Biodiversity: Function and Assessment in Agricultural Areas: A Review. Agronomie. 2005;25(1):6119–25.
- 41. Gurr GM, Wratten SD, Luna JM. Multi-function agricultural biodiversity: pest management and other benefits. Basic Appl Ecol. 2003;4(2):107–16.
- 42. Angle JS. Release of transgenic plants: biodiversity and population-level considerations. Mol Ecol. 1994;3(1):45–50.
- 43. Dunfield KE, Germida JJ. Impact of genetically modified crops on soil- and plant-associated microbial communities. J Environ Qual. 2004;33(3):806–15. Epub 2004/07/01. pmid:15224914
- 44. Arias-Martín M, García M, Luciáñez MJ, Ortego F, Castañera P, Farinós GP. Effects of three-year cultivation of Cry1Ab-expressing Bt maize on soil microarthropod communities. Agric, Ecosyst Environ. 2016;220:125–34.
- 45. Zhou D, Xu L, Gao S, Guo J, Luo J, You Q, et al. Cry1Ac Transgenic Sugarcane Does Not Affect the Diversity of Microbial Communities and Has No Significant Effect on Enzyme Activities in Rhizosphere Soil within One Crop Season. Frontiers in plant science. 2016;7(265). pmid:27014291
- 46. Han C, Liu B, Zhong W. Effects of transgenic Bt rice on the active rhizospheric methanogenic archaeal community as revealed by DNA-based stable isotope probing. J Appl Microbiol. 2018. Epub 2018/05/31. pmid:29846995
- 47. Song L, Zhao DG, Jin DC. Impact of T8 Transgenic Rice Containing an Isopentenyl Transferase Gene on Soil Bacterial Biomass. Advanced Materials Research. 2012;455–456:1404–9.
- 48. Khan MS, Sadat SU, Jan A, Munir I. Impact of TransgenicBrassica napusHarboring the Antifungal Synthetic Chitinase (NiC) Gene on Rhizosphere Microbial Diversity and Enzyme Activities. Frontiers in plant science. 2017;8:1307. pmid:28791039
- 49. Mocali S, Dentice A, Marcucci A, Benedetti A. The impact of post-harvest treatments of transgenic eggplant residues on soil quality and microbial diversity. Ultrasound Med Biol. 2009;53(5):296–307.
- 50. George TS, Richardson AE, Li SS, Gregory PJ, Daniell TJ. Extracellular release of a heterologous phytase from roots of transgenic plants: does manipulation of rhizosphere biochemistry impact microbial community structure? FEMS Microbiol Ecol. 2009;70(3):433–45. Epub 2009/09/12. pmid:19744239
- 51. Lu GH, Tang CY, Hua XM, Cheng J, Wang GH, Zhu YL, et al. Effects of an EPSPS-transgenic soybean line ZUTS31 on root-associated bacterial communities during field growth. PLoS One. 2018;13(2):25. pmid:29408918
- 52. Schmalenberger A, Tebbe CC. Bacterial community composition in the rhizosphere of a transgenic, herbicide-resistant maize (Zea mays) and comparison to its non-transgenic cultivar Bosphore. FEMS Microbiol Ecol. 2002;40(1):29–37. pmid:19709208
- 53. Sohn SI, Oh YJ, Ahn BO, Ryu TH, Cho HS, Park JS, et al. Soil Microbial Community Assessment for the Rhizosphere Soil of Herbicide Resistant Genetically Modified Chinese Cabbage. Korean Journal of Environmental Agriculture. 2012;31(1):52–9.
- 54. Sohn SI, Kwon JS, Woen HY, Noh HJ, Kim KH, Baek HJ, et al. Assessment of Microbial Community in the Rhizoplane and Rhizosphere Soil of Herbicide-Resistant Transgenic Perilla. Journal of the Korean Society of International Agriculture. 2009.
- 55. Rousk J, Brookes PC, Bååth E. The microbial PLFA composition as affected by pH in an arable soil. Soil Biol Biochem. 2010;42(3):516–20.
- 56. Noah F, Jackson RB. The diversity and biogeography of soil bacterial communities. Proc Natl Acad Sci U S A. 2006;103(3):626–31. pmid:16407148
- 57. Trungchanh T, MichioTsutsumi, KinkichiKurihara. Comparative study on the response of Indica and Japonica rice plants to ammonium and nitrate nitrogen. Soil Sci Plant Nutr. 1981;27(1):83–92.
- 58. Zhang J, Liu Y-X, Zhang N, Hu B, Jin T, Xu H, et al. NRT1.1B is associated with root microbiota composition and nitrogen use in field-grown rice. Nat Biotechnol. 2019;37(6):676–84. pmid:31036930
- 59. Hazen TC, editor Field-Integrated Studies of Long-Term Sustainability of Chromium Bioreduction at Hanford 100H Site. AGU Fall Meeting; 2006.
- 60. Wu Q, Knowles R, Niven DF. O2 regulation of denitrification in Flexibacter canadensis. Can J Microbiol. 1994;40(40):916–21.
- 61. Lucas JA, Garcíavillaraco A, Ramos B, Garcíacristobal J, Algar E, Gutierrezmañero J. Structural and functional study in the rhizosphere of Oryza sativa L. plants growing under biotic and abiotic stress. J Appl Microbiol. 2013;115(1):218–35. pmid:23594253
- 62. Tripathi BM, Kim M, Laihoe A, Shukor NA, Rahim RA, Go R, et al. pH dominates variation in tropical soil archaeal diversity and community structure. FEMS Microbiol Ecol. 2013;86(2):303–11. pmid:23773164
- 63. Long XE, Yao H, Wang J, Huang Y, Singh BK, Zhu YG. Community structure and soil pH determine chemoautotrophic carbon dioxide fixation in drained paddy soils. Environ Sci Technol. 2015;49(12):7152–60. pmid:25989872
- 64. Duan G-L, Ding L-J, Cui H-L, Zhu Y-G, Nie S-A, Long X-E. Microbiomes inhabiting rice roots and rhizosphere. 2019.
- 65. Mendes R, Garbeva P, Raaijmakers JM. The rhizosphere microbiome: significance of plant beneficial, plant pathogenic, and human pathogenic microorganisms. FEMS Microbiol Rev. 2013;37(5):634–63. pmid:23790204