Response of soil microbial community to application of biochar in cotton soils with different continuous cropping years

The bacterial community in soils of cotton that have continuously been cropped for 2 years, 6 years, 11 years and 14 years and treated with biochar (B0, 0 t·ha−1; B1, 12.5 t·ha−1; and B2, 20 t·ha−1) was investigated using next-generation sequencing. Of the 45 bacterial genera (relative abundance ratio of genera greater than 0.3%), 21 genera were affected (p < 0.05) by the biochar treatment, whereas 20 genera were affected by the continuous cropping. Between the soils that have been continuously cropped for 2 years and 14 years, 12 different genera were significantly observed (p < 0.05), and 6 genera belonged to the phylum Acidobacteria. The relative abundance of Sphingomonas and Pseudomonas in the biochar-treated soils was significantly higher than that in the soil without biochar treatment (p < 0.05), and the relative abundance of Sphingomonas and Pseudomonas in soils that have been continuously cropped for 2 years and 6 years was significantly higher than that in the soils continuously cropped for 11 years and 14 years (p < 0.05). The results suggest that the biochar application has a significant impact on the soil bacterial community, which may improve the microbial diversity of continuous cropping systems in cotton soils.

continuously cropped system in Hubei Province. This study aimed to characterize the soil bacterial communities among the different continuous cotton soil field cropping systems treated with biochar to answer the following questions: (i) Does the overall soil bacterial community differ among different continuous cotton field systems? (ii) Can biochar treatments change the soil bacterial community of different continuous cotton field systems? This information will improve our understanding of the relationship between biochar application and soil microbial ecology in continuous crop systems of cotton.

Results
Sequence data and bacterial taxonomic richness. A total of 1,669,315 paired-end 250 bp reads were acquired, and the average read length per sample was 0.16 Gb, with 87,323 to 204,627 raw reads in cotton soils with different continuous cropping years and biochar treatments (Table S1). After the initial quality control process, 1,608,450 high-quality sequences were obtained. On average, 134,038 sequences were obtained per sample. Based on 97% species similarity, 6552 to 9420 operational taxonomic units (OTUs) were separately obtained from the samples of different continuously cropped cotton soils with biochar treatment (Table S1). The average length of the sequence reads was 440 bp, and they were classified into different taxonomic groups using Uclust 17 . The bacterial diversity is reflected by the Chao1 index; the Chao1 index in the soils cropped continuously for 2 years and 6 years was higher than that in the soils continuously cropped for 11 years and 14 years (Fig. 1), with significant differences (except for the 6-B0 soil). This finding indicates that the bacterial diversity decreased during the continuous years. Fig. S1 describe the distribution of the DNA sequences into phyla. A total of 34 phyla were shared by the 12 soil samples. The main phyla were as follows: Proteobacteria, Acidobacteria, Actinobacteria, Nitrospira, Bacteroidetes, Gemmatimonadetes, Planctomycetes, Firmicutes and Verrucomicrobia. Proteobacteria was the most dominant among the 34 phyla (p < 0.05) in the samples and comprised 39% to 54.5% of the total sequences. Acidobacteria was the second-largest phylum in all groups, comprising approximately 14.2% to 21.3% of the different groups (Fig. 2). Seven phyla (Proteobacteria, Acidobacteria, Nitrospira, Bacteroidetes, Planctomycetes, Firmicutes and Verrucomicrobia) differed (p < 0.05) between the continuous cropping year and biochar treatment (Table S2). The relative abundance of Proteobacteria, Nitrospira, Planctomycetes and Firmicutes was significantly reduced by the continuous year, while the relative abundance of Acidobacteria and Bacteroidetes increased by the continuous year (Fig. 2). Biochar treatment increased the relative abundance of Proteobacteria in the soils continuously cropped for 2 years and 6 years but not in the continuous 11-year and 14-year soils. Biochar treatment also increased the relative abundance of Planctomycetes and Firmicutes, especially for the B1 (12.5 t·ha −1 ) treatment regarding Planctomycetes. However, biochar treatment reduced the relative abundance of Acidobacteria and Bacteroidetes in different continuous-year soils.

Effect of biochar on the bacterial community composition. Results shown in
At the genus level, a total of 853 genera were identified from all the samples, regardless of the treatments administered. The 46 most abundant genera (with relative abundance of more than 0.3%), comprising more than 66.6% of the total sequences, are shown in Fig. S2. Of the 46 genera, 29 were affected by the continuous cropping year, biochar treatment and interactions between the continuous cropping year and biochar treatment (p < 0.05). Among these genera, 20 were affected by the continuous cropping year, 21 by the biochar treatment and 12 by the interactions between continuous cropping year and biochar treatment ( Table 1). The relative abundance of Gemmatimonas, Nitrospira, Sphingomonas, Pseudomonas, Dongia, Phaselicystis, Kofleria, Nitrosospira, Geobacter, Ramlibacter, Novosphingobium and Ohtaekwangia was significantly reduced by the continuous year, while the relative abundance of Gp6, Gp4 and Gp10 was increased by the continuous year. Biochar treatment could increase the relative abundance of Gemmatimonas, Nitrospira, Sphingomonas, Pseudomonas and Anaeromyxobacter but reduce the relative abundance of Rhodoplanes, Ilumatobacter, Gp6, Gp4 and Gp10 ( Table 2).
The five most abundant genera were Sphingomonas, Gemmatimonas, Nitrospira, Pseudomonas and Gp6 (Fig. S2). The relative abundance of Sphingomonas and Pseudomonas in the biochar-treated soils was significantly higher than that the soils without biochar treatment (Fig. 3, p < 0.05), and the relative abundance in the cotton soils continuously cropped for 2 years and 6 years was significantly higher than that of the 11-year and 14-year continuously cropped soils (p < 0.05). However, the relative abundance of Nitrospira in the soils without biochar treatment in different continuous cropping years was significantly higher compared with that of the biochar-treated soils (Fig. 3, p < 0.05).
Eight genera significantly differed (p < 0.05) between soil continuously cropped for 2 years and 6 years, and 10 genera differed between 2-year and 11-year continuously cropped soils ( Table 3). Most of the genera belonged to the phylum Proteobacteria. However, between soils continuously cropped for 2 years and 14 years, 12 significantly different genera (p < 0.05) were identified, with 6 of those genera belonging to the phylum Acidobacteria. The relative fold changes in Pseudomonas between soils continuously cropped for 11 years and 2 years and between 14-year and 2-year soils were −19.81 and −19.02, respectively, and the differences were significant (p < 0.05). The soil samples were separated into categories, which matched their different continuous cropping year (Fig. 4). The separation was clearer for the cotton soils continuously cropped for 11 years and 14 years than for that of 6-year and 2-year soils, suggesting that the number of years of continuous cropping influenced the bacterial community composition. Biochar has a significant effect on soil bacterial composition. Twenty-one genera were affected (p < 0.05) by biochar treatment (Table 1). Specifically, 10 different genera were observed (p < 0.05) between B1and B0-treated soils, and 13 genera between B2-and B0-treated soils based on T-test results (Table 4). Among the 21 genera, the relative fold changes of Pseudomonas, Sphingomonas, Pasteuria and Nitrospira between B2and B0-treated soils were 12.45, 19.40, −19.48 and −17.75, respectively, and the differences were significant (p < 0.05). Our data suggest that the genera in the biochar-treated soil were significantly different. Pseudomonas, a major rhizosphere-promoting bacteria, was significantly improved by the biochar treatment, suggesting that the biochar treatment changes the relative abundance of Pseudomonas. The soil samples were separated into categories, which matched the biochar treatments (Fig. 4). Principle component analysis separated the biochar treatment and the control soils, suggesting that biochar treatment influences the bacterial community composition.

Discussion
Bacterial community composition and its relative proportion in soil microbial communities varied and was influenced by both the biochar treatment and number of continuous cropping years. This result is in accordance with the findings of previous studies 18,19 that also investigated the changes in the taxonomy of soil microbial communities after biochar amendment. Possible causes of the ecological shifts in the relative abundance of the bacterial community observed may include effects of the root exudates on a portion of the microbial community as a result of continuous cropping; changes to the physiochemical state of the soil environment, including pH, mineral content, pore and particle size; and changes to the water and nutrient availability due to the application of biochar and to the continuous cropping system, all of which have been shown to influence the composition of the bacterial community in soils 20 . Further work will be required to determine the relative importance of these factors in altering the composition of the microbial community due to biochar amendments to soils.
The relative abundance of Acidobacteria in the cotton soil continuously cropped for 14 years was approximately 57.3% to 184.3% higher compared with that of the 2-year continuously cropped soil (Fig. 2). Of the 15 significantly different genera found in the soils continuously cropped for 14 years and 2 years, 6 belong to Acidobacteria (Table 3). This result is probably due to the soil organic matter and especially the pH. The pH in cotton soil that has been continuously cropped for 14 years was significantly lower than that in the 2-year continuously cropped soil (Table S3). Soil pH has been recently documented in various soil samples as the major factor that determines soil bacterial diversity and composition. Previous studies [21][22][23] reported that soil pH influences bacterial communities in soils across North and South America, in Britain and on Changbai Mountain. The effects of soil pH on the relative abundance of some bacterial groups in this study are consistent with these studies, which indicate that the relative abundance of Acidobacteria tends to increase with lower pH values [22][23][24][25][26] . Thus, our results further emphasize that soil pH plays an important role in shifting the composition of the bacterial community in the cotton soils with different continuous cropping years. However, Chan et al. 27 reported that the addition of biochar to soil can significantly increase the pH of the bulk soil, which potentially provides a more favorable habitat for microbial organisms, especially bacteria that are sensitive to pH. This result is in accordance with the findings of our study. The application of biochar increased the pH of cotton soils that have been continuously cropped for a different number of years (Table S3). Previous studies 3,4,28,29 have reported that high porosity, cation exchange capacity and sorption capacity of biochar provide a suitable habitat for microorganisms, promoting their activity in soil and affecting different microbial processes involved in nutrient cycling and organic matter decomposition. In this study, the four most abundant genera that were significantly affected by the biochar treatments were Sphingomonas, Gemmatimonas, Nitrospira and Pseudomonas (Table S2). Biochar treatment improved the relative abundance of Sphingomonas and Pseudomonas (Fig. 2), possibly due to biochar providing a suitable habitat. At the same time, the soil microbial communities are mostly limited by carbon sources, and biochar could provide abundant carbon resources for microbial growth; thus, the greater quantity of a few dominant microorganisms, such as Sphingomonas and Pseudomonas, as a result of the biochar treatment may also be due to improved carbon sources.
genera abundance 1, 3 , which contributes to increased cotton productivity (Table S4). In addition, biochar is highly recalcitrant to microbial decomposition and thus guarantees a long-term benefit to soil fertility 33 . The actual effects of biochar application depend on various factors, such as the soil type and the water balance at a given site and possibly even the cultivated genotype, which currently require further studies.

Material and Methods
Site description. The   Soil and Biochar. The soil was collected from the surface layer of cotton soils (0-15 cm) that have been continuously cropped for 2, 6, 11 and 14 years at the cotton research station in Qianjiang City (Fig. 5). The collected soil was classified as acrisols according to the FAO, and the clay content was 66.3%. The original years of '2, 6 and 11-year soil' was planted with corn, and the culturing and soil management were consistent. Corncob was used as the feedstock for biochar. The corncob was first air-dried and then pyrolyzed under controlled conditions to ensure uniform heating and treatment conditions. Biochar production was carried out using a traditional kiln reactor (Fengben Biological Technology Co., Ltd, Shandong, China) at a heating rate of 10 °C min −1 up to 550 °C. The basic properties of the collected soil and biochar are presented in Table 5.

Pot experimental design. A pot trial was conducted at the experimental base of the Industrial Crops
Institute, Hubei Academy of Agricultural Sciences in Hubei Province, China. This region has a typical subtropical monsoon climate, with an average annual precipitation of 1269 mm and average temperature between 15.8 °C and 17.5 °C.
After the removal of plant debris and stones, soils collected from different continuous cropping cotton fields (2, 6, 11 and 14 years) were mixed with biochar. The additive amount of collected soil in each experimental pot (the upper bore and pot height were 40.0 cm and 27.0 cm, respectively) was 7.5 kg of dry soil. The biochar application rates of each continuously cropped soil were 0 t·ha −1 (B0), 12.5 t·ha −1 (B1) and 20 t·ha −1 (B2), resulting in biochar weights of 0%, 1.538% and 3.077%, respectively, which accounted for the dry weights of the potted soil. In total, there were four continuous cropping soils, and each had three application rates of biochar; thus, there were twelve treatments in this experiment. Each treatment was replicated in ten pots; therefore, the twelve treatments employed 120 pots. Basal fertilizer was added to all pots. Equal amounts (10 g) of compound fertilizer (N:P:K = 15:15:15) were applied to all pots. The cotton seed was sown on 1 May 2014. One cotton (Gossypium      tions. The quality of the extracted DNA was preserved using 1% agarose gels. The V3-V4 hypervariable regions of 16 S rRNA were amplified via PCR from the microbial genomic DNA using barcoded fusion primers (forward primers: 341 F CCTACACGACGCTCTTCCGATCTN (barcode) CCTACGGGNGGCWGCAG, reverse primers: 805 R GACTGGAGTTCCTTGGCACCCGAGAATTCCAGACTACHVGGGTATCTAATCC). The reaction mixtures (50 µl) contained 5 µl of 10 × PCR reaction buffer (TakaRa, Japan), 10 ng of DNA template, 0.5 µl of each primer, 0.5 µl of dNTPs and 0.5 µl of Platinum Taq DNA polymerase (TakaRa, Japan). The PCR conditions were as follows: 94 °C for 3 min, 94 °C for 30 s, annealing at 45 °C for 20 s and 65 °C for 30 s, which was repeated for 5 cycles, followed by 94 °C for 20 s, 55 °C for 20 s and 72 °C for 30 s, which was repeated for 20 cycles, before a final elongation at 72 °C for 5 min. The PCR product was excised from the 1.5% agarose gel and purified using a QIAquick Gel Extraction Kit.
Amplicon sequence and sequence data processing. The barcoded V3 and V4 amplicons were sequenced using the paired-end method with an Illumina MiSeq (Illumina, San Diego, CA, USA) system with a 6-cycle index. Sequences with an average Phred score of less than 25 that contain ambiguous bases, a homopolymer run exceeding 6, mismatches in primers or a length of less than 100 bp were removed using Prinseq software (PRINSEQ-lite 0.19.5). For the V3 and V4 paired-end reads, only the sequences that overlapped by more than 10 bp and without any mismatch were assembled according to their overlapping sequences using Flash software (FLASH v1.2.7). Reads that could not be assembled were discarded. Barcode and sequencing primers were trimmed from the assembled sequence (V3 and V4).   Table 5. Basic chemical properties of biochar and soil. The data are expressed as mean ± SD (n = 3). Superscript letters that differ within rows indicate significant differences between treatments (p < 0.05).
Sequences were clustered and assigned to operational taxonomic units (OTUs) at a 3% dissimilarity level using Uclust software (Uclust v1.1.579). Taxonomic ranks were assigned to each sequence using the Ribosomal Database Project (RDP) Naïve Bayesian Classifier v.2.2 trained on the Greengenes database (October 2012 version) (Lan et al. 2012). The relative abundance count at the genus level was log2-transformed and then normalized, as described in the following. The arithmetic mean of all transformed values was subtracted from each log-transform measured, and the difference was divided by the standard deviation of all log-transformed values for a given sample. After this procedure, the relative abundance profiles for all the samples exhibited a mean of 0 and a standard deviation of 1. Principal coordinates analysis (PCoA) at the genus level was performed using Bray-Curtis distances with Mothur 1.29.2 software. Statistical analysis. The results were analyzed using the SPSS software program (v10.0 for Windows, Chicago, IL, USA). The differences in the relative abundance of individual genera and the treatment means among plant age were tested using one-way variance analysis (ANOVA), and significant differences among the means were determined using the LSD test. Normal distribution and homogeneity of variance were verified by the Bartlett and Dunnett tests. The differences were considered statistically significant when p < 0.05. T-tests and Metastats (http://metastats.cbcb.umd.edu/) in Mothur were used to compare the differences, and all p-values were adjusted with the false discovery rate (FDR) using the BH method with the mt.rawp2adjp function in R.