Soil properties and agricultural practices shape microbial communities in flooded and rainfed croplands

Applied Soil Ecology xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Applied Soil Ecology journal homepage: www.elsevier.com/locate/apsoil

Soil properties and agricultural practices shape microbial communities in flooded and rainfed croplands Xiaoyan Wanga,1, Tianhua Heb,1, Shiying Gena, Xiao-Qi Zhangb, Xiao Wangc, Dong Jiangc, ⁎ ⁎⁎ Chunyan Lid, Chaosu Lie, Jianlai Wangf, Wenying Zhanga, , Chengdao Lib,g, a

College of Agriculture, Yangtze University, No. 88 Jingmi Road, Jingzhou, Hubei 434025, China Western Barley Genetic Alliance, Murdoch University, Western Australia 6150, Australia c National Technique Innovation Center for Regional Wheat Production/Key Laboratory of Crop Eco-physiology and Production, Ministry of Agriculture, Nanjing Agricultural University, No. 1 Weigang Road, Nanjing, Jiangsu 210095, China d College of Agriculture, Yangzhou University, No. 88 Daxuenan Road, Yangzhou 225009, China e Sichuan Academy of Agricultural Sciences, No. 20 Jiangjusi Road, Chengdu, Sichuan 610066, China f Anhui Academy of Agricultural Sciences, No. 40 Nongkenan Road, Hefei, Anhui 230031, China g Department of Primary Industry and Regional Development, Government of Western Australia, South Perth WA6155, Australia b

ARTICLE INFO

ABSTRACT

Keywords: Acidobacteriaceae Gemmatimonadaceae Gene subsystem Metagenomics MG-RAST Soil pH Shotgun sequencing Soil microbial community Yangtze River Basin

Understanding the dynamics of soil microbial communities and the factors that affect those dynamics is crucial for comprehending the processes affecting soil fertility in agricultural ecosystems. Here, we used shotgun DNA sequencing to characterise 29 soil microbial communities in flooded paddies in China's Yangtze River Basin, and comparatively analysed the composition and function of microbial communities with 132 communities from North and South America's rainfed cropland. We hypothesised that soil microbial community diversity and functional composition are predominantly determined by edaphic properties and land-use history, rather than by spatial distance and climate. We revealed significant differences in taxonomic structure and functional composition among the microbial communities collected from a 2000 km transect along the Yangtze River and found that taxonomic diversity and genomic functional composition of the soil microbial communities were predominantly defined by soil pH. The significant correlation between soil pH and microbial community diversity can be extended to soils from different continents. Microbial communities in flooded paddies in China differed significantly from those in rainfed croplands in North and South America, while the communities from rainfed croplands in North and South America were similar despite their significant differences in geographic distance. Together with available evidence, our results suggest that soil microbial diversity is controlled primarily by edaphic variables rather than by climate, which differs fundamentally from the global biogeography of macro-organisms. The predominant element of soil properties (soil pH in particular) and the response of particular taxa to changing pH provides insight for the selection of agronomic practice. Agronomic practice and fertilisation may change soil pH and therefore alter soil microbial diversity and composition, which could have an impact on soil nutrient supply in agricultural ecosystems. Significance: Traditional methods may underestimate the microbial diversity. Metagenome analysis using whole genome sequencing allows quantification of the diversity more accurately and characterisation of the function of the mega-diverse soil microbial communities simultaneously. Using metagenome analysis on 141 soil microbial communities collected from global agriculture soil system with 32 from flooded paddies in China and 129 from rain-fed farmland in North and South America. We found that the significant correlation between soil pH and diversity of microbial communities can be extended to soils from different continents, which is fundamentally different from the diversity and community structure of macro-organisms. Moreover, we revealed that land use and cultivation history have shaped the taxonomic and functional composition in the soil microbial community. Our study deepens our understanding of the structure and function of the soil microbial community in general, and have particularly important implications in agricultural practice.

Corresponding author. Correspondence to: Chengdao Li, Western Barley Genetic Alliance, Murdoch University, Western Australia 6150, Australia. E-mail addresses: [email protected] (W. Zhang), [email protected] (C. Li). 1 These authors contributed equally to this work. ⁎

⁎⁎

https://doi.org/10.1016/j.apsoil.2019.103449 Received 4 June 2019; Received in revised form 30 October 2019; Accepted 14 November 2019 0929-1393/ © 2019 Published by Elsevier B.V.

Please cite this article as: Xiaoyan Wang, et al., Applied Soil Ecology, https://doi.org/10.1016/j.apsoil.2019.103449

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

1. Introduction

between soil microbial community and soil fertility, it is paramount to detail the soil microbial communities and their pattern of composition and function across the vast area of YRB. Apart from edaphic properties (soil pH in particular), evidence exists that land-use history was a critical factor driving the diversity and structure of soil microbial communities (e.g. Suleiman et al., 2013; Goss-Souza et al., 2017). Soils in YRB have been used for flooded rice paddies, with the sporadic rotation of dryland crops, for almost 10,000 years (Zong et al., 2007). In contrast, traditional agriculture in both North and South America has used crops in rainfed farming (maybe with irrigation, but not flooded) for hundreds of years (Grau and Aide, 2008), potentially creating soils that differ fundamentally from the flooded paddies in YRB. It is of particular interest to compare the soil microbial communities under the long-term practice in flooded paddies and those with rainfed land. We hypothesised that soil microbial community diversity and functional composition are predominantly determined by edaphic properties and land-use history, rather than by spatial distance and climate. We aimed to: 1) determine the pattern of microbial community composition and structure in soils across the vast YRB using shotgun metagenomic sequencing methodology; 2) compare changes in microbial community composition and function diversity, and determine the influence of soil properties and climatic factors on microbial community composition and diversity; and 3) test the general pattern of influence of edaphic properties and land-use history on microbial community diversity and functional composition. To this end, we collected soil samples from croplands across YRB and characterised the taxonomic composition and functional pattern of soil microbial communities using shotgun metagenomics. We also retrieved comparable data of soil microbial communities across continents from the public depository, the Meta-Genome Rapid Annotation with Subsystem Technology (MG-RAST) server (Meyer et al., 2008), to test for the general driver of taxonomic composition and functional pattern.

Increasing and maintaining crop production for future food and energy demands while preserving soil quality and function is the major challenge in modern agriculture (Doran, 2002; Godfray et al., 2010). Soil fertility is the key for sustaining crop production. In modern agriculture, soil fertility is maintained predominantly by external fertiliser inputs. However, with increased water and environmental contamination (Bennett et al., 2001), it is vital to understand the dynamics of lifesupporting elements in soil, such as nitrogen, phosphorus and carbon, which lies in the connection between their forms in soil and is modulated by soil microbial community biology (Scholes and Scholes, 2013). Even though croplands are intensively managed, they rely on microbialmediated processes for overall soil sustainability (Osler and Sommerkorn, 2007; Bissett et al., 2013). Therefore, understanding the dynamics of soil microbial communities and the factors that influence those dynamics is crucial for comprehending the processes that affect soil fertility (Carbonetto et al., 2014). Ecological approaches, such as diversity and function analyses of soil microbial communities and causal analysis of the influence of external factors on their composition and structure, have the potential to address those questions (Fierer et al., 2012, 2013; Carbonetto et al., 2014). Studies have revealed that soil microbial diversity and biogeography are controlled by soil properties, such as soil pH, soil carbon and nitrogen (Fierer and Jackson, 2006; Rousk et al., 2010; Zhalnina et al., 2015; Yuan et al., 2017; Ren et al., 2018). Among climates and seasons, soil types and structures, and aboveground vegetation—which have been investigated for determinants of soil microbial community—aboveground plants are thought to have a significant effect on microbial structure and diversity patterns (Garbeva et al., 2004; Rasche et al., 2006; Ren et al., 2018). It is believed that each plant species may be associated with a microbial population (Bever et al., 2012; Leff et al., 2018; Schmid et al., 2019). For example, a review by Berg and Smalla (2009) showed that each plant species is colonised by specific rhizosphere microorganisms. Soil properties and aboveground plant species are, therefore, the two most significant variables influencing the structure and function of soil microbial communities. To fully understand the extent to which both factors contribute to microbial communities, it would be beneficial to control one element and investigate the influence of the other. Croplands offer a unique opportunity as the aboveground vegetation is easily controlled, usually as a uniform crop species or several crop species in rotation over a long period (Wu et al., 2009). Therefore, differences in the structure and function of soil microbial communities are more likely the result of varying soil properties. In return, knowledge of soil microbial community dynamics and the factors that influence those dynamics in croplands help us to understand the processes that affect soil fertility, which is fundamentally important in cropland production systems. Soil microbial diversity and community structure was quantified using Terminal-Restriction Fragment Length Polymorphism (T-RFLP) analyses in earlier studies (e.g. Fierer and Jackson, 2006), and later by 16s RNA amplicon sequencing (e.g. Rousk et al., 2010; Tripathi et al., 2012; Ren et al., 2018). The T-RFLP method could underestimate microbial diversity due to its limit on the number of DNA bands on gel and the homology of restricted DNA (Fierer and Jackson, 2006), while the more robust taxonomic assignment using 16s RNA amplicon sequencing could be limited by the databases used for sequence comparisons (Poretsky et al., 2014). Shotgun metagenomic sequencing allows for quantification of the diversity and characterisation of function simultaneously, which is very powerful as shifts in taxonomy are usually accompanied by changes in the metabolic functional potential of soil microbial communities (Urich et al., 2008). China's Yangtze River Basin (YRB) is one of the most densely populated and agriculturally productive areas in the world (Yan et al., 2011), but there are long-term trends in reduced soil fertility and stability in the region (Zhang et al., 2015). Given the close connection

2. Materials and methods 2.1. Soil properties and sampling procedure Soil samples were obtained from five locations along the Yangtze River catchment in China, separated by approximately 80 to 2000 km (Table S1). Sampling was carried out from 4 to 6 December 2017 after rice had been harvested for about 50 days and wheat had been sown for about 20 days. At the time of sampling, the soil was not submerged and wheat was in the over-winter stage. At each of the five locations, six soil samples (~50 g each) were taken at 10 cm depth using a soil corer. For each sample, half of the soil was used for DNA extraction and later sequencing, and the remaining half for soil properties analysis. The soil samples contained no plant root residues. Soil properties, including soil pH, readily available phosphorus (AP), alkaline hydrolysis nitrogen (available nitrogen, AN) and organic matter (OM), were measured for three replicates from each site. Determination of AP, AN and OM followed the methods described in Qiao (2012). Meteorological data (daily temperature and rainfall between 1 Oct 2017 and 30 June 2018) were obtained from the nearest weather station (within 10 km) at each location. 2.2. Metagenome sequencing, initial quality control and assembly We used shotgun sequencing metagenomics and computational analysis to compare the taxonomic and functional profiles of microbial communities in the 29 soil samples and determine the influence of soil properties and climate on taxonomic and functional composition. Bulk nucleic acids (DNA) in each sample were extracted using an E.Z.N.A. DNA Kit (Omega Bio-Tek Inc.) following the manufacturer's protocol. Sequencing of the metagenome and initial quality control were carried out at the Beijing Genomic Institute (BGI, Shenzhen, China) following their standard procedures. Briefly, qualified bulk DNA from each sample was sheared into smaller fragments by nebulisation. The 2

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

overhangs resulting from fragmentation were converted into blunt ends using T4 DNA polymerase, Klenow fragment and T4 polynucleotide kinase. An A (adenine) base to the 3′ end of the blunt phosphorylated DNA fragments was then added to the adapters that were ligated to the ends of the DNA fragments. The short DNA fragments were extracted with Ampure beads. An Agilent 2100 Bioanalyzer and ABI StepOnePlus Real-Time PCR System were used to qualify and quantify the sample libraries. The qualified libraries were then sequenced with an Illumina HiSeqTM platform at BGI. The DNA sequences were binned by barcode, quality filtered, and the first 15 base pairs (bp) were trimmed to remove bias introduced by the Illumina primers. Further quality control included discarding reads containing: a) 10% or more ambiguous bases (N base), b) adapter sequences (default: 15 bases overlapped by reads and adapter) and c) 50% or more low quality (Q < 20) bases. The short reads were then assembled into longer contigs with IDBA_UD (Peng et al., 2012). For each sample, reads were assembled with a series of different k-mer sizes in parallel and mapped back to each assembly result with SOAP2 (Li et al., 2009), before selecting the optimal k-mer size and assembly result depending on both contig N50 and mapping rate. During the assembly process, only contigs with no < 500 bp were kept for further analysis. Contigs were analysed using the MG-RAST server (Meyer et al., 2008), as described below.

Table 1 Soil chemical and nutrient characteristics at the 29 sampling sites. S.D.: standard deviation.

2.3. Metagenome analysis pipeline in MG-RAST Taxonomic assignment and function characterisation for each sample (referred to as a microbial community) was implemented using the MG-RAST server, an open-source service for pipelines built for highperformance computing of taxonomic assignment and functional analysis of metagenomes and a public repository for metagenomic data. Clean sequences generated by BGI were uploaded to the MG-RAST server for analysis. MG-RAST compares DNA sequences against a large group of protein and nucleotide databases for the automated assignment of metagenomic sequences to their respective taxonomic and functional groups. An average of 10.6 Gb of clean sequencing data per sample was uploaded to the MG-RAST server, and the metagenome of each soil sample was assigned a unique and retrievable MG-RAST reference ID (Supplementary Table S1). We used the pipeline options according to the default settings in MG-RAST. Further data quality control was carried out in MG-RAST, along with a normalisation step. For taxonomic assignment, the MG-RAST server identifies candidate RNA genes by comparing the sequence data against rRNA databases Greengenes (DeSantis et al., 2006), the European 16S RNA database (Wuyts et al., 2002), RDP-II (Cole et al., 2006), and boutique databases (Leplae et al., 2004) simultaneously. For functional characterisation, we examined the gene subsystems—groups of genes with functional roles that act collectively in a biological process, for example, in a metabolic pathway. These genes are grouped into functional subsystem categories. The MG-RAST server uses the Basic Local Alignment Search Tool X (BLASTX) and MD5 non-redundant database M5nr to assign genes to clustering-based subsystems as functional groups. To do so, sequence similarity was searched against several databases simultaneously—KEGG (Kyoto Encyclopedia of Genes and Genomes), NCBI (National Center for Biotechnology Information), RDP (Ribosomal Database Project), SEED (The SEED Project, Overbeek et al., 2005), UniProt (UniProt Knowledgebase), and eggNOG (evolutionary genealogy of genes Non-supervised Orthologous Groups) (Meyer et al., 2008). Genes were assigned to a subsystem by homology searches of protein sequences encoded by the reads. For both taxonomic assignment and function annotation, the match threshold was set at an expected value of < 1 × 10−5, similarity of > 50 bp, and a minimum identity of 65%. If no match was found in the relevant database, the sequence was classified as ‘unclassified’. The taxonomic assignment and functional annotation results for each sample are deposited and retrievable from MG-RAST following the reference ID (Supplementary Table S1).

Sites

Soil pH

Available nitrogen (mg/kg)

Available phosphorus (mg/kg)

Available potassium (mg/kg)

Organic matter (g/kg)

GHNSR2 GHNSR4 GHNSR6 GHSR2 GHSR4 GHSR6 HFNSR2 HFNSR4 HFNSR6 HFSR2 HFSR4 HFSR6 JLNSR2 JLNSR4 JLNSR6 JLSR2 JLSR4 JLSR6 JTNSR4 JTSR6 JTSR2 JTSR4 JTSR6 YZNSR2 YZNSR4 YZNSR6 YZSR2 YZSR4 YZSR6 Average S.D.

8.02 7.99 8.11 7.93 8.02 8.05 6.82 6.11 5.75 6.59 6.81 6.48 8.12 8.11 8.15 7.94 7.98 7.99 7.81 7.88 8.00 8.23 8.09 7.27 7.09 7.50 7.64 7.69 7.43 7.57 0.12

88.4 84.2 94.2 81.4 85.9 104.0 56.1 97.7 76.9 53.7 43.7 55.4 64.8 65.1 54.4 81.4 80.4 74.8 81.4 73.5 36.0 33.6 40.2 50.2 78.0 71.0 69.3 60.6 64.1 68.9 3.4

23.0 24.5 33.4 17.0 17.9 20.3 12.2 10.7 13.1 15.2 10.7 18.8 16.1 15.8 18.5 28.0 14.3 16.1 25.3 23.3 14.0 8.9 20.6 89.8 86.8 71.0 58.5 57.6 62.6 29.1 4.3

89.1 96.5 106.5 94.1 94.1 96.5 89.1 96.5 84.1 101.5 86.6 96.5 119.0 91.6 89.1 126.5 101.5 84.1 66.6 69.1 69.1 86.6 153.9 119.0 136.5 146.4 126.5 161.4 183.9 105.5 5.3

31.38 30.51 35.63 28.98 33.59 36.79 16.41 38.10 24.87 14.02 11.86 15.84 23.42 25.28 22.47 28.00 29.39 26.89 29.72 25.52 13.37 8.96 14.58 25.31 25.97 26.42 23.63 25.51 26.07 24.78 1.40

Table 2 Ten most abundant bacteria families in flooded paddies in the Yangtze River Basin. Family

Average abundance (%)

Gemmatimonadaceae Acidobacteriaceae Anaerolineaceae Sphingomonadaceae Burkholderiaceae Geobacteraceae Anaeromyxobacteraceae Rhodocyclaceae Solibacteraceae Comamonadaceae

3.0269 1.3491 1.3258 1.2920 1.2709 1.2336 1.2125 1.1821 1.1649 1.0822

Table 3 Analysis of similarity between soil microbial community composition (at the family level) at five locations. p-Values are shown (with Bonferroni correction).

Guang Han He Fei Jiang Ling Yang Zhou

Jin Tan

Guang Han

He Fei

Jiang Ling

0.033 0.021 0.034 0.052

0.024 0.025 0.022

0.024 0.031

0.021

2.4. Characterisation of community taxonomy and function and the influence of soil properties Taxonomic assignments were extracted from the MG-RAST server using representative match classification at the domain and family level, as taxonomy is less ambiguous at the family level and above in metagenomics analysis (Escobar-Zepeda et al., 2018). The relative 3

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

Fig. 1. Principal component analysis showing the top ten families differentiating the 29 microbial communities.

Fig. 2. Relative abundance of the ten most differentiated families from the sampling sites.

abundance of taxa was extracted as the proportion of sequences assigned to each matched microbial family relative to the total number of taxonomic assignments from MG-RAST (details in Supplementary Table S2). The relative abundance of functional gene subsystems is defined as the proportion of protein features assigned to each functional subsystem relative to the total number of annotated features assigned to functional categories (details in Supplementary Table S3). Functional assignments reported in the text are based on the subsystem hierarchy of MG-RAST and extracted from MG-RAST using the same match filters as for taxonomic assignments. Further data statistics were carried out with software package PAST V3 (Hammer et al., 2001). We first performed a one-way analysis of

similarity (ANOSIM) between pairwise of location samples (with six site samples in a location group) with 9999 permutations. We further employed principal component analysis (PCA) to explore and illustrate the relationship of taxa abundance and composition of all identified families among the 29 soil samples and reveal the taxa that differed significantly among soil microbial communities. Alpha diversity (a within-sample diversity calculated in MG-RAST), index of dominance and Shannon's index for diversity (at the family level) for each soil microbial community were calculated. The relative influence of elements of soil properties and climatic factors—including soil pH, average temperature during the coldest month (January), total rainfall during recording period, soil available nutrient and phosphorus levels, and soil 4

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

organic matter level—on alpha diversity, dominance and Shannon's index were tested through multiple regression analysis assuming a linear relationship. The relationships between microbial alpha diversity with soil pH (the soil property revealed to influence the microbial community, see Results) were tested with Pearson's correlation analyses. The relationships between the taxonomic distance (Euclidian distance) between two sites with environmental distance (Euclidian distance in soil pH) and the physical distance (logarithm transformed) between two sites were tested using regression analysis with linear or quadrate relationship assumed, and the relationship with smaller Akaike IC value was chosen. Significance was taken at p < 0.05. When multiple comparisons were made, the Bonferroni correction of p-value was used. To compare taxonomic composition and functional structure of soil microbial communities in agricultural fields globally, we extracted available datasets archived in MG-RAST using the criteria of “metagenomes” and “agricultural feature (or agricultural field)”. Only datasets with > 1 million reads per sample were retained. Consequently, data for 132 samples were retrieved (as of 31 December 2018), including 91 from North America (the USA and Canada), 36 from Brazil, three from China and two from France. Their MG-RAST sample ID, library ID and other details are provided in Supplementary Table S1. For each sample, alpha diversity was retrieved as a diversity index, relative abundance of families was retrieved as taxonomic composition, and the relative abundance of gene subsystem retrieved as functional structure. Soil pH was recorded where available (52 samples). These data and our data were initially analysed using the MG-RAST server with the same pipeline and against the same set in the reference database. For comparison in taxonomic composition, only families with relative abundance > 0.01% (377 families) were retained. The previous data were combined with our data to compare taxonomic composition and functional structure using PCA analysis in PAST V3 (Hammer et al., 2001). Significance was taken at p < 0.05. When multiple comparisons were made, the Bonferroni correction of p-value was used. 3. Results The soil from the 29 sites (five sites from location Jin Tan due to sequencing failure for one sample) varied in soil pH from acidic (at location He Fei) to alkaline in other locations (Table 1). Major nutrient elements (available nitrogen, phosphorus and potassium, and organic matter) differed considerably across the sampling sites, between and among the five locations. Shotgun sequencing generated an average of 70 million (M) paired-end reads, ranging from 56 to 83 M in the 29 bulk soil samples. Adaptor sequences were removed first in the data cleanup. Low-quality reads were removed using a Q score of 20 (Q20, implying 99.0% of base call accuracy). As a result, an average of 69.5 M reads and 10.4 Gb pairs were retained as clean data. Further details of retailed DNA sequences are in Supplementary Table S1. 3.1. Soil microbial community diversity, composition and structure in the paddy field For the microbial communities in the 29 soil samples, the MG-RAST server predicted an average of 28.2 M protein features (23.3–34.9 M) and an average of 38,686 rRNA features (29,059–45,583). By aligning the sequencing data against several rRNA databases, MG-RAST identified an average of 13,161 rRNA features in the 29 samples. MG-RAST also identified 13.0 M protein features by searching for sequence similarity against several databases (KEGG, UniProt, eggnog, SEED) simultaneously. Details for each sample are in Supplementary Table S3. Among the identified rRNA features, 98.16% ( ± 0.35%) of the sequences were bacteria, 1.04% ( ± 0.31%) were Archaea, 0.59% were Eukaryotes, and < 0.02% were viruses. An average of 36% (26–50%) DNA sequences were unclassified. The taxonomic assignment was accepted at the family level. Among

Fig. 3. Soil pH influences soil microbial community diversity and composition at the sites. The fitted trend line in red assumes a linear relationship, blue lines indicate 95% confidence. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 5

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

Fig. 4. A. A unimodal correlation between pairwise differences of soil microbial communities and soil pH values. Both differences were measured as Euclidean distance. B. Pairwise difference of soil microbial communities and pairwise spatial distance between sampling sites. Blue markers are the pairwise distance between sites within a location, and orange markers are between sites among different locations. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3.2. Influence of soil properties on microbial community composition, structure and function Among the soil chemical properties and climatic factors, soil pH was the only significant element influencing microbial community alpha diversity (p = 0.0011), Dominance (p < 0.0001), and Shannon index of diversity (p < 0.0001) (Table 4). With an increase in soil pH from acidic to basic, soil microbial community diversity, as measured by alpha diversity, increased. Increasing soil pH resulted in increased dominance, suggesting some specific taxa became more dominant in alkaline soils. As a result, Shannon's index of diversity decreased in more basic soils (Fig. 3). The extent of the difference in soil pH also significantly influenced the similarity of species composition in the soil communities, and the relationship became unimodal (Fig. 4A). Spatial distance between the two sampling sites tended to influence similarity of species composition in the soil communities, but only within the same sampling location—the closer the two sampling sites were, the more similar their microbial species composition. However, the relationship between spatial distance and microbial composition disappeared between pairs of sites from different locations (Fig. 4B). In 23 of the 530 identified microbial families, their relative abundance in each soil microbial community was related to soil pH (Bonferroni corrected p < 0.05). The six families with an overall abundance > 0.5% are shown in Fig. S1. With the shifting soil pH, the relative abundance of those six families also changed, with some families (such as Planctomycetaceae and Phyllobacteriaceae) becoming more abundant with an increase in soil pH, while the opposite occurred for families such as Acidobacteriaceae, Hyrogenphilaceae, Beijerinckiaceae and Bradyrhizobiaceae (Fig. S1). The complete list of the 23 families is in Supplementary Table S4. The relative abundance of 32 families was linked to available P level in the soil, though most of those families had relatively lower relative abundance in the microbial communities to have an impact on overall community diversity and structure. The complete list of the 32 families is in Supplementary Table S4. To examine whether the pattern of correlation between soil microbial diversity and soil pH hold in microbial communities collected globally (with most datasets from North and South America), we correlated soil pH to the alpha diversity of microbial communities in our experimental soils with those archived in MG-RAST. The relationship between soil pH and microbial community diversity (alpha diversity) in global agricultural soils was significant, whether for the data from MGRAST alone (r = 0.304, p = 0.029) or combined with our data (r = 0.372, p = 0.0006) (Fig. 5). In other words, the influence of soil

Fig. 5. Influence of soil pH on alpha diversity of soil microbial communities across continents, with 52 samples from North and South America (soil pH and alpha diversity retrieved from MG-RAST on 30 December 2018), and 29 samples in this study. Pearson's correlation coefficient (r) and probability (p) are shown.

the identified taxa, Gemmatimonadaceae was the most abundant bacteria family in the soil microbial communities in our study, being an average of 3.02% of the 530 families identified, followed by Acidobacteriaceae (1.35%), Anaerolineaceae (1.32%), and Sphingomonadaceae (1.29%) (Table 2). The average abundance of each family in the 29 samples is in Supplementary. The analysis of similarity (ANOSIM) between soil microbial communities suggested that microbial communities from different locations differ significantly at the family level, except between Jin Tan and Yang Zhou that were only ~80 km apart (Table 3). The UPGMA clustering of species composition showed that most of the sites from the same location were clustered, with a few exceptions; for instance, sites from Jin Tan that were scattered in the PCA plot (Fig. 1). The PCA also identified that families differed significantly in their relative abundance (Fig. 1). Fig. 2 shows the relative abundance of the top ten families differing in their relative abundances. For instance, the relative abundance of Sphingomonadaceae was close to 41% (in the top ten families) at site JTNSR4, but only 1% at HFNSR2, while the abundance of Gallionellaceae had the opposite pattern. 6

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

Fig. 6. Principal component analysis plot showing differences between soil microbial communities from different continents with contrasting land-use history. Yellow dots indicate samples from France (rainfed dryland agriculture). The rainfed dryland may be irrigated, but not flooded over a prolonged period as the flooded paddies. Convex hull for each group of samples is shown. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

America (ANOSIM p = 0.578) regardless of the geographic distances > 7000 km between locations. The taxonomic composition of the North and South American groups differed from those in flooded paddy fields in China's YRB (ANOSIM p = 0.002 and p = 0.008, respectively). The distribution of samples from the North and South American groups was continuous in the PCA plot, despite most of the samples coming from two main locations in each continent (Boone in Iowa and Williston in North Dakota in North America, Sao Paulo and Amazonas in Brazil, see Table S1), suggesting the dominant role of agricultural practice (either flooded paddy practice or rainfed cropping) shaping the taxonomic composition of soil microbial communities at a continental scale. As support, data for two samples from France's agricultural dry cropland were retrieved from MG-RAST; the taxonomic composition of their soil microbial communities was similar to those in North and South America, as evident in the PCA plot (a formal ANOSIM was not performed due to the small number of samples). For three other samples from China's paddy fields (from Cai et al., 2014), one clustered with our current sample group, while the other two clustered with the North and South American groups (Fig. 6). The pattern of agricultural practice shaping the taxonomic composition of soil microbial communities at a continental scale was not evident for functional composition. The functional composition of soil microbial communities, as measured by the relative abundance of gene subsystems, differed between communities in rainfed dry croplands and flooded paddies (Fig. 7). Genes involved in Carbohydrate Metabolism, Nitrogen Metabolism, and Respiration differed significantly between the two agricultural systems. Within the North American group, the samples clustered into two groups, generally corresponding to the two major sampling locations (see Table S1), and a similar pattern was evident for the samples from Brazil (Fig. 7), suggesting additional environmental influences on the soil microbial community functional profile.

Table 4 Multiple regressions showing the effect (p-values) of soil properties and local climate factors on soil microbial community diversity and composition. Tmin was the average temperature in the coldest month in the sampling year (January 2017). Rainfall is the accumulated precipitation from 1 October 2016 to 30 June 2017. Bold indicates significant p-values.

pH Tmin (°C) Rainfall (mm) Available nitrogen (mg/kg) Available phosphorus (mg/ kg) Organic matter (g/kg)

Alpha diversity

Dominance index

Shannon index

0.0011 0.8881 0.4218 0.8321 0.5103

< 0.0001 0.4618 0.8621 0.4906 0.8237

< 0.0001 0.0841 0.0717 0.4501 0.2351

0.3941

0.8843

0.9070

pH on microbial community diversity is qualitatively similar across biomes separated by many thousands of kilometres in different continents. Soil pH also has a profound influence on gene function and life history pathways of microbial communities. Twenty-eight different functional subsystems were assigned to the 29 soil metagenomes. The relative abundance of eight categories was significantly related to soil pH (Table 5), but none were related to other soil property parameters or climatic factors. Key metabolic pathways associated with Amino Acids and Derivatives, and Iron Acquisition and Metabolism, were enriched as soil pH increased, while genes in biological processes related to Cell Division and Cell Cycle, and Cell Wall and Capsule, are more abundant in acidic soils. 3.3. Influence of agricultural practice on global soil microbial community Agricultural practice had a profound impact on the taxonomic composition of soil microbial communities under contrasting practices: flooded paddies and rainfed dryland (Fig. 6). The taxonomic composition (families with relative abundance > 0.01%) in North American dry croplands did not differ from those in Brazilian Amazonia in South

4. Discussion The metagenomic analysis in this study suggested that soil pH is a significant driver of microbial diversity and structure in croplands from 7

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

Fig. 7. Principal component analysis plot showing difference between gene subsystems in soil microbial communities from different continents with contrasting land-use history. Yellow dots indicate samples from France (rainfed dryland agriculture). The rainfed dryland may be irrigated, but not flooded over a prolonged period as the flooded paddies. Convex hull for each group of samples is shown. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

with comparable data—sampled from agriculture land, using wholegenome sequencing metagenomes, and analysed with the same pipeline—retrieved from the MG-RAST server, the significant correlation between soil pH and alpha diversity remained evident in the large dataset with 81 samples (with soil pH value available) from South and North America, and Asia (this study). Using either T-RFLP or 16s rRNA amplicon sequencing, previous studies suggested the importance of pH in controlling total microbial community structure in diverse soils across various spatial scales (Fierer and Jackson, 2006; Jangid et al., 2008; Lauber et al., 2009; Rousk et al., 2010). Rousk et al. (2010) revealed significant differences in soil bacterial community composition across a mere 180 m distance in a liming experiment, suggesting the dominance of pH in structuring bacterial communities at a local scale. At the continental level, Fierer and Jackson (2006) reported that differences in the diversity of soil bacterial communities across North and South America were attributed mainly to soil pH. In contrast, Faoro et al. (2010) reported no significant impact of pH on diversity in Brazilian Atlantic Forest using 16S rRNA gene sequences, which may have been due to the low sample size (10) and narrow pH range (3.7–4.4) in their analysis. Our study and those retrieved from MG-RAST used a much more powerful wholegenome sequencing metagenome approach that spanned a spatial scale from 10 m to thousands of kilometres across different continents with different environment and climates. Together with existing evidence, our analysis supports the notion that soil pH is a more significant driver of bacterial community diversity and composition than other environmental factors. The pattern is consistent both across biomes and within individual soil types, regardless of the spatial scale or methods used for taxonomy assignment. It has been hypothesised that soil pH may influence the composition of soil microbial communities by modifying enzyme activity through controlling the accessibility of nutrient and moisture by changing the ionisation balance in soil (Rousk et al., 2010). For example, Hu et al. (2013) reported that soil pH influenced the ionisation equilibrium of nitrate and ammonia in soils and consequently drove community composition and the biogeography of ammonia oxidisers. Indeed, our analysis of functional composition (gene subsystems) revealed that soil pH influenced a considerable proportion of gene subsystems that regulate metabolic and reproductive pathways in soil microbial communities. It is likely that most bacterial taxa have relatively narrow growth

Table 5 Effect of soil pH on gene subsystems in microbial communities. Bold indicates significance after Bonferroni correction (Bonferroni adjusted p-value was 0.0018 with significance at p < .05). Gene subsystems

p-Value

Coefficient

Amino acids and derivatives Carbohydrates Cell division and cell cycle Cell wall and capsule Clustering-based subsystems Cofactors, vitamins, prosthetic groups, pigments DNA metabolism Dormancy and sporulation Fatty acids, lipids and isoprenoids Iron acquisition and metabolism Membrane transport Metabolism of aromatic compounds Miscellaneous Motility and chemotaxis Nitrogen metabolism Nucleosides and nucleotides Phages, prophages, transposable elements, plasmids Phosphorus metabolism Photosynthesis Potassium metabolism Protein metabolism Regulation and cell signalling Respiration RNA metabolism Secondary metabolism Stress response Sulphur metabolism Virulence, disease and defence

0.0000 0.3452 0.0001 0.0001 0.5894 0.0295 0.7959 0.1449 0.0002 0.0000 0.0847 0.0004 0.3743 0.0000 0.8864 0.0001 0.5395 0.6516 0.0020 0.3501 0.4417 0.0059 0.0335 0.1795 0.0293 0.3037 0.3554 0.0253

0.7308 0.1818 −0.6529 −0.6639 −0.1045 0.4045 −0.0502 −0.2775 0.6456 0.6843 0.3256 0.6122 −0.1713 −0.7129 −0.0277 0.6631 0.1187 0.0875 0.5491 −0.1800 −0.1486 0.4987 −0.3959 −0.2563 0.4051 0.1977 −0.1780 −0.4146

China's Yangtze River Basin. The results indicate that soil microbial community diversity was lower in acidic soils than alkaline soils, when measured by alpha diversity. We, however, observed the opposite trend when diversity was estimated by Shannon's index of diversity. Increased soil pH may encourage the presence of more microbial taxa with low abundance. Meanwhile, a few common bacteria families are favoured in non-acidic soil with higher relative abundance than in acid soil. Both may contribute to the observed pattern of negative correlation between Shannon's index of diversity and soil pH. When incorporating our data 8

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

tolerances to pH in their environments, which is consistent with the narrow pH range (between 2 and 4 pH units) in the laboratory culture of many bacterial species (Rosso et al., 1995). Different bacterial strains may have different and narrow pH optima, which leads to the strong relationship between microbial community composition and soil pH. The pattern of soil pH and microbial community composition is stable and consistent across spatial scales and biomes (Fierer and Jackson, 2006; Lauber et al., 2009; Rousk et al., 2010; Ren et al., 2018). Soil pH could maintain a stable feedback loop that determines the fate of bacterial populations by either facilitating or inhibiting their growth (Ratzke and Gore, 2018). The stable feedback loop, therefore, maintains the stable and consistent soil pH–microbial community composition pattern. Indeed, we observed a significant shift in vital metabolic and reproductive pathways (abundance of gene subsystems) in response to shifts in soil pH in the microbial communities. We also observed changes in the relative abundances of specific taxonomic groups across pH gradient, as reported in other studies; for example, decreasing pH increased the relative abundance of Acidobacteria (Männistö et al., 2007; Jones et al., 2009; Dimitriu and Grayston, 2010) and we observed similar trends in Acidobacteriaceae, Hydrogenophilaceae, Beijerinckiaceae and Bradyrhizoiaceae. The relative abundances of Planctomycetaceae and Phyllobacteriaceae tended to be positively related to increasing pH across our sampling sites. How the change in diversity and abundance of these bacterial groups influences soil quality is not clear as the ecology and metabolism of these bacteria is not well understood (Kielak et al., 2016). However, Nannipieri et al. (2003) believed that soil tends to be redundant in functions and that a change in abundance of a group of taxa may have little effect on overall soil processes because other microbial groups can take on its function. The climatic factor played an important role in some biomes and had a significant and direct influence on soil microbial communities (Balser et al., 2010). Soil carbon and nitrogen availability also had strong positive associations with some soil microbial communities (Li et al., 2017; Ren et al., 2018). He et al. (2013) reported that nitrogen additions significantly reduced the diversity of a soil microbial community, while Yuan et al. (2017) observed that the abundance of some microbial families increased with high nitrogen supply in Alpine Tundra. We found that other soil properties (AP, AN and organic matter) and climatic factors (minimum temperature and rainfall) play a negligible role in affecting overall microbial community structure. Agricultural soils are commonly fertilised with nitrogen, and usually excessively. The effect of nitrogen on soil microbial communities may be saturated, and its impact was, therefore, not differentiated across the sampling sites. Studies have reported high abundances of proteobacterial groups due to higher availability of organic matter (soil carbon) (Axelrood et al., 2002; Fazi et al., 2005; Fierer et al., 2007). The influence of organic matter on soil microbial communities was not evident in our study. It is likely that the effect of organic matter could be saturated, as organic matter content is usually high in agricultural soil. The relative abundance of several bacteria families was linked to the level of available P in soil in our study. However, most of these families had relatively lower relative abundance in the communities; therefore, available P had little impact on the overall microbial community diversity and structure in our study. By retrieving comparable data from a public depository (MG-RAST) with contrasting land-use history, we revealed significant differences in taxonomic structure and functional composition between our samples (collected from flooded rice paddies) and those from primarily rainfed dry croplands in South America, North America and Europe. Land-use history has previously been shown to drive the diversity and structure of soil microbial communities (Tripathi et al., 2012; Suleiman et al., 2013; Goss-Souza et al., 2017). Agronomic land use and cultivation history may have induced shifts in both taxonomic composition and metabolic function (Carbonetto et al., 2014; Li et al., 2018), as observed here between farming systems. Rice cultivation in flooded paddies in China's YRB has a history of almost 10,000 years (Zong et al., 2007).

Such long and consistent agricultural land use must have created a persistent yet unique microbial community with a stable structure and function profile. The close community structure and composition of the two samples from Europe (France) to those from the North and South American communities suggest that similar agricultural land-use types may have resulted in a similar assembly of soil microbial community, though a formal analysis with more samples from Europe and other continents (Asia, Africa and Australia) is needed for a robust conclusion. Elsewhere, Breidenbach et al. (2017) reported different microbial community structures in soils for rainfed maize cultivation and continuous wetland cultivation for rice, which may be due to different environmental conditions, such as the availability of oxygen in the soil. Metagenomic analysis using shotgun sequencing identified some significant taxa groups in our study. Some taxa groups are noticeable, with relative abundances differing significantly between sites or being related to soil pH. Acidobacteriaceae was the second most abundant bacteria family in our samples, and its relative abundance was significantly associated with soil pH. It is known that soil pH controls the abundance of this family (Dedysh et al., 2006; Sait et al., 2006; Jones et al., 2009). All species of Acidobacteriaceae family are found in acidic environments. Indeed, it had a higher abundance in soil microbial communities from He Fei, where soils are acidic (pH < 7). Other families with relative abundances that differed among samples included Hydrogenophilaceae and Gallionellaceae. The presence of sulfides as electron donors and either molecular oxygen or nitrogen oxides as electron acceptors is a crucial factor for the presence and growth of members of Hydrogenophilaceae (Orlygsson and Kristjansson, 2014), though Lipson et al. (2015) reported that the abundance of Hydrogenophilaceae was related positively to the level of organic matter in arctic wet tundra soils. Geobacteraceae was abundant across the soil microbial communities in our study. Lovely et al. (2011) suggested that the elevated presence of members of Geobacteraceae may be related to Fe-reducing environments as a result of anthropogenic increases in organic matter. The high abundance of Geobacteraceae may stimulate the growth of other microbes through the release of soluble Fe(II) from insoluble Fe(III) into their environment (Röling, 2014), and improve nutrient accessibility by aboveground plants, which is beneficial in cropping systems. The family Gemmatimonadaceae, the most abundant family identified in our paddy field samples, made up an average of 3% of the soil bacterial communities, which corroborates the previous estimate that Gemmatimonadaceae is one of the top families found in soil (Fawaz, 2013). However, Gemmatimonadaceae was not related to soil pH in our study. Some studies have reported that members of Gemmatimonadaceae prefer soils near-neutral pH (Lauber et al., 2009), DeBruyn et al. (2011) found that soil moisture was the most significant factor constraining the abundance of Gemmatimonadaceae in a global biogeography analysis. Yuan et al. (2017) observed that the abundance of Gemmatimonadaceae responded positively to nitrogen supply, which was not seen in our study. It is likely that the effect of nitrogen is saturated as nitrogen fertilisation is usually common in croplands. Sphingomonadaceae was one of the most abundant bacterial families identified, with an average relative abundance of 1.29% in our study, and its relative abundance was not related to soil pH or other soil properties. The abundance of Sphingomonadaceae may be linked to the utilisation of small organic substances resulting from the degradation of humic substances (Hutalle-Schmelzer et al., 2010). Organic matter content is highly abundant in the croplands of our study system, and its influence on the abundance of Sphingomonadaceae may have been saturated. Anaerolineaceae is abundant in cropland soil, being the third most abundant with an average abundance of 1.32% in our study. Its relatively high abundance has been reported (Kirkegaard et al., 2016; McIlroy et al., 2017), while their physiology and ecology are less known. Most species of the Anaerolineaceae use carbohydrates and proteinaceous carbon sources under anaerobic conditions (Yamada et al., 2007; Sun et al., 2016). Habitats with abundant crop residues and 9

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

sometimes submerged soil (in rice paddies) may encourage the growth of members of the Anaerolineaceae.

database project (RDP-II): introducing myRDP space and quality controlled public data. Nucleic Acids Res. 35, D169–D172. DeBruyn, J.M., Nixon, L.T., Fawaz, M.N., Johnson, A.M., Radosevich, M., 2011. Global biogeography and quantitative seasonal dynamics of Gemmatimonadetes in soil. Appl. Environ. Microbiol. 15, AEM–05005. Dedysh, S.N., Pankratov, T.A., Belova, S.E., Kulichevskaya, I.S., Liesack, W., 2006. Phylogenetic analysis and in situ identification of bacteria community composition in an acidic Sphagnum peat bog. Appl. Environ. Microbiol. 72, 2110–2117. DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K., Huber, T., Dalevi, D., Hu, P., Andersen, G.L., 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72, 5069–5072. Dimitriu, P.A., Grayston, S.J., 2010. Relationship between soil properties and patterns of bacterial b-diversity across reclaimed and natural boreal forest soils. Microb. Ecol. 59, 563–573. Doran, J.W., 2002. Soil health and global sustainability: translating science into practice. Ag. Ecosyst. Environ. 88, 119–127. Escobar-Zepeda, A., Godoy-Lozano, E.E., Raggi, L., Segovia, L., Merino, E., Gutiérrez-Rios, R.M., Juarez, K., Licea-Navarro, A.F., Pardo-Lopez, L., Sanchez-Flores, A., 2018. Analysis of sequencing strategies and tools for taxonomic annotation: defining standards for progressive metagenomics. Sci. Rep. 8, 12034. Faoro, H., Alves, A.C., Souza, E.M., Rigo, L.U., Cruz, L.M., Al-Janabi, S.M., Monteiro, R.A., Baura, V.A., Pedrosa, F.O., 2010. Influence of soil characteristics on the diversity of bacteria in the Southern Brazilian Atlantic Forest. Appl. Environ. Microbiol. 76, 4744–4749. Fawaz, M., 2013. Revealing the Ecological Role of Gemmatimonadetes Through Cultivation and Molecular Analysis of Agricultural Soils. Master's Thesis. University of Tennessee. Fazi, S., Amalfitano, S., Pernthaler, J., Puddu, A., 2005. Bacterial communities associated with benthic organic matter in headwater stream microhabitats. Environ. Microbiol. 7, 1633–1640. Fierer, N., Jackson, R.B., 2006. The diversity and biogeography of soil bacterial communities. Proc. Natl. Acad. Sci. U. S. A. 103, 626–631. Fierer, N., Bradford, M.A., Jackson, R.B., 2007. Toward an ecological classification of soil bacteria. Ecology 88, 1354–1364. Fierer, N., Lauber, C.L., Ramirez, K.S., Zaneveld, J., Bradford, M.A., Knight, R., 2012. Comparative metagenomic, phylogenetic and physiological analyses of soil microbial communities across nitrogen gradients. ISME J 6, 1007–1017. Fierer, N., Ladau, J., Clemente, J.C., Leff, J.W., Owens, S.M., Pollard, K.S., Knight, R., Gilbert, J.A., McCulley, R.L., 2013. Reconstructing the microbial diversity and function of pre-agricultural tallgrass prairie soils in the United States. Science 342, 621–624. Garbeva, P., van Veen, J.A., van Elsas, J.D., 2004. Microbial diversity in soil: selection of microbial populations by plant and soil type and implications for disease suppressiveness. Annu. Rev. Phytopathol. 42, 243–270. Godfray, H.C., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F., Pretty, J., Robinson, S., Thomas, S.M., Toulmin, C., 2010. Food security: the challenge of feeding 9 billion people. Science 327, 812–818. Goss-Souza, D., Mendes, L.W., Borges, C.D., Baretta, D., Tsai, S.M., Rodrigues, J.L., 2017. Soil microbial community dynamics and assembly under long-term land use change. FEMS Microbiol. Ecol. 93, fix109. Grau, H.R., Aide, M., 2008. Globalization and land-use transitions in Latin America. Ecol. Soc. 13, 16. Hammer, O., Harper, D.A.T., Ryan, P.D., 2001. PAST: paleontological statistics software package for education and data analysis. Paleontologica Electronica 4, 9. He, Y., Qi, Y., Dong, Y., Xiao, S., Peng, Q., Liu, X., Sun, L., 2013. Effects of nitrogen fertilization on soil microbial biomass and community functional diversity in temperate grassland in Inner Mongolia, China. CLEAN–Soil. Air. Water 41, 1216–1221. Hu, H.W., Zhang, L.M., Dai, Y., Di, H.J., He, J.Z., 2013. pH-dependent distribution of soil ammonia oxidizers across a large geographical scale as revealed by high-throughput pyrosequencing. J. Soils Sediments 13, 1439–1449. Hutalle-Schmelzer, K.M., Zwirnmann, E., Krüger, A., Grossart, H.P., 2010. Enrichment and cultivation of pelagic bacteria from a humic lake using phenol and humic matter additions. FEMS Microbiol. Ecol. 72, 58–73. Jangid, K., Williams, M.A., Franzluebbers, A.J., Sanderlin, J.S., Reeves, J.H., Jenkins, M.B., Endale, D.M., Coleman, D.C., Whitman, W.B., 2008. Relative impacts of landuse, management intensity and fertilization upon soil microbial community structure in agricultural systems. Soil Biol. Biochem. 40, 2843–2853. Jones, R.T., Robeson, M.S., Lauber, C.L., Hamady, M., Knight, R., Fierer, N., 2009. A comprehensive survey of soil acidobacterial diversity using pyrosequencing and clone library analyses. ISME J 3, 442–453. Kielak, A.M., Barreto, C.C., Kowalchuk, G.A., van Veen, J.A., Kuramae, E.E., 2016. The ecology of Acidobacteria: moving beyond genes and genomes. Frontiers in Microbiol 7, 744. Kirkegaard, R.H., Dueholm, M.S., McIlroy, S.J., Nierychlo, M., Karst, S.M., Albertsen, M., Nielsen, P.H., 2016. Genomic insights into members of the candidate phylum Hyd2412 common in mesophilic anaerobic digesters. ISME J 10, 2352–2364. Lauber, C., Hamady, M., Knight, R., Fierer, N., 2009. Pyrosequencing based assessment of soil pH as a predictor of soil bacterial community structure at the continental scale. Appl. Environ. Microbiol. 75, 5111–5120. Leff, J.W., Bardgett, R.D., Wilkinson, A., Jackson, B.G., Pritchard, W.J., Long, J.R., Oakley, S., Mason, K.E., Ostle, N.J., Johnson, D., Baggs, E.M., 2018. Predicting the structure of soil communities from plant community taxonomy, phylogeny, and traits. The ISME J 9, 1. Leplae, R., Hebrant, A., Wodak, S.J., Toussaint, A., 2004. ACLAME: A CLAssification of Mobile Genetic Elements. Nucleic Acids Res. 32, D45–D49.

5. Conclusions and implications In this study, using shotgun sequencing and metagenomic analysis, we characterised soil microbial communities in croplands in China's major agricultural region in YRB. The taxonomic diversity of the microbial communities in this region was predominantly defined by soil pH, rather than by climatic factors. The significant correlation between soil pH and microbial community diversity was extended to soils from different continents, when datasets generated with similar methodology were retrieved from MG-RAST and incorporated into the correlation analysis. Together with available evidence, our findings support the notion that microbial biogeography is controlled primarily by edaphic variables, in particular soil pH, regardless of the spatial scale, which differs fundamentally from the biogeography of other macro-organisms. Our comparative analysis also revealed that agronomical land use and cultivation history might have induced shifts in both taxonomic composition and metabolic function. The results of our study deepen our understanding of the structure and functions of the soil microbial community in croplands (rice paddies) in China. The predominant element of soil property (pH) and the response of particular taxa to changing pH provide insight for the selection of agricultural practices. Agricultural practices and fertilisation could change soil pH and, therefore, alter soil microbial diversity and composition, which could have a potential impact on soil quality and nutrient supply. Supplementary data to this article can be found online at https:// doi.org/10.1016/j.apsoil.2019.103449. Declaration of competing interest The authors declare no conflict interest. Acknowledgments This work was supported by the National Key Research and Development Program of China (2016YFD0300107) and National Natural Science Foundation of China (31871578). We thank the support from Hubei Collaborative Innovation Center for Grain Industry. References Axelrood, P.E., Chow, M.L., Radomski, C.C., McDermott, J.M., Davies, J., 2002. Molecular characterization of bacterial diversity from British Columbia forest soils subjected to disturbance. Can. J. Microbiol. 48, 655–674. Balser, T.C., Gutknecht, J.L.M., Liang, C., 2010. How will climate change impact soil microbial communities? In: Dixon, G.R., Tilston, E. (Eds.), Soil Microbiology and Sustainable Crop Production. University of Reading Press, Reading, pp. 373–397. Bennett, E.M., Carpenter, S.R., Caraco, N.F., 2001. Human impact on erodable phosphorus and eutrophication: a global perspective: increasing accumulation of phosphorus in soil threatens rivers, lakes, and coastal oceans with eutrophication. BioScience 51, 227–234. Berg, G., Smalla, K., 2009. Plant species and soil type cooperatively shape the structure and function of microbial communities in the rhizosphere. FEMS Microbiol. Ecol. 68, 1–13. Bever, J.D., Platt, T.G., Morton, E.R., 2012. Microbial population and community dynamics on plant roots and their feedbacks on plant communities. Ann. Rev. Microbiol. 66, 265–283. Bissett, A., Brown, M.V., Siciliano, S.D., Thrall, P.H., 2013. Microbial community responses to anthropogenically induced environmental change: towards a systems approach. Ecol. Lett. 16, 128–139. Breidenbach, B., Brenzinger, K., Brandt, F.B., Blaser, M.B., Conrad, R., 2017. The effect of crop rotation between wetland rice and upland maize on the microbial communities associated with roots. Plant Soil 419, 435–445. Cai, Y., Wu, Y., Wang, S., Yan, X., Zhu, Y., Jia, Z., 2014. Microbial metabolism in typical flooded paddy soils. Acta Microbial. Sinica 54, 1033–1044. Carbonetto, B., Rascovan, N., Álvarez, R., Mentaberry, A., Vázquez, M.P., 2014. Structure, composition and metagenomic profile of soil microbiomes associated to agricultural land use and tillage systems in argentine pampas. PLoS One(6), e99949. Cole, J.R., Chai, B., Farris, R.J., Wang, Q., Kulam-Syed-Mohideen, A.S., McGarrell, D.M., Bandela, A.M., Cardenas, E., Garrity, G.M., Tiedje, J.M., 2006. The ribosomal

10

Applied Soil Ecology xxx (xxxx) xxxx

X. Wang, et al.

8, 5619. Röling, W.F.M., 2014. The family Geobacteraceae. In: Rosenberg, E., DeLong, E.F., Lory, S., Stackebrandt, E., Thompson, F. (Eds.), The Prokaryotes. Springer, Berlin, Heidelberg. Rosso, L., Lobry, J.R., Bajard, S., Flandrois, J.P., 1995. Convenient model to describe the combined effects of temperature and pH on microbial growth. Appl. Environ. Microbiol. 61, 610–616. Rousk, J., Bååth, E., Brookes, P.C., Lauber, C.L., Lozupone, C., Caporaso, J.G., Knight, R., Fierer, N., 2010. Soil bacterial and fungal communities across a pH gradient in an arable soil. ISME J 4, 1340–1351. Sait, M., Davis, K.E., Janssen, P.H., 2006. Effect of pH on isolation and distribution of members of subdivision 1 of the phylum Acidobacteria occurring in soil. Appl. Environ. Microbiol. 72, 1852–1857. Schmid, M.W., Hahl, T., van Moorsel, S.J., Wagg, C., De Deyn, G.B., Schmid, B., 2019. Feedbacks of plant identity and diversity on the diversity and community composition of rhizosphere microbiomes from a long-term biodiversity experiment. Mol. Ecol. 28, 863–878. Scholes, M.C., Scholes, R.J., 2013. Dust unto dust. Science 342, 565–566. Suleiman, A.K., Manoeli, L., Boldo, J.T., Pereira, M.G., Roesch, L.F., 2013. Shifts in soil bacterial community after eight years of land-use change. Syst. Appl. Microbiol. 36, 137–144. Sun, L., Toyonaga, M., Ohashi, A., Matsuura, N., Tourlousse, D.M., Meng, X.Y., Tamaki, H., Hanada, S., Cruz, R., Yamaguchi, T., Sekiguchi, Y., 2016. Isolation and characterization of Flexilinea floccule gen. nov., sp. nov., a filamentous anaerobic bacterium belonging to the class Anaerolineae in the phylum Chloroflexi. Int. J. Syst. Evol. Microbiol. 66, 988–996. Tripathi, B.M., Kim, M., Singh, D., Lee-Cruz, L., Lai-Hoe, A., Ainuddin, A.N., Go, R., Rahim, R.A., Husni, M.H., Chun, J., Adams, J.M., 2012. Tropical soil bacterial communities in Malaysia: pH dominates in the equatorial tropics too. Microb. Ecol. 64, 474–484. Urich, T., Lanzén, A., Qi, J., Huson, D.H., Schleper, C., Schuster, S.C., 2008. Simultaneous assessment of soil microbial community structure and function through analysis of the meta-transcriptome. PLoS One 3, e2527. Wu, J.X., Cheng, X., Xiao, H.S., Wang, H., Yang, L.Z., Ellis, E.C., 2009. Agricultural landscape change in China’s Yangtze Delta, 1942–2002: a case study. Agric. Ecosyst. Environ. 129, 523–533. Wuyts, J., Van de Peer, Y., Winkelmans, T., De Wachter, R., 2002. The European database on small subunit ribosomal RNA. Nucleic Acids Res. 30, 183–185. Yamada, T., Imachi, H., Ohashi, A., Harada, H., Hanada, S., Kamagata, Y., Sekiguchi, Y., 2007. Bellilinea caldifistulae gen. nov., sp. nov. and Longilinea arvoryzae gen. nov., sp. nov., strictly anaerobic, filamentous bacteria of the phylum Chloroflexi isolated from methanogenic propionate-degrading consortia. Int. J. Syst. Evol. Microbiol. 57, 2299–2306. Yan, Y., Wang, S., Chen, J., 2011. Spatial patterns of scale effect on specific sediment yield in the Yangtze River basin. Geomorphology 130, 29–39. Yuan, X., Knelman, J.E., Wang, D., Goebl, A., Gasarch, E., Seastedt, T.R., 2017. Patterns of soil bacterial richness and composition tied to plant richness, soil nitrogen, and soil acidity in alpine tundra. Arctic. Antarctic. Alpine Res. 49, 441–453. Zhalnina, K., Dias, R., de Quadros, P.D., Davis-Richardson, A., Camargo, F.A., Clark, I.M., McGrath, S.P., Hirsch, P.R., Triplett, E.W., 2015. Soil pH determines microbial diversity and composition in the park grass experiment. Microbial Ecol 69, 395–406. Zhang, K., Dearing, J.A., Dawson, T.P., Dong, X., Yang, X., Zhang, W., 2015. Poverty alleviation strategies in eastern China lead to critical ecological dynamics. Sci. Total Environ. 506, 164–181. Zong, Y.Q., Chen, Z., Innes, J.B., Chen, C., Wang, Z., Wang, H., 2007. Fire and flood management of coastal swamp enabled first rice paddy cultivation in east China. Nature 449, 459–462.

Li, R., Yu, C., Li, Y., Lam, T.W., Yiu, S.M., Kristiansen, K., Wang, J., 2009. SOAP2: an improved ultrafast tool for short read alignment. Bioinformatics 25, 1966–1967. Li, F., Chen, L., Zhang, J., Yin, J., Huang, S., 2017. Bacterial community structure after long-term organic and inorganic fertilization reveals important associations between soil nutrients and specific taxa involved in nutrient transformations. Front. Microbiol. 8, 187. Li, X., Jousset, A., de Boer, W., Carrión, V.J., Zhang, T., Wang, X., Kuramae, E.E., 2018. Legacy of land use history determines reprogramming of plant physiology by soil microbiome. ISME J 27, 1. Lipson, D.A., Raab, T.K., Parker, M., Kelley, S.T., Brislawn, C.J., Jansson, J., 2015. Changes in microbial communities along redox gradients in polygonized Arctic wet tundra soils. Environ. Microbiol. Rep. 7, 649–657. Lovely, D.R., Ueki, T., Zhang, T., Malvankar, N.S., Shrestha, P.M., Flanagan, K.A., Aklujkar, M., Butler, J.E., Giloteaux, L., Rotaru, A.E., Holmes, D.E., Franks, A.E., Orellana, R., Risso, C., Nevin, K.P., 2011. Geobacter: the microbe electric’s physiology, ecology, and practical applications. Adv. Microb. Physiol. 59, 1–100. Männistö, M.K., Tiirola, M., Häggblom, M.M., 2007. Bacterial communities in Arctic fields of Finnish Lapland are stable but highly pH dependent. FEMS Microbiol. Ecol. 59, 452–465. McIlroy, S.J., Kirkegaard, R.H., Dueholm, M.S., Fernando, E., Karst, S.M., Albertsen, M., Nielsen, P.H., 2017. Culture-independent analyses reveal novel Anaerolineaceae as abundant primary fermenters in anaerobic digesters treating waste activated sludge. Front. Microbiol. 8, 1134. Meyer, F., Paarmann, D., D'Souza, M., Olson, R., Glass, E.M., Kubal, M., Paczian, T., Rodriguez, A., Stevens, R., Wilke, A., Wilkening, J., Edwards, R.A., 2008. The metagenomics RAST server–a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 9, 386. Nannipieri, P., Ascher, J., Ceccherini, M., Landi, L., Pietramellara, G., Renella, G., 2003. Microbial diversity and soil functions. European J. Soil Sci. 54, 655–670. Orlygsson, J., Kristjansson, J.K., 2014. The family Hydrogenophilaceae. In: The Prokaryotes. Springer, Berlin, Heidelberg, pp. 859–868. Osler, G.H., Sommerkorn, M., 2007. Toward a complete soil C and N cycle: incorporating the soil fauna. Ecology 88, 1611–1621. Overbeek, R., Begley, T., Butler, R.M., Choudhuri, J.V., Chuang, H.Y., Cohoon, M., de Crécy-Lagard, V., Diaz, N., Disz, T., Edwards, R., Fonstein, M., Frank, E.D., Gerdes, S., Glass, E.M., Goesmann, A., Hanson, A., Iwata-Reuyl, D., Jensen, R., Jamshidi, N., Krause, L., Kubal, M., Larsen, N., Linke, B., McHardy, A.C., Meyer, F., Neuweger, H., Olsen, G., Olson, R., Osterman, A., Portnoy, V., Pusch, G.D., Rodionov, D.A., Rückert, C., Steiner, J., Stevens, R., Thiele, I., Vassieva, O., Ye, Y., Zagnitko, O., Vonstein, V., 2005. The subsystems approach to genome annotation and its use in the project to annotate 1000 genomes. Nucleic Acids Res. 33, 5691–5702. Peng, Y., Leung, H.C., Yiu, S.M., Chin, F.Y., 2012. IDBA-UD: a de novo assembler for single-cell and metagenomic sequencing data with highly uneven depth. Bioinformatics 28, 1420–1428. Poretsky, R., Rodriguez-R, L.M., Luo, C., Tsementzi, D., Konstantinidis, K.T., 2014. Strengths and limitations of 16S rRNA gene amplicon sequencing in revealing temporal microbial community dynamics. PLoS One 9, e93827. Qiao, S.Y., 2012. Experiment Instruction of Soil Physical and Chemical Properties. China University of Geosciences Press, Beijing. Rasche, F., Hödl, V., Poll, C., Kandeler, E., Gerzabek, M.H., van Elsas, J.D., Sessitsch, A., 2006. Rhizosphere bacteria affected by transgenic potatoes with antibacterial activities compared with the effects of soil, wild-type potatoes, vegetation stage and pathogen exposure. FEMS Microbiol. Ecol. 56, 219–235. Ratzke, C., Gore, J., 2018. Modifying and reacting to the environmental pH can drive bacterial interactions. PLoS Biol. 16, e2004248. Ren, B., Hu, Y., Chen, B., Zhang, Y., Thiele, J., Shi, R., Liu, M., Bu, R., 2018. Soil pH and plant diversity shape soil bacterial community structure in the active layer across the latitudinal gradients in continuous permafrost region of Northeastern China. Sci. Rep.

11