FEMS Microbiology Ecology, 93, 2017, fiw222

doi: 10.1093/femsec/fiw222
Advance Access Publication Date: 27 October 2016
Research Article

RESEARCH ARTICLE

Successive soybean-monoculture cropping assembles

rhizosphere microbial communities for the soil
suppression of soybean cyst nematode
M. Imran Hamid1,2, , Muzammil Hussain1,3, , Yunpeng Wu1 ,
Xiaoling Zhang1 , Meichun Xiang1, and Xingzhong Liu1
1
State Key Lab of Mycology, Institute of Microbiology, Chinese Academy of Sciences, No 3 Park 1, Beichen West
Rd., Chaoyang District, Beijing 100101, China, 2 Department of Plant Pathology, University College of
Agriculture, University of Sargodha, Sargodha, 40100, Pakistan and 3 University of Chinese Academy of
Sciences, Beijing 100039, China

Corresponding author: State Key Laboratory of Mycology, Institute of Microbiology, Chinese Academy of Sciences, No 3 Park 1, Beichen West Rd.,
Chaoyang District, Beijing 100101, China. Tel/Fax: +861064807512; E-mail: xiangmc@im.ac.cn

These authors contributed equally to this work.

One sentence summary: Soybean monoculturing engaged exclusive bacterial and fungal taxa in the rhizosphere microbiome for soil suppression
of Heterodera glycines.
Editor: Angela Sessitsch

ABSTRACT
One of the mechanisms of disease suppressiveness in soils is long-term monoculture (LTM) cropping to dissuade pathogen
infestation. However, the linkage between monoculturing and microbial community assemblage in the rhizosphere for
disease suppression remains unclear. To decipher this potential relationship, soil samples were collected from seven
locations in northeastern China, where LTM (638 yr) and short-term monoculture (STM 5 yr) cropping of soybean showed
varying degrees of soil suppressiveness to the soybean cyst nematode (SCN; Heterodera glycines). Using high-throughput
pyrosequencing to examine bacterial 16S rRNA and fungal ITS1 genes, we observed substantial variation in the species
richness and relative abundance of taxa in the rhizosphere across different sampling sites. At the genus level, the genera
Pseudomonas, Purpureocillium and Pochonia, which have been documented to suppress SCN in earlier studies, were much
more abundant in LTM soils than in STM soils. Moreover, the relative abundance of several bacterial and fungal genera with
metabolic, biocidal and parasitic activities was also monitored in the rhizosphere. In this study, we provide additional
evidence that plants shift the structural and functional composition of the rhizosphere microbiota to suppress pathogen
infection in LTM cropping soils.

Keywords: Monoculture; Heterodera glycines; rhizosphere; community composition; disease suppression

INTRODUCTION deposition of plant mucilage and root exudates directly influ-

ences the assemblage and activities of the rhizosphere micro-
Plants can shape the soil microbial community structure to
biota (Kent and Triplett 2002). The root exudate composition can
recruit special microbial groups in the rhizosphere against
vary between plant species, cultivars and developmental stages
pathogen infection (Cook et al. 1995; Mendes et al. 2011). The

Received: 12 June 2016; Accepted: 25 October 2016


C FEMS 2016. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

1
 2 FEMS Microbiology Ecology, 2017, Vol. 93, No. 1
(Mark et al. 2005; Cavaglieri, Orlando and Etcheverry 2009; DeAn-
gelis et al. 2009; Micallef, Shiaris and Colon-Carmona 2009). Sev-
eral studies have shown the involvement of the rhizosphere mi-
crobiota in the specific suppression of root pathogens in natural
disease suppressive soils (Weller et al. 2002; Mendes et al. 2011;
Cha et al. 2016; Chapelle et al. 2016). Specific suppression is gen-
erally developed during successive monoculture of a susceptible
cultivar in field soils infested with certain pathogens (Raaijmak-
ers and Mazzola 2016). However, an understanding of the im-
pact of monoculture cropping on the assembly of bacterial and
fungal communities in relation to disease suppression remains
unclear.
Plant parasitic nematodes affect the quality and quantity of
root exudates, which may influence rhizosphere microorgan-
isms (Kerry 2000). Altered root exudation may activate antag-
onistic microbes to establish a parasitic relationship with plant
parasitic nematodes. Although microbial antagonists parasitize
and protect plants against pathogen infections, both high abun-
dance and activity are needed for efficient antagonism (Bull,
Weller and Thomashow 1991; Raaijmakers et al. 1995; Berend-
sen, Pieterse and Bakker 2012). Wheat take-all disease is a clas-
sic example of an increase in the abundance of antagonis-
tic microbes during continuous wheat monoculture resulting
in an increase of 2,4-diacetylphloroglucinol-producing Pseu-
domonas spp., thereby contributing to the decline in take-all dis-
ease (Weller et al. 2002). Microbiome studies have indicated that
the mechanism of suppressiveness of Rhizoctonia root rot disease
during sugarbeet monoculture involves fluorescent pseudomon-
ads via the production of lipopeptide antibiotics (Mendes et al.
2011). A number of studies have revealed that natural disease
suppression does not reflect the presence of a single taxon but
rather the high abundance of several taxa or special functional
groups of microbes corresponding to disease suppressiveness
(Sanguin et al. 2009; Mendes et al. 2011; Chapelle et al. 2016).
The soybean cyst nematode (SCN; Heterodera glycines Ichi-
nohe) is an obligate sedentary parasite and accounts for sub-
stantial economic losses worldwide (Ma et al. 2005). The con-
tinuous planting of soybean in northeastern China, where SCN
occurs widely, has resulted in the development of a number
of suppressive soils (Sun and Liu 2000). In these soils, a de-
cline in SCN population densities is frequently observed when
cultivation is sustained for more than 8 years, and less SCN
eggs were recorded compared with conducive soils (Zhu et al.
2013). A similar phenomenon concerning the population de-
cline of H. schachtii has been reported in fields with the subse-
quent cropping of sugarbeet, demonstrating that antagonistic
microorganisms cause the suppressiveness observed in these
crops (Borneman and Becker 2007). So far, the culture-dependent
and culture-independent microbiome assessments of nematode
suppressive soils have shown several functional groups of bac-
teria and fungi that could potentially be involved in the suppres-
sion of nematodes in suppressive soils (Westphal and Becker
2001; Yin et al. 2003a,b; Borneman and Becker 2007; Xiang et al.
2010; Zhu et al. 2013; Adam et al. 2014). However, both bacterial
and fungal microbial communities have not been analyzed us-
ing next-generation sequencing of phylogenetic markers, and
the assembly of these microbial communities after nematode
suppression in monoculture field soils has not yet been investi-
gated. Based on previous studies, we hypothesized that the shift
of a soil from conducive to suppressive state after the continu-
ous monoculture of plants results in specific microbial changes
in the rhizosphere. Hence, the aim of this study was to inves-
tigate the impact of monoculture cropping on the assembly of
bacterial and fungal communities in the plant rhizosphere for
disease suppression. To this end, we conducted growth room
pot experiments with naturally infested SCN soils collected from
experimental soybean-monoculture fields located in northeast-
ern China to assess the impact of short-term monoculture (STM)
and long-term monoculture (LTM) cropping on the composition
of rhizospheric bacterial and fungal communities and SCN sup-
pression.
MATERIALS AND METHODS
Soil sample collection and physiochemical analyses
Soil samples were collected from short-term (5 years or less) and
long-term (6 years or more) soybean-monoculture fields, located
in Anda (AD), Baicheng (BC), Fulaerji (FL), Heihe (HH), Hailun
(HL), Sunwu (SW) and Wudalianchi (WD) counties, northeast-
ern China. The preferred sampling sites and fields within each
location were close enough to minimize climatic and edaphic
variations (Table S1, Supporting Information). The soils were col-
lected at a depth of 030 cm from 20 sites (targeted rows of
plants) across the field in a zigzag sampling pattern immediately
after soybean harvesting. Each soil sample was air-dried and
mixed thoroughly prior to passing through 2-mm pored mesh to
remove plant debris and stones and was then placed in a plastic
bag and transported to the laboratory in a cool box.
A portion (ca. 500 g) of each soil sample was applied for the
analysis of physical and chemical properties. The total N and
total P of the soils were measured using an AA3 HR AutoAna-
lyzer (SEAL Analytical, Mequon, WI, USA) and analyzed using
Windows-based AACE software according to the manufacturers
instructions. The remaining analyses of the physical and chem-
ical properties of each soil sample were conducted at Advanced
Standard Technical Services (Beijing, P.R. China).
SCN extraction and initial egg density determination
A subsample of 500 g was used to extract SCN cysts by decant-
ing the soil suspension through an 850-m-diameter aperture
sieve onto a 250 m aperture sieve, followed by sucrose floata-
tion and centrifugation. The SCN eggs were released from cysts
after breaking in a 5-ml glass tissue grinder and subsequently
suspended in a 37% sucrose solution after centrifugation at 2500
g for 5 min. The extracted eggs were collected on a 25-m
aperture sieve and counted in a 12-well tissue culture plate (Nest
Biotechnology) under an inverted microscope (Olympus CK40).
Growth chamber bioassay to assess disease
suppressiveness of soils
The soil suppressiveness was evaluated by adding the fresh SCN
eggs to the soil samples at 1500 eggs/100 g of dry soil, followed by
thorough mixing. A control treatment (CK) was designed by au-
toclaving the soil sample from the longest soybean-monoculture
field (BC-38yr) at 121 C for 1 hour prior to the inoculation of
healthy eggs. Two soybean seeds of the SCN susceptible variety
Sturdy were sown in PVC pots (11 cm in diameter and 9 cm in
height) containing 450 g dry soil with an initial moisture con-
tent of 10% (v/w), and the pots were covered with polyethy-
lene bags to retain moisture for seed germination. The plants
were grown in a growth chamber at 24 C 25 C with a photope-
riod of 16 h light/8 h dark. Three days after seed germination,
each pot was thinned to one seedling and watered at 2-day in-
tervals. Three pots (replicates) for each soil sample were pre-
pared, and all pots were distributed on the bench in a completely

randomized design. The soybean plants were harvested after 56
days to measure the SCN egg densities.
Collection of rhizosphere soil and extraction of DNA
Plant roots were gently collected from soil discarded from the
pot. After gently shaking to remove the loosely attached soil par-
ticles, the roots were vortexed in 50-ml tubes containing 0.1%
Tween 20 solution to wash off the tightly adhered rhizosphere
soil and facilitate soil descent for 5 min. The soil suspension in
the tube was centrifuged at 12 000g for 15 min after the roots
were carefully removed. The supernatant was discarded, and the
rhizosphere soil was collected and stored at 80 C for future DNA
extraction.
The collected rhizosphere soils were subjected to DNA ex-
traction. Three replicates of rhizosphere soil from each sample
were used to isolate DNA using the FastDNA Spin Kit for Soil
(MP Biomedicals) according to the manufacturers instructions.
The DNA concentrations were measured using a NanoVue plus
(Biochrom Ltd, UK).
Ultra-high-throughput barcoded amplicon sequencing
High-throughput sequencing was adopted to investigate bac-
terial and fungal richness, diversity and composition in
soybean-monoculture fields across seven locations in north-
eastern of China. The bacterial V4 region of the 16S rRNA
gene and the ITS1 region of the ITS locus were targeted
and PCR amplified using the individually bar-coded for-
ward primers V4-F, 5 -GTGCCAGCMGCCGCGGTAA-3 ; ITS1-F,
5 -CTTGGTCATTTAGAGGAAGTAA-3 and reverse primers V4-
R, 5 -GGACTACHVGGGTWTCTAAT -3 ; ITS2, 5 -GCTGCGTTC
TTCATCGATGC-3 . The DNA concentrations were adjusted to 20
ngl1 for each sample in the PCR reaction. Briefly, the PCR reac-
tion (50 l) contained 200 moles of dNTPs, 0.4 moles of each
gene-specific primer, 1X reaction buffer and 3 unit of Taq DNA
polymerase (TransGen, Beijing). The PCR amplifications were
performed in triplicate reactions on a Veriti Thermal Cycler (Ap-
plied Biosystems) under the following conditions: initial dena-
turing at 94 C for 5 min, followed by 35 cycles at 94 C for 50 s,
54 C for 30 s and 72 C for 40 s, with a final extension step at 72 C
for 10 min. Negative controls with no template DNA were also in-
cluded in the PCR amplification reactions. The amplified prod-
ucts were pooled for each monoculture year and subjected to gel
electrophoresis for purified amplicons using the EasyPure Quick
Gel Extraction Kit (TransGen, Beijing). The purified amplicons
were adjusted to the same concentration and products encod-
ing different barcodes were combined into a single library for se-
quencing on the MiSeq system. The pooled samples containing
different barcodes were loaded onto the MiSeq reagent cartridge,
and the cartridge was then loaded into the instrument along
with the flow cell. The automated cluster generation and paired-
end sequencing with a 13-cycle index read were performed with-
out any further user intervention (Novogene, China).
Analysis of 16S rRNA and ITS gene sequence
Primary data analysis (image analysis, base calling) was
performed on a MiSeq instrument. The sequences were
analyzed using the Quantitative Insights Into Microbial Ecol-
ogy (QIIME) toolkit with default parameters for each step,
as previously described (Caporaso et al. 2010). The bacterial
and fungal sequences at lengths shorter than 250 and 200
nucleotides, respectively, were removed. De novo and reference-
Hamid et al. 3
based chimera checking was performed, and sequences
characterized as chimeric were also removed. More than
0.47 million quality-filtered reads for bacteria and 1.1 million
quality-filtered reads for fungi were obtained for the rhizo-
spheric soil samples, with an average of 29 463 reads per
sample for bacteria (min = 11 540 and max = 36 919) and 72 729
reads per sample for fungi (min = 19,535 and max = 1,64,304).
The bacterial reads were clustered into OTUs at the 97%
sequence similarity level using the open-reference OTU picking
protocol in UPARSE-pipeline (Edgar 2013), and each cluster
was represented by its most abundant sequence. Taxonomic
configuration of bacterial OTUs was performed by searching
the Basic Local Alignment Search Tool for each representative
sequence against a subset of the Silva database (Quast et al.
2013). The fungal sequences were subjected to 97% sequence
similarity single-linkage clustering using GenBank (Benson et al.
2008) and UNITE (Abarenkov et al. 2010) as reference taxonomic
databases.
Data analysis
To understand the changes in bacterial and fungal communi-
ties during assembly in the rhizosphere after LTM and identify
significant changes in the relative abundance of key taxa, we
used custom R scripts executed using R v3.0.2 to calculate the
percentage of classified reads. Mothur v 1.34 was used to cal-
culate the alpha diversity within samples. Before analysis, the
samples were divided or multiplied to normalize the abundance
to the equivalent of 11 540 reads for bacteria and 19 535 reads for
fungi. Rarefaction analyses were performed separately for bac-
teria and fungi to discern the level of OTUs richness. Diversity
indices Shannon H and Dominance D were also calculated with
respect to sequence depth. Rarefaction curves were generated
using the R package vegan 2.2-1 to determine bacterial and fun-
gal OTUs richness within each sample.
Beta-diversity calculations were initially computed using
non-rarefied OTU counts. The R package Phyloseq with the
function ordinate was used to simultaneously calculate un-
weighted and weighted UniFrac dissimilarity matrices and the
principal coordinate analyses (PCoA). The proportions of total
variance explained by the PCs were recorded. We further con-
structed unweighted and weighted UniFrac dissimilarity ma-
trices in QIIME using the command Jackknifed beta diversity.py
on rarefied sequence counts (1000 in each sample). BrayCurtis
dissimilarity matrices for bacterial and fungal communities
were constructed based on the log2 -transformed OTU relative
abundance. Only OTUs with a relative abundance above 0.5%
in at least one sample were included in the analysis. Con-
strained PCoA were performed in R using the function capscale,
constrained by monoculture samples with different field en-
vironments. The significant difference between the ordination
and confidence intervals for the variance was determined us-
ing an ANOVA-like permutation test (functions permutest and
anova.cca) with 999 permutations. All the R functions used for
the beta diversity calculations were retrieved from the R Pack-
age vegan 2.2-1. A multivariate statistical analysis, including
multiple regression and correlation, was performed to under-
stand the relationship between SCN densities and soybean-
monoculture years. Significant differences between SCN densi-
ties across seven locations were evaluated using ANOVA, and
Duncans multiple range test was used to compare treatment
means. The treatments were considered significant when P <
0.05.
 4 FEMS Microbiology Ecology, 2017, Vol. 93, No. 1
Figure 1. Map of northeastern China with sampling sites and soybean cyst ne-
matode egg densities. (A) Geographical locations of sampling region with differ-
ent soybean-monoculture history. The different colors of map represent three
provinces of northeastern China. (B) The soybean cyst nematode egg densities
were recorded at 56 dpi after greenhouse pot experiment (n = 3). The soil sam-
ples from different locations with STM and LTM year (yr) were placed on X-axis
and SCN egg densities on Y-axis. Different letters on the bars show significant
differences among soil samples (P < 0.05).
RESULTS
Bioassay of disease suppression in the soil samples
One or two soil samples from LTM fields (638 yr) and STM fields
(5 yr or less) in each of seven different locations in northeast-
ern China were collected (Fig. 1A). The bioassay of the soil sup-
pressiveness against Heterodera glycines in growth chamber pot
experiments showed that the control treatment (CK), involving
the autoclaving of LTM soil (BC-38yr), significantly increased SCN
egg densities (4010 467.65 eggs/100 g of dry soil) compared
with other soil treatments. The SCN egg densities in all of the
LTM soil samples were significantly suppressed and ranged from
26.66 8.81 in AD-13yr to 650 45.82 in HL-8yr in the 100-g
dry soil samples. In contrast, STM field soils resulted in rela-
tively higher egg densities (P < 0.05), ranging from 220 41.63
to 2380 254.03 in 100 g of dry soil (Fig. 1B). Overall, in five out
of the seven locations, LTM soils had a significantly lower num-
ber of SCN eggs compared with STM soils. In a previous study,
we demonstrated that the turnaround time for the induction of
soil suppressiveness against SCN was 8 years, and soils with <8
years of soybean-monocultures were considered to be STM (Zhu
et al. 2013). However, the turnaround period for SCN suppres-
sion was 6 to 8 years in this study, suggesting a slight influence
of the localities. The regression analysis also indicated that egg
densities were significantly reduced from STM to LTM (Fig. S1,
Supporting Information). These results indicated that the sup-
pressiveness of successive monoculture soils against H. glycines
has a microbial basis because the suppressive activity was lost
after autoclaving.
A soil physiochemical analysis was also performed to under-
stand the potential role of abiotic factors in the soil suppres-
sion of SCN. However, there were no significant differences in
the physiochemical properties between soil samples of the same
location for short- and long-term monoculturing (Table S2, Sup-
porting Information), indicating that soil abiotic factors are not
involved in the soil suppression of SCN.
Impact of soybean monoculture on the assembly of
bacterial communities
The rhizospheric bacterial communities were analyzed by py-
rosequencing the 16S amplicon, recovering 471 419 high-quality
16S rRNA sequence reads. These reads were clustered into 3703
OTUs at 97% sequence similarity and used to calculate phy-
lum level abundance within kingdom bacteria. This analysis
revealed that assembly of the rhizosphere bacterial communi-
ties was markedly affected by the LTM and STM of soybean. To
compare the alpha-diversity (within-sample diversity) of sam-
ples with uneven sequence counts, we rarefied the data using
Mothur, and rarefaction curves were generated to observe the
OTU richness in each sample. The results showed that LTM crop-
ping decreased alpha diversity in assembled bacterial commu-
nities, except for the AD-13yr sample (Fig. 2A and Table S3, Sup-
porting Information). In addition, the rhizosphere of the soy-
bean plant grown in autoclaved soil (CK; BC-38yr) showed higher
species richness than in LTM soil (BC-38yr). We also estimated
beta-diversity using unweighted (based on presence or absence
of taxa) and weighted (based on abundance of taxa) UniFrac dis-
tance metrics and visualized PCoA to determine variation be-
tween rhizobacterial communities of all soil samples from the
seven locations. Overall, the results showed little variation, but
no significant differences were observed in the rhizobacterial
communities in LTM and STM soils samples measured by un-
weighted UniFrac and weighted Unifrac (Fig. 2B and C), respec-
tively. CAP analyses revealed that different locations with mono-
culture samples explain 54% of the overall variance of the data
based on BrayCurtis metrics (Fig. 2D), and the constrained or-
dination showed the clustering of the samples with respect to
closed geographical locations (Fig. 2D and Fig. S2).
Bacterial taxa involved in nematode suppression
To further our understanding of the impact of successive mono-
culture cropping on the assembly of bacterial communities and
disease suppression, rhizobacterial taxa with increased relative
abundance after LTM cropping were detected. In all treatments,
the soybean rhizosphere was dominated by phylum Proteobac-
teria, followed by Actinobacteria and Bacteroidetes (Fig. 3A and
Table S4, Supporting Information). The high abundance of Pro-
teobacteria in rhizosphere largely reflects the abundance of Al-
phaproteobacteria, followed by Gammaproteobacteria and Be-
taproteobacteria. Compared with STM samples, the abundance
of Alphaproteobacteria, Betaproteobacteria and Actinobacteria
in all samples significantly varied, while Gammaproteobacteria
showed an increase in relative abundance in LTM soils (Fig. 3B
and Table S5, Supporting Information). At the family level, the
abundance of Pseudomonadaceae gradually increased with in-
creasing years of soybean monoculture at each location (Fig. 3C
 Hamid et al. 5

Figure 2. Bacterial richness and beta diversity of STM and LTM samples collected from seven locations in northeastern China. (A) Rarefaction curves shown bacterial
diversity within each sample based on normalized pyrosequencing reads. (B) PCoA of unweighted UniFrac by monoculture samples revealed variation in rhizospheric
bacterial microbiota among different year of soybean monoculture soils. The percentage variation explained by the PCs is indicated on axes. (C) PCoA of weighted
UniFrac by monoculture samples revealed variation in rhizospheric bacterial microbiota among different year of soybean monoculture soils. The percentage variation
explained by the PCs is indicated on axes. (D) Constrained PCoA on the rhizospheric bacterial microbiota. Variation in BrayCurtis distance by monoculture samples
within all field environments explained 54% of the overall variance (P < 0.003). The percent variation explained by each axis refers to the fraction of the total variance of
the data explained by monoculture samples by field. In all panels, circles correspond to STM samples and triangles to LTM samples. The samples within each location
are mark with same color.

and Table S6, Supporting Information). Specifically, the abun-
dance of the genus Pseudomonas was significantly higher in rhi-
zosphere of soybean seedlings in LTM soils than that in STM soils
and CK (Fig. 3D and Table S7, Supporting Information). These
results revealed that LTM cropping increases the abundance of
special bacterial taxa in the rhizosphere that may be involved in
the soil suppression of H. glycines.
Subsequent analyses focused on other microbial genera with
biocidal, metabolic and parasitic capabilities. Bacterial genera,
such as Burkholderia, Chitinophaga and Streptomyces, have antago-
nistic traits (El-Banna and Winkelmann 1998; Garbeva et al. 2011)
and were also enriched in the rhizosphere of soybean in LTM
soils of AD-13yr and SW-15yr but decreased in FL-30yr and WD-
6yr. The genera Burkholderia, Chitinophaga and Streptomyces also
showed an increased relative abundance in HH-10yr, BC-38yr
and HL-8yr, respectively (Fig. 3D and Table S7). This result in-
dicated that some key taxa may respond to SCN suppression in
certain locations, except taxa such as Pseudomonas, which were
commonly involved in all samples.
Assembly of fungal communities in successive soybean
monoculturing
The fungal communities in the soybean rhizosphere in LTM and
STM soils were investigated. A total of 1 163 674 high-quality ITS
sequencing reads were recovered and clustered into 1439 OTUs
at 3% distance sequence dissimilarity. The samples with uneven
sequence counts were initially rarefied using Mothur to compare
alpha-diversity of the samples. Rarefaction analysis showed that
samples with successive monoculture cropping for more than
 6 FEMS Microbiology Ecology, 2017, Vol. 93, No. 1
Figure 3. Relative abundance of bacterial taxa in the rhizosphere of soybean growing in STM and LTM soil samples. (A) Relative abundance of bacterial phyla. (B)
Relative abundance of bacterial classes. (C) Relative abundance of the dominant bacterial families that showed variation between STM and LTM samples within each
location and across location. (D) Relative abundance of the dominant bacterial genera. The same colors in each bar across different locations are representing same
taxa.
13 years cause a decrease of alpha diversity in the assembled
fungal communities in the rhizosphere (Fig. 4A and Table S3).
Generally, the fungal taxa were fewer in the rhizosphere of
plants in LTM soils compared with STM soils. We further used
the unweighted and weighted UniFrac distance metrics to es-
timate beta-diversity. The variation between the rhizospheric
fungal communities of LTM and STM samples was visualized
by generating PCoA. The results of these analyses showed lit-
tle variation of fungal communities along PCoA 1 and PCoA
2 between LTM and STM samples measured by unweighted
UniFrac. However, there was no significant difference observed
between STM and LTM. Interestingly, we observed the cluster-
ing of LTM samples from three locations AD-13yr, BC-38yr and
FL-30yr (Fig. 4B), despite differences in soil composition (Table
S2), implying that the shared climate likely drives this pattern.
The weighted UniFrac distance metric also derived a similar pat-
tern and divided the samples into two distinct clusters based
on monoculture samples and geographical location (Fig. 4C).
We also used the canonical analysis of principal coordinates
to quantify the influence of geographical locations on mono-
culture samples. CAP analysis revealed that different locations
with monoculture samples explained 56% of the overall variance
of the data based on BrayCurtis metrics (Fig. 4D), 47% with the
unweighted UniFrac and 55% with weighted UniFrac dissimilar-
ity matrices (Fig. S3). These results indicated that monoculture
cropping and geographic heterogeneity potentially interact to
shape the fungal communities at the rootsoil interface.
The fungal taxa contributing to soil suppressiveness
against Heterodera glycines
To obtain a more in-depth understanding of the soybean-
monoculture cropping on assembly of fungal communities in
the rhizosphere microbiome, we focused on changes in the
abundance of rhizospheric mycotaxa after LTM. The overall phy-
logenetic distribution of the fungal communities across differ-
ent samples was associated with the dominance of phylum As-
comycota, followed by Zygomycota, Basidiomycota and Chytrid-
iomycota (Fig. 5A and Table S4). Although the fungal commu-
nity was dominated by members of Sordariomycetes, Euro-
tiomycetes, Agaricomycetes and Leotiomycetes at the class level
(Fig. 5B and Table S5), considerable variation in their compo-
sition across entire samples was observed at the family level
(Fig. 5C and Table S6). Among the dominant fungal genera, the
relative abundance of nematophagous fungi in the genera of
Purpureocillium, Pochonia and Clonostachys in the soybean rhizo-
sphere generally increased in the LTM soils (Fig. 5D and Table
S7). However, the dominant fungi were not detected in all geo-
graphical locations. The genus Purpureocillium was highly abun-
dant in LTM samples, except for WD, where the Purpureocil-
lium was not detected in STM and LTM soil samples. The rel-
ative abundance of Pochonia gradually increased in the north
LTM soils of HH-10yr, SW-15yr and WD-6yr, and Clonostachys was
gradually increased in their relative abundance in BC-38yr, HH-
10yr, SW-15yr and WD-6yr. In addition, the relative abundance
of Mortierella gradually decreased in LTM compared with STM
soils, except BC-38yr and HL-8yr, and the abundance of Fusar-
ium showed significant variation across sampling sites among
the dominant fungal genera (Fig. 5D and Table S7). These re-
sults suggest that successive monoculture cropping and envi-
ronmental heterogeneity are the main driving factors to exploit
the functional fungal taxa in soybean rhizosphere to induce soil
suppressiveness.
Current results revealed that some special taxa, such as
Pseudomonas, Purpureocillium, Pochonia and Clonostachys, were
gradually increased in abundance with monoculture years of
soybean in most of the soil samples tested and should be
the core microbes commonly responding to SCN suppression.
 Hamid et al. 7

Figure 4. Fungal richness and beta diversity of STM and LTM samples collected from seven locations in northeastern China. (A) Rarefaction curves shown fungal
diversity within each sample based on normalized pyrosequencing reads. (B) PCoA of unweighted UniFrac by monoculture samples revealed variation in rhizospheric
fungal microbiota among different year of soybean monoculture soils. The percentage variation explained by the PCs is indicated on axes. (C) PCoA of weighted UniFrac
by monoculture samples revealed variation in rhizospheric fungal microbiota among different year of soybean monoculture soils. The percentage variation explained
by the PCs is indicated on axes. (D) Constrained PCoA on the rhizospheric fungal microbiota. Variation in BrayCurtis distance by monoculture samples within all field
environments explained 56% of the overall variance (P < 0.001). The percent variation explained by each axis refers to the fraction of the total variance of the data
explained by monoculture samples by field. In all panels, circles correspond to STM samples and triangles to LTM samples. The samples within each location are mark
with same color.

However, some key taxa, such as Fusarium, Mortierella, Burkholde-
ria, Chitinophaga and Streptomyces, with biocontrol functions in
certain suppressive soils, are also a prerequisite for the induc-
tion of soil suppressiveness. These data also indicated that bac-
terial and fungal communities are jointly involved in the soil
suppression of SCN in the LTM cropping system.
DISCUSSION
Mechanisms underlying the development of disease suppres-
sive soils by continuous monoculture are complex. The decline
of nematode Heterodera avenae was observed in the UK (Colling-
wood 1962) and throughout northern Europe after the succes-
sive monoculture of wheat, and microbial taxa such as Verti-
cillium chlamydosporium, Catenaria anguillulae and Nematophthora
gynophila were considered causative agents (Kerry, Crump and
Mullen 1982; Kerry and Crump 1998). Over the last decade, it has
been documented that disease suppressive soils are of biological
origin in nature, and some dominant microbial taxa are involved
in this suppression (Weller et al. 2002; Mazzola 2007). Recent
microbiome studies on disease suppressive soils against fungal
root pathogens have also demonstrated that special microbial
groups are also involved in this suppression (Mendes et al. 2011;
Cha et al. 2016; Chapelle et al. 2016). The results of this study
indicated that some core microbes, such as Purpureocillium lilac-
inum and Pseudomonas spp., consistently corresponded to SCN
suppression, while other key microbes, such as species in Pocho-
nia, Clonostachys, Fusarium, Burkholderia, Chitinophaga and Strep-
tomyces, were dominant in some suppressive soils. The involve-
ment of case-dependent dominant microbes in suppressive
 8 FEMS Microbiology Ecology, 2017, Vol. 93, No. 1
Figure 5. Relative abundance of fungal taxa in the rhizosphere of soybean growing in STM and LTM soil samples. (A) Relative abundance of fungal phyla. (B) Relative
abundance of fungal classes. (C) Relative abundance of the dominant fungal families (D) Relative abundance of the dominant fungal genera that showed variation
between STM and LTM samples within each location and across location. The same colors in each bar across different locations are representing same taxa.
soils may well explain why the turnaround period for the de-
velopment of nematode suppression is also case dependent. Al-
though Zhu et al. (2013) reported that the turnaround period is
8 years for SCN suppressive soils, the egg densities of SCN in
soils with 6 years of monoculture history (WD-6yr) were not
significantly different from those of LTM soils (BC-38yr and FL-
30yr); thus, we proposed 6 to 8 years for the turnaround period.
Overwhelming scientific evidence has shown that soil micro-
biota in successive monoculturing cropping systems are contin-
uously exposed to the roots of host plants with natural capabil-
ity to recruit certain microbial groups for protection against soil-
borne pathogens (Cook 2006). However, the impact of successive
monoculture cropping on assembly of rhizospheric microbial
communities and disease suppression is not well understood.
The results obtained from soil samples under monoculture for
different lengths of time in seven locations indicated that soy-
bean monoculturing can constantly exploit specific functional
taxa for protection of roots against nematode infection, and the
soybean rhizosphere and root exudates may have more influ-
ence on bacteria than fungi. However, the differences in func-
tional fungal taxa that respond to suppressiveness across var-
ious locations suggested that geography also plays a key role,
such as the preference of Purpureocillium in southern LTM soils
and Pochonia in northern LTM soils. It has previously been re-
ported that P. chlamydospora prefers a lower temperature than P.
lilacinum (Manzanilla et al. 2013). Moreover, variations across ge-
ographic patterns have also been observed to shape the micro-
biota of the maize rhizosphere and their interactions in nature
(Green and Bohannan 2006; Berg and Smalla 2009). Overall, the
microbes responding to soil suppressiveness could be different
across different geographical locations.
Although Actinobacteria and Bacteroidetes were the dom-
inant phyla detected in the soybean rhizosphere in all soil
samples tested, the phylum Proteobacteria was the most domi-
nant. In particular, the class Gammaproteobacteria in this phy-
lum has shown an increased abundance in LTM soils compared
with STM soils. Further analysis showed that Pseudomonas in
the Pseudomonadaceae family was the only dominant genus in
the rhizosphere that showed a gradual increase in abundance
in LTM samples across all locations. The contribution of antago-
nistic pseudomonads was declared important against soil-borne
plant pathogens (Mazzola et al. 1992; Chapon et al. 2002; Weller
et al. 2007; Sanguin et al. 2009), and Pseudomonadaceae has also
been implicated in other plant disease suppressive soils (Mendes
et al. 2011), suggesting that some plants secrete similar compo-
nents to enhance this type of bacteria for root protection. How-
ever, Burkholderia, Chitinophaga, Streptomyces and Variovorax were
also highly abundant in the rhizosphere of soybean in some LTM
soils, suggesting that soybean roots selectively enhance certain
bacteria for root protection at different locations. These taxa
produce diverse metabolites, exhibit biocidal activities and in-
duce antagonistic traits in other bacterial taxa (El-Banna and
Winkelmann 1998; Garbeva et al. 2011; Han et al. 2011). In addi-
tion, Pasteuria, an obligate parasite of plant parasitic nematodes,
has been associated with SCN suppression in the USA as well as
eastern and southeastern China (Sayre et al. 1991; Atibalentja
et al. 1998; Ma et al. 2005). We also detected this parasite in the
LTM soils of AD-13yr, FL-30yr and WD-6yr. The contribution of
juvenile parasites to SCN suppressiveness should be specifically
investigated in the future.
The rhizosphere microbiota has been extensively studied in
different disease suppressive soils, primarily focusing on bacte-
rial communities. However, as one of the most important group
of soil microbial community, fungal communities have been
overlooked. The animal nature of plant-parasitic nematodes is
different from pathogenic fungi and bacteria. Nematophagous

fungi have been extensively investigated for several decades as
biological control agents of nematodes (Barron 1977). Genera
such as Purpureocillium, Pochonia and Clonostachys are the most
encountered fungi associated with eggs, females and cysts of
sedentary nematodes, and Pochonia chlamydosporia is one of the
fungi that contribute to the decline of cereal cyst nematode in
Europe (Kerry 1975). Fungi were an essential part in soil sup-
pressiveness against SCN in this study. The distinct variations in
fungal communities along LTM and STM and across geograph-
ical locations suggest that fungi may play more functions in
the suppressiveness than bacterial community. The high rela-
tive abundance of nematophagous fungal genus Purpureocillium,
Pochonia and Clonostachys was observed in the rhizosphere of
soybean plants. These functional taxa may accumulate biomass
in the rhizosphere, reflecting the increased parasitism rate with
increasing monoculture years. The fungus Purpureocillium is a
promising biological control agent that effectively parasitizes
the eggs of the plant parasitic nematodes and is occasionally as-
sociated with female of nematodes (Prasad, Varshney and Adho-
leya 2015). The genus Pochonia, particularly the species Pochonia
chlamydosporia, is a promising antagonist of cyst and root knot
nematodes (Sung et al. 2007; Manzanilla et al. 2013), and this
species parasitizes nematode eggs (Gine et al. 2016). The genera
Clonostachys and Fusarium have also been implicated in the sup-
pression of plant parasitic nematodes (Zhang et al. 2008; Waweru
et al. 2014). We also detected Hirsutella spp. across all locations in
this study. The species of this genus are the dominant biocontrol
agents of H. glycines (Liu and Chen 2001; Hussain et al. 2016; Wang
et al. 2016) and were considered to respond to the SCN suppres-
sive soils because they have been detected in the soybean mono-
culture fields in northeastern China (Xiang et al. 2010). Overall,
the fungal community is one of the most important contribu-
tors to nematode suppressive soils, except the bacterial com-
munity, which is more important in fungal disease suppressive
soils.
In conclusion, successive monoculture clearly affected the
assembly of bacterial and fungal communities in the rhizo-
sphere for disease suppression. The variations in microbial com-
munity structure range from STM to LTM, and the increased
abundance of exclusive microbial taxa in successive monocul-
ture systems provides a detailed description of the develop-
ment of suppressive soils against SCN. These findings clarify the
basic concepts that SCN suppressive soils developed with
successive monoculture systems that may be a complex
community-based phenomenon. However, future studies are re-
quired to understand the functional attributes of these microbial
communities and develop synthetic microbial communities for
biological control of plant parasitic nematodes.
SUPPLEMENTARY DATA
Supplementary data are available at FEMSEC online.
ACKNOWLEDGEMENTS
The authors would like to thank the two anonymous review-
ers for their valuable suggestions for improving the manuscript.
MIH would like to thank the Chinese Academy of Sciences for
the award of CAS Fellowship for Postdoctoral and Visiting Schol-
ars from Developing Countries, and MH would like to thank the
CAS-TWAS for the PhD Fellowship.
Hamid et al. 9
FUNDING
This research was financially supported by a grant from the Na-
tional Key Basic Research Program of China (Program no. 973,
Grant No. 2013CB127506).
Conflict of interest. None declared.
REFERENCES
Abarenkov K, Henrik Nilsson R, Larsson KH et al. The
UNITE database for molecular identification of fungirecent
updates and future perspectives. New Phytol 2010;186:
2815.
Adam M, Westphal A, Hallmann J et al. Specific Microbial Attach-
ment to Root Knot Nematodes in Suppressive Soil. Appl Env-
iron Microb 2014;80:267986.
Atibalentja N, Noel GR, Liao TF et al. Population changes in Het-
erodera glycines and its bacterial parasite Pasteuria sp. in nat-
urally infested soil. J Nematol 1998;30:8192.
Barron GL. The Nematode-destroying Fungi. Topics in Mycobiol-
ogy No. 1. Guelph:Canadian Biological Publications Ltd., 1977,
140.
Benson DA, Karsch-Mizrachi I, Lipman DJ et al. GenBank. Nucleic
Acids Res 2008;36:2530.
Berendsen RL, Pieterse CM, Bakker PA. The rhizosphere micro-
biome and plant health. Trends Plant Sci 2012;17:47886.
Berg G, Smalla K. Plant species and soil type cooperatively shape
the structure and function of microbial communities in the
rhizosphere. FEMS Microbiol Ecol 2009;68:113.
Borneman J, Becker JO. Identifying microorganisms involved
in specific pathogen suppression in soil. Phytopathology
2007;45:153.
Bull CT, Weller D, Thomashow LS. Relationship between root
colonization and suppression of Gaeumannomyces graminis
var. tritici by Pseudomonas fluorescens strain 2-79. Phytopathol-
ogy 1991;81:9549.
Caporaso JG, Kuczynski J, Stombaugh J et al. QIIME allows anal-
ysis of high-throughput community sequencing data. Nat
Methods 2010;7:3356.
Cavaglieri L, Orlando J, Etcheverry M. Rhizosphere microbial
community structure at different maize plant growth stages
and root locations. Microbiol Res 2009;164:3919.
Cha J-Y, Han S, Hong H-J et al. Microbial and biochemical
basis of a Fusarium wilt-suppressive soil. ISME J 2016;10:
11929.
Chapelle E, Mendes R, Bakker PAH et al. Fungal invasion of the
rhizosphere microbiome. ISME J 2016;10:2658.
Chapon A, Guillerm A-Y, Delalande L et al. Dominant colonisa-
tion of wheat roots by Pseudomonas fluorescens Pf29A and se-
lection of the indigenous microflora in the presence of the
take-all fungus. Eur J Plant Pathol 2002;108:44959.
Collingwood C. Continuous corn growing and cereal root eel-
worm in the South-West. NAAS Q Rev 1962;58:70.
Cook RJ. Toward cropping systems that enhance productivity
and sustainability. P Natl Acad Sci USA 2006;103:1838994.
Cook RJ, Thomashow LS, Weller DM et al. Molecular mechanisms
of defense by rhizobacteria against root disease. P Natl Acad
Sci USA 1995;92:4197201.
DeAngelis KM, Brodie EL, DeSantis TZ et al. Selective progressive
response of soil microbial community to wild oat roots. ISME
J 2009;3:16878.
Edgar RC. UPARSE: highly accurate OTU sequences from micro-
bial amplicon reads. Nat Methods 2013;10:9968.
 10 FEMS Microbiology Ecology, 2017, Vol. 93, No. 1
El-Banna N, Winkelmann G. Pyrrolnitrin from Burkholderia cepa-
cia: antibiotic activity against fungi and novel activities
against streptomycetes. J Appl Microbiol 1998;85:6978.
Garbeva P, Silby MW, Raaijmakers JM et al. Transcriptional
and antagonistic responses of Pseudomonas fluorescens Pf0-
1 to phylogenetically different bacterial competitors. ISME J
2011;5:97385.
Gine A, Carrasquilla M, Martnez-Alonso M et al. Characteri-
zation of soil suppressiveness to root-knot nematodes in
organic horticulture in plastic greenhouse. Front Plant Sci
2016;7:164.
Green J, Bohannan BJ. Spatial scaling of microbial biodiversity.
Trends Ecol Evol 2006;21:5017.
Han J-I, Choi H-K, Lee S-W et al. Complete genome sequence
of the metabolically versatile plant growth-promoting endo-
phyte Variovorax paradoxus S110. J Bacteriol 2011;193:118390.
Hussain M, Hamid MI, Wang N et al. The transcription factor
SKN7 regulates conidiation, thermotolerance, apoptotic-like
cell death and parasitism in the nematode endoparasitic
fungus Hirsutella minnesotensis. Sci Rep 2016;6:30047.
Kent AD, Triplett EW. Microbial communities and their interac-
tions in soil and rhizosphere ecosystems. Annu Rev Microbiol
2002;56:21136.
Kerry BR. Fungi and the decrease of cereal cyst-nematode pop-
ulations in cereal monoculture. EPPO Bull 1975;5:35361.
Kerry BR. Rhizosphere interactions and the exploitation of mi-
crobial agents for the biological control of plant-parasitic ne-
matodes. Annu Rev Phytopathol 2000;38:42341.
Kerry BR, Crump DH. The dynamics of the decline of the cereal
cyst nematode, Heterodera avenae, in four soils under inten-
sive cereal production. Fund Appl Nematol 1998;21:61725.
Kerry B, Crump D, Mullen L. Natural control of the cereal cyst
nematode, Heterodera avenae Woll., by soil fungi at three sites.
Crop Prot 1982;1:99109.
Liu XZ, Chen SY. Screening isolates of Hirsutella species for
biocontrol of Heterodera glycines. Biocontrol Sci Techn 2001;11:
15160.
Ma R, Liu X, Jian H et al. Detection of Hirsutella spp. and Pasteuria
sp. parasitizing second-stage juveniles of Heterodera glycines
in soybean fields in China. Biol Control 2005;33:2239.
Manzanilla R, Esteves I, Finetti M et al. Pochonia chlamydosporia:
advances and challenges to improve its performance as a
biological control agent of sedentary endo-parasitic nema-
todes. J Nematol 2013;45:17.
Mark GL, Dow JM, Kiely PD et al. Transcriptome profiling of
bacterial responses to root exudates identifies genes in-
volved in microbe-plant interactions. P Natl Acad Sci USA
2005;102:174549.
Mazzola M. Manipulation of rhizosphere bacterial communities
to induce suppressive soils. J Nematol 2007;39:21320.
Mazzola M, Cook RJ, Thomashow LS et al. Contribution of
phenazine antibiotic biosynthesis to the ecological compe-
tence of fluorescent pseudomonads in soil habitats. Appl En-
viron Microb 1992;58:261624.
Mendes R, Kruijt M, de Bruijn I et al. Deciphering the rhizo-
sphere microbiome for disease-suppressive bacteria. Science
2011;332:1097100.
Micallef SA, Shiaris MP, Colon-Carmona A. Influence of Arabidop-
sis thaliana accessions on rhizobacterial communities and
natural variation in root exudates. J Exp Bot 2009;60:172942.
Prasad P, Varshney D, Adholeya A. Whole genome annotation
and comparative genomic analyses of bio-control fungus Pur-
pureocillium lilacinum. BMC Genomics 2015;16:1004.
Quast C, Pruesse E, Yilmaz P et al. The SILVA ribosomal RNA gene
database project: improved data processing and web-based
tools. Nucleic Acids Res 2013;41:5906.
Raaijmakers JM, Mazzola M. Soil immune responses. Science
2016;352:13923.
Raaijmakers JM, Leeman M, Van Oorschot MM et al. Dose-
response relationships in biological control of Fusarium
wilt of radish by Pseudomonas spp. Phytopathology 1995;85:
107580.
Sanguin H, Sarniguet A, Gazengel K et al. Rhizosphere bacte-
rial communities associated with disease suppressiveness
stages of take-all decline in wheat monoculture. New Phytol
2009;184:694707.
Sayre RM, Wergin WP, Schmidt JM et al. Pasteuria nishizawae sp.
nov., a mycelial and endospore-forming bacterium parasitic
on cyst nematodes of genera Heterodera and Globodera. Res Mi-
crobiol 1991;142:55164.
Sun MH, Liu XZ. Suppressive soils of soybean cyst nematode in
China. Acta Phytopathol Sin 2000;30:3536.
Sung GH, Hywel-Jones NL, Sung JM et al. Phylogenetic classifica-
tion of Cordyceps and the clavicipitaceous fungi. Stud Mycol
2007;57:559.
Wang N, Zhang Y, Jiang X et al. Population genetics of Hir-
sutella rhossiliensis, a dominant parasite of cyst nematode ju-
veniles on a continental scale. Appl Environ Microb 2016;82:
631725.
Waweru B, Turoop L, Kahangi E et al. Non-pathogenic Fusarium
oxysporum endophytes provide field control of nematodes,
improving yield of banana (Musa sp.). Biol Control 2014;74:
828.
Weller DM, Landa B, Mavrodi O et al. Role of 2,
4-diacetylphloroglucinol-producing fluorescent Pseu-
domonas spp. in the defense of plant roots. Plant Biol 2007;9:
420.
Weller DM, Raaijmakers JM, Gardener BBM et al. Microbial popu-
lations responsible for specific soil suppressiveness to plant
pathogens 1. Annu Rev Phytopathol 2002;40:30948.
Westphal A, Becker J. Components of soil suppressiveness
against Heterodera schachtii. Soil Biol Biochem 2001;33:916.
Xiang M, Xiang P, Jiang X et al. Detection and quantification of the
nematophagous fungus Hirsutella minnesotensis in soil with
real-time PCR. Appl Soil Ecol 2010;44:1705.
Yin B, Valinsky L, Gao X et al. Bacterial rRNA genes associ-
ated with soil suppressiveness against the plant-parasitic
nematode Heterodera schachtii. Appl Environ Microb 2003a;69:
157380.
Yin B, Valinsky L, Gao X et al. Identification of fungal rDNA
associated with soil suppressiveness against Heterodera
schachtii using oligonucleotide fingerprinting. Phytopathology
2003b;93:100613.
Zhang L, Yang J, Niu Q et al. Investigation on the infection
mechanism of the fungus Clonostachys rosea against nema-
todes using the green fluorescent protein. Appl Microbiol Biot
2008;78:98390.
Zhu Y, Shi F, Tian J et al. Effect of soybean monoculture on the
bacterial communities associated with cysts of Heterodera
glycines. J Nematol 2013;45:22835.