Science of the Total Environment 543 (2016) 155–160

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Glyphosate effects on soil rhizosphere-associated bacterial communities

Molli M. Newman a,⁎, Nigel Hoilett b, Nicola Lorenz c, Richard P. Dick c, Mark R. Liles d,
Cliff Ramsier e, Joseph W. Kloepper a
a
Department of Entomology and Plant Pathology, Auburn University, CASIC Building, Auburn, AL 36849, USA
b
Department of Agricultural Sciences, Northwest Missouri State University, 800 University Drive, Maryville, MO 64468, USA
c
School of Environment and Natural Resources, The Ohio State University, 2021 Coffey Road, Columbus, OH 43210, USA
d
Department of Biological Sciences, Auburn University, CASIC Building, Auburn, AL 36849, USA
e
Ag Spectrum, 428 East 11th Street, DeWitt, IA 52742, USA

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• We examined the rhizosphere bacterial

community composition response to
glyphosate.
• Next-generation sequencing was used
to examine the rhizosphere bacterial
community.
• Relative abundance of Acidobacteria
decreased in response to glyphosate
exposure.
• Long-term glyphosate application could
affect rhizosphere nutrient status.

a r t i c l e i n f o a b s t r a c t

Article history: Glyphosate is one of the most widely used herbicides in agriculture with predictions that 1.35 million metric tons
Received 21 September 2015 will be used annually by 2017. With the advent of glyphosate tolerant (GT) cropping more than 10 years ago,
Received in revised form 2 November 2015 there is now concern for non-target effects on soil microbial communities that has potential to negatively affect
Accepted 3 November 2015
soil functions, plant health, and crop productivity. Although extensive research has been done on short-term re-
Available online 12 November 2015
sponse to glyphosate, relatively little information is available on long-term effects. Therefore, the overall objec-
Editor: D. Barcelo tive was to investigate shifts in the rhizosphere bacterial community following long-term glyphosate
application on GT corn and soybean in the greenhouse. In this study, rhizosphere soil was sampled from
Keywords: rhizoboxes following 4 growth periods, and bacterial community composition was compared between glypho-
Bacterial community composition sate treated and untreated rhizospheres using next-generation barcoded sequencing. In the presence or absence
Soil rhizosphere of glyphosate, corn and soybean rhizospheres were dominated by members of the phyla Proteobacteria,
Next-generation sequencing Acidobacteria, and Actinobacteria. Proteobacteria (particularly gammaproteobacteria) increased in relative abun-
Glyphosate dance for both crops following glyphosate exposure, and the relative abundance of Acidobacteria decreased in re-
16S rDNA
sponse to glyphosate exposure. Given that some members of the Acidobacteria are involved in biogeochemical

⁎ Corresponding author at: Department of Entomology and Plant Pathology, Auburn University, CASIC Building, Auburn, AL 36849, USA.
E-mail address: ramsemm@auburn.edu (M.M. Newman).

http://dx.doi.org/10.1016/j.scitotenv.2015.11.008
0048-9697/© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
156 M.M. Newman et al. / Science of the Total Environment 543 (2016) 155–160

processes, a decrease in their abundance could lead to significant changes in nutrient status of the rhizosphere.
Our results also highlight the need for applying culture-independent approaches in studying the effects of pesti-
cides on the soil and rhizosphere microbial community.
© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction composition as a whole. Such approaches may actually cause the effects
on lesser-abundant, yet still significant, taxa to be overlooked (Johnsen
Pesticides are substances or mixtures of substances intended for et al., 2001).
preventing, destroying, repelling or mitigating pests, and the major Mijangos et al. (2009) used DGGE in combination with Biolog
groups of pesticides are fungicides, herbicides, and insecticides (Grube Ecoplates™ and microbial biomass to assess the effects of glyphosate
et al., 2011). A recent comprehensive study by BCC Research of the glob- on rhizosphere soil microbial properties and observed a glyphosate-
al biopesticide and synthetic pesticide market estimated the global mar- induced stimulation of microbial activity and functional diversity
ket of pesticides in 2014 at $61.8 billion, with a projected increase to 15 days after glyphosate treatment in the culturable portion of the soil
$83.7 billion by 2019 (Lehr, 2014). Pesticides are typically used in the microbial community. But, this response was inconsistent when exam-
agricultural industry for improving crop yield and quality while also ining the microbial community 30 days after glyphosate addition. Using
maximizing economic returns. Herbicides are the most widely used PLFA and bacterial 16S rRNA genotyping via T-RFLP, Widenfalk et al.
class of pesticides in agriculture (Grube et al., 2011), and of all herbi- (2008) showed that the herbicide glyphosate increased the abundance
cides, glyphosate has the highest use world-wide with the global of branched, saturated fatty acids typical of Gram-positive bacteria in
market projected to reach 1.35 million metric tons by 2017 (Global freshwater sediment. Nearly all of the research reported above on
Industry Analysts, 2011). glyphosate was done under short-term conditions where a single or
Examining the effects of pesticides, such as glyphosate, on soil and one season application of glyphosate was applied, and as mentioned
rhizosphere microbial communities is important due to the critical above, often with integrative methods that might have missed subtle
role of microorganisms in driving biogeochemical processes, controlling effects on the soil microbial community. This misses the actual field con-
pathogens, and ultimately enabling ecosystems to function and provide ditions in the U.S. where glyphosate tolerant (GT) cropping has now
services to humanity. The soil microbial community, especially the been extensively used in the major agricultural regions for 10–15
rhizosphere microbial community, impacts soil quality through its in- years. In addition, common agricultural practices apply commercial for-
volvement in biogeochemical and nutrient cycling, long-term soil sus- mulations containing glyphosate, rather than the active ingredient
tainability, and resistance to perturbations (Prashar et al., 2014; Topp, alone. Given that the toxicity of commercial formulations may differ
2003). Within the rhizosphere, microorganisms positively affect plant from that of pure glyphosate (Sihtmäe et al., 2013; Tsui and Chu,
health through a variety of mechanisms, including mineralization of 2003), it is important to use commercial formulations in studies inves-
nutrients, suppression of disease, improving plant stress tolerance, tigating the effects of glyphosate-based pesticides.
and production of phytohormones (Berendsen et al., 2012; Figueiredo Recently, Nye et al. (2014) found on the same soil type that more
et al., 2011; Gupta et al., 2000). In agricultural systems, these effects than 10 years of GT cropping shifted the microbial PLFA diversity com-
on plant health have a major impact on crop production. pared to soil that had no history of glyphosate exposure. Although ef-
Numerous studies have investigated the impacts of glyphosate fects on overall microbial community composition and associated
on soil microbial properties using broad-scale or integrative bacterial subgroups were noted as a result of glyphosate exposure,
methods such as microbial biomass, enzyme activity, and respira- specific bacterial taxa affected were not identified. To fill this gap, a
tion. Bünemann et al. (2006) and Johnsen et al. (2001) provide ex- greenhouse study was conducted subjecting soil that had no history of
ceptional reviews of this literature. Typically the results of these glyphosate applications to GT cropping over 8 growing periods, simulat-
studies have shown no or transitory effects of glyphosate on the ing long-term field conditions. In this study, we examined the bacterial
above mentioned microbial properties. However, the effects of community composition from rhizosphere soil samples collected from
glyphosate may be masked by “functional redundancy” where overall the fourth growth period of this larger greenhouse study. And more spe-
soil functions are unaffected while microbial community composition cifically, we used next-generation barcoded sequencing, which permits
is altered and key functions mediated by specific microbial populations detailed phylogenetic diversity analysis (Imfeld and Vuilleumier, 2012).
are affected (Imfeld and Vuilleumier, 2012). Alterations to soil microbial Therefore, the objective of this particular study was to use next-
community composition and subsequent changes in microbial diversity generation barcoded sequencing to identify specific bacterial taxa shifts
could potentially have pronounced long-term effects on soil quality as in the rhizosphere bacterial community in response to repeated glyph-
well as impact plant health and therefore crop production (Bending osate exposure on corn and soybeans.
et al., 2007; Lynch et al., 2004).
Many studies examining the effects of glyphosate on the microbial 2. Materials and methods
community have used culture-based methods to target specific bacteri-
al populations of functional significance in the soil environment. For ex- 2.1. Greenhouse study
ample, a study by Zobiole et al. (2011) targeted populations of Fusarium,
fluorescent pseudomonads, Mn-transforming bacteria, and indoleacetic The soil used for the study was a Blount silt loam (fine, illitic mesic
acid-producing bacteria in rhizosphere soils of soybean receiving Aeric Epiaqualf). Soil pH was 6.95, and soil total C was 1.47%. Soil texture
glyphosate treatment and found that glyphosate treatment nega- was 11% sand, 48% silt, and 41% clay. Typical Blount soil clay mineralogy
tively impacted the interactions of these microbial groups, leading is characterized by illite, hydroxyl-interlayered vermiculite, kaolinite,
to increased Fusarium spp. abundance and reduced abundances of and quartz (Dontsova and Norton, 2002). Soil was collected in 2-cm in-
fluorescent pseudomonads, Mn-reducing bacteria and indole acetic crements to a depth of 39 cm, with 37 cm from the A horizon and the
acid-producing rhizobacteria. Johnsen et al. (2001) suggests, however, remaining 2 cm from the O horizon, from soil pits at a farm undergoing
that by targeting specific rhizosphere bacterial populations, little infor- organic management in Delaware County, OH. This field site was previ-
mation is gained regarding effects on rhizosphere bacterial community ously under rotation of alfalfa–orchard grass–corn, oats–alfalfa–orchard
 M.M. Newman et al. / Science of the Total Environment 543 (2016) 155–160 157

grass, spelt–timothy–clover, and timothy–clover. The soil had never rhizosphere. DNA was extracted from 500 mg of each soil rhizosphere
been exposed to glyphosate. Once collected, soil was stored in sealed sample using the UltraClean Microbial DNA Isolation Kit (MoBio Labora-
plastic bags returned to the lab on ice and placed in rhizoboxes starting tories, CA, USA) and eluted in 50 μl. DNA extracts were quantified using
with the 38–39-cm increment, using ~62 g of soil per cm fill height. The a Qubit Fluorometer and the dsDNA HS Assay kit (Life Technologies, CA,
soil was evenly distributed in the rhizobox and compacted to a bulk USA).
density of 1.3 g cm−3 and a total fresh soil weight within the rhizobox
of 2500 g. A total of eight rhizoboxes were constructed as described by 2.3. Sequencing library construction
Bott et al. (2008). Four rhizoboxes were planted for each of two crops,
corn and soybean. Two rhizoboxes per crop were treated with glypho- PCR primers (515F/806R) designed by Caporaso et al. (Caporaso
sate (Roundup PowerMax, Monsanto Company, MO, USA; active ingre- et al., 2012) were used to amplify the bacterial V4 hypervariable region
dient: glyphosate, N-(phosphonomethyl) glycine, in the form of its of the 16S rRNA gene. Each primer contained the sequence adapter re-
potassium salt), and two rhizoboxes served as untreated plant controls. gions used by Caporaso et al. (Caporaso et al., 2012), and the reverse
These eight rhizoboxes were part of a larger ongoing research project PCR primers contain a 12-base Golay barcode. Three sequencing
that utilized all available rhizoboxes, leading to two rhizoboxes per primers were designed based on those of Caporaso et al. (Caporaso
treatment combination in this study. et al., 2012) to yield the 5′ read, the 3′ read, and the index read. See
Plants were grown in eight growth periods over three years, with Table 2 for a description of the primers used.
each growth period lasting 58 days. Plants were fertilized twice per PCR reagent mixes contained 12.5 μl KAPA HiFi HotStart Ready Mix
growth period by applying 25 mL of fertilizer solution per rhizobox. (2×), 0.75 μl each of the forward and reverse primers (10 μM final con-
Fertilizer solution was prepared by dissolving 3.745 g of Peters® centration), 10 ng genomic DNA, and PCR water for a total reaction vol-
20/20/20 Professional fertilizer per liter, equaling 0.749 mg N, ume of 25 μl. The following touchdown PCR conditions were used:
0.749 mg P, and 0.749 mg K mL−1 of fertilizer solution. Fertilizer trace initial denaturation at 95 °C for 2 min followed by 32 cycles of denatur-
element concentrations were magnesium (0.019 mg mL− 1), boron ation at 98 °C for 20 s, annealing beginning at 61 °C and ending at 50 °C
(0.749 μg mL− 1), copper (0.002 mg mL− 1), iron (0.004 mg mL− 1), for 30 s, and extension at 72 °C for 30 s. The annealing temperature was
manganese (0.002 mg mL− 1), molybdenum (0.019 μg mL− 1), and lowered 1 °C every cycle until reaching 50 °C, which was used for the re-
zinc (0.002 mg mL−1). The fertilizer was applied on days 30 and 50. maining cycles. Following this, a final extension of 72 °C for 10 min was
The schedule for each period is outlined in Table 1. used. PCR products were purified by ethanol precipitation and verified
On day 1, before planting, all rhizoboxes were sprayed with glypho- on a 1% agarose gel. Positive amplicons were quantified using a Qubit
sate except for the controls. Glyphosate was applied at the recommend- Fluorometer and the dsDNA HS Assay kit (Life Technologies, CA, USA).
ed field rate (300.79 mL ha−1). Corn and soybean seedlings germinated Amplicons were pooled at equimolar concentrations, and the
on cotton tissue were transplanted into rhizoboxes (2 plants/box) on resulting pooled library was size-selected to remove smaller primer
day 10. Roundup Ready corn (Zea mays; DeKalb hybrid seed brand dimers. Since the 16S rRNA gene amplicon was approximately 420 bp,
DKC62-54 (VT3)) and soybean (Glycine max; OX 20-8 RR) were used. the E.Z.N.A. Size Select-IT Kit (Omega Bio-Tek, GA, USA) was used on
Growth stages were estimated using the shortest periods given in the the pooled bacterial 16S rRNA gene library, targeting 150–500 bp frag-
Ontario Agronomy Guide (Baute et al., 2002) for corn and soybean. On ments. The library was quantified using a Qubit Fluorometer and
days 30 and 51 (when plants reached the V–5 and V–7 growth stages, dsDNA HS Assay kit (Life Technologies, CA, USA). The library was
respectively), glyphosate was applied on plant leaves using a cell denatured with 0.2 N NaOH and diluted with pre-chilled HT1 buffer
spreader. Soil rhizosphere samples were collected on days 31, 37, 52, (Illumina, CA, USA) to a final concentration of 8 pM. The denatured
and 58. This schedule was then repeated for a total of eight growth pe- and diluted library was spiked with 40% denatured PhiX and sequenced
riods. The rhizosphere soil samples used in this study were collected on separately on an Illumina MiSeq (Illumina, CA, USA) using the sequenc-
day 58 of the fourth growth period. ing primers mentioned above and a 300-cycle (2 × 150) MiSeq Reagent
Kit v2 (Illumina, CA, USA).
2.2. Sample collection and DNA extraction
2.4. Data analysis
Samples for this study were collected in the fourth growth period for
corn and soybean. For the collection of rhizosphere soil samples, Paired-end reads were assembled using PANDAseq (Bartram et al.,
rhizoboxes were placed horizontally on the lab bench and clamps and 2011; Masella et al., 2012), and all downstream processing of sequences
the top acrylic plate were removed. Three 5-g subsamples of soil were was completed using the QIIME pipeline v1.5.0 (Caporaso et al., 2010b).
collected using a spatula to recover soil within a 1-mm vicinity of the Assembled sequences were quality filtered using USEARCH v7 (Edgar,
primary and lateral roots, avoiding the areas around the root tips and 2010), retaining only sequences N 75 bases in length with expected
stored at −80 °C until further processing. These subsamples were proc-
essed separately, and the resulting sequence data was combined to
account for variability in bacterial community composition within the Table 2
Primers used for amplification of bacterial 16S rRNA gene V4 hypervariable region.

Primers Sequence (5′ to 3′)a

Table 1
Forward aatgatacggcgaccaccgagatctacacTATGGTAATTgtGTGCCAGCMGCCGCGGTAA
Schedule of events per growth period.
(515F)
Day Event Reverse caagcagaagacggcatacgagatNNNNNNNNNNNNbAGTCAGTCAGccGGAC
(806R) TACHVGGGTWTCTAAT
1 Glyphosate burn down spray
10 Corn and soybean planted Sequencing
30 Glyphosate spraya Read 1 TATGGTAATTgtGTGCCAGCMGCCGCGGTAA
31 Collection of rhizosphere soil samples Read 2 AGTCAGTCAGccGGACTACHVGGGTWTCTAAT
37 Collection of rhizosphere soil samples Index ATTAGAWACCCBDGTAGTCCggCTGACTGACT
51 Glyphosate sprayb a
Lowercase letters denote adapter sequences, underlined letters are pad regions, low-
52 Collection of rhizosphere soil samples
ercase bold letters are linker regions, and uppercase letters are primer sequences specific
58 Collection of rhizosphere soil samples
to the V4/ITS1 regions.
a b
Exact application at V3–V5 growth stages. N's represent location of 12-base Golay barcode; see Caporaso et al. (2012) for a listing
b
Exact application at V6–V7 growth stages. of the barcodes used.
158 M.M. Newman et al. / Science of the Total Environment 543 (2016) 155–160

errors N 1.0. Chimeric sequences were identified and removed using

USEARCH v6.1 (Edgar, 2010). Sequences were assigned to operational
taxonomic units (OTUs) using open reference OTU picking in
USEARCH v6.1 at a 97% similarity threshold. Reads were first clustered
against the Greengenes 16S rRNA gene database (Aug 2013 release)
(DeSantis et al., 2006), and then all remaining reads that did not cluster
were clustered de novo. A representative sequence was chosen for each
OTU based on which sequence was the most abundant for that given
OTU. Taxonomy was assigned to the representative sequence of each
OTU using uclust with the Greengenes 16S rRNA gene database (Aug
2013 release) (DeSantis et al., 2006). In addition, sequences identified
as chloroplast following taxonomic assignment were removed. Reads
were aligned using PyNAST (Caporaso et al., 2010a) against a reference
alignment of the Greengenes core set (McDonald et al., 2011).
The generated OTU table was rarified to an even sampling depth
(29,205 sequences per sample) using the single_rarefaction.py script
in the QIIME pipeline, and this rarified OTU table was used to calculate
alpha diversity metrics, including OTU abundance, Chao1 (Chao, 1984),
Faith's phylogenetic diversity (Faith, 1992), and Shannon's index
(Shannon, 1948). Alpha diversity metrics were compared between Fig. 1. PCoA plot based on weighted Unifrac distances generated for control (solid sym-
control and glyphosate-treated samples for each crop using the bols) and glyphosate-treated (hollow symbols) rhizosphere bacterial communities of
compare_alpha_diversity.py script in the QIIME pipeline which imple- corn (squares) and soybean (circles) following four growth periods.

ments a nonparametric two-sample t-test with 999 Monte Carlo

permutations. Beta diversity metrics were also estimated using the soybean. Chao1 richness estimates increased in corn following glypho-
rarified OTU table, including unweighted and weighted UniFrac dis- sate treatment (e.g. 5872 to 6154) but decreased in soybean (e.g. 5946
tances (Lozupone and Knight, 2005). Weighted UniFrac distances to 5754). Phylogenetic diversity observed within the corn rhizosphere
were compared using a multiple response permutation procedure was 182.29 in controls and 189.66 in glyphosate-treated samples, and
(MRPP) with 999 permutations. Principal coordinates analysis was per- there was no increase in phylogenetic diversity for soybean rhizosphere
formed using the beta_diversity_through_plots.py script in the QIIME samples following glyphosate treatment. Shannon's diversity estimates
pipeline which uses the weighted UniFrac distances to generate plots were similar between control and glyphosate-treated rhizosphere
and aid in visualization of the relationships among the various samples samples with a mean Shannon's diversity estimate of 10.2. Weighted
and treatments. All sequences obtained in this study were submitted to UniFrac distances showed that rhizosphere beta diversity varied by
the NCBI Sequence Read Archive (SRA) and are available under the plant species (p = 0.029; α = 0.10) but was fairly similar overall be-
study accession number PRJNA284763. tween control and glyphosate-treated samples (p = 0.78; α = 0.10).
Fig. 1 contains a PCoA plot of these results.
3. Results
3.3. Rhizosphere bacterial community composition
3.1. Sequencing summary
Control samples and those receiving long-term glyphosate ap-
Following assembly and quality filtering, a total of 505,391 bacterial plications showed similarities in bacterial community composition
16S rRNA gene sequences were obtained with a range of 29,205 to at the phylum level. Control and treatment rhizosphere samples
66,702 sequences per rhizobox and a mean of 49,249 sequences per for both corn and soybean were dominated by members of the
rhizobox. All rarefaction curves tended to approach a plateau, indicating phyla Proteobacteria, Acidobacteria, and Actinobacteria (Fig. 2). The
that the number of sequences obtained was sufficient to describe the abundance of Proteobacteria-affiliated sequences increased in re-
bacterial diversity within these samples (Supplemental Fig. S1). sponse to glyphosate treatment (p = 0.096). Corn rhizosphere
samples showed an increase from an average of 22.9 ± 1.5%
3.2. Rhizosphere bacterial community diversity Proteobacteria sequences to 25.9 ± 0.9%. Soybean rhizosphere sam-
ple Proteobacteria sequences increased from an average of 25.4 ±
Alpha diversity estimates were similar between corn and soybean 1.2% to 27.2 ± 0.2%. Within the sequences identified as belonging
rhizospheres as well as among the control and glyphosate-treated sam- to the phylum Proteobacteria, no one bacterial class dominated for ei-
ples (Table 3). Mean OTU abundance within corn and soybean control ther corn or soybean. The alphaproteobacteria, betaproteobacteria, and
rhizospheres was 3814 and 3849, respectively. This increased slightly gammaproteobacteria classes were present in controls samples for
following glyphosate treatment to 4001 OTUs in corn and 3893 in corn and soybean (ranging from approximately 5.2–8.3% relative

Table 3
Alpha diversity metrics for rhizosphere samples collected from control and glyphosate-treated rhizospheres of corn and soybean. Values represent mean ± 1SE.

Observed OTUs Chao1 richness estimate Faith's phylogenetic diversity Shannon's index

Corn
Control 3814 ± 60 5872 ± 233 182.3 ± 3.1 10.2 ± 0.03
Glyphosate 4001 ± 86 6154 ± 111 189.7 ± 4.3 10.3 ± 0.05

Soybean
Control 3849 ± 2 5946 ± 49 184.2 ± 0.2 10.2 ± 0.04
Glyphosate 3893 ± 101 5754 ± 210 184.2 ± 4.0 10.2 ± 0.08
Crop Effecta 0.656 0.428 0.604 0.91
Treatment Effecta 0.201 0.928 0.318 0.241
a
Values represent p-values calculated using a nonparametric two-sample t-test with 999 Monte Carlo permutations.
 M.M. Newman et al. / Science of the Total Environment 543 (2016) 155–160 159

example, many Gram-negative fluorescent Pseudomonas species fall

within the phylum Proteobacteria and have been reported to benefit
plants by stimulating plant growth and exhibiting traits involved in bi-
ological control of plant diseases (Lugtenberg and Kamilova, 2009).
The results of this study showed subtle alterations to rhizosphere
bacterial community composition following the application of the her-
bicide glyphosate. The largest shifts in relative abundance were
observed for Proteobacteria (specifically gammaproteobacteria) and
Acidobacteria. The increase in γ-Proteobacteria relative abundance for
both corn and soybean rhizosphere samples was driven by increases
in bacteria from the family Xanthomonadaceae following glyphosate
treatment, suggesting that Xanthomonadaceae are adapted to and/
or enriched by environments containing glyphosate. Previous stud-
ies also noted an increased abundance of bacteria from the family
Xanthomonadaceae in response to long-term fertilization and have
cited a potential importance of Xanthomonadales in the bacterial
population dynamics of altered soils (Campbell et al., 2010). Inter-
estingly, members of the novel family Sinobacteraceae (Order
Fig. 2. Relative abundance of bacterial phyla present in control and glyphosate-treated rhi- Xanthamonadales) have been isolated from a soil that was adjacent
zosphere bacterial communities of corn and soybean following four growth periods.
to and contaminated by a chemical factory that produced herbicides,
including glyphosate (personal communication Zhou et al., 2008),
suggesting that the members of this family are increased in their
abundance for each), as well as to a lesser extent the deltaproteobacteria abundance and metabolic activity in response to herbicide contami-
(3.2% corn, 4.1% soybean). nation of soils.
Following glyphosate treatment, all classes of Proteobacteria in- Concomitantly, upon treatment with glyphosate, in both corn
creased in relative abundance. Gammaproteobacteria sequences in- and soybean there were decreases in the relative abundance of
creased the most for both crops with an increase of 1.5% in corn and Acidobacteria, particularly the Acidobacteria subgroup 6. Acidobacteria
0.7% in soybean. The majority of gammaproteobacteria sequences in con- have been found to be dominant members of rhizosphere soil and are
trol samples (2.6% corn, 3.2% soybean) were identified as belonging to believed to be highly involved in biogeochemical processes within
the order Xanthomonadales, mainly from the families Sinobacteraceae the rhizosphere particularly for cellulose degradation (Eichorst et al.,
and Xanthomonadaceae. Within the families Sinobacteraceae and 2011; Lee et al., 2008; Štursová et al., 2012). Long-term decreases in
Xanthomonadaceae, the majority of sequences matched reference the abundance of these bacteria could impair the ability of soil to per-
sequences from unidentified genera. Other lower abundance genera form certain biogeochemical reactions performed by these organisms.
present from these families included Steroidobacter within the The decrease of Acidobacteria was more dramatic in corn, suggesting
family Sinobacteraceae and Arenimonas, Dokdonella, Luteibacter, that any subsequent effects on biogeochemical processes due to re-
Lysobacter, Pseudoxanthomonas, and Thermomonas within the family duced Acidobacteria taxa abundance and/or activity would be more pro-
Xanthomonadaceae. The relative abundance of Xanthomonadaceae nounced in corn. This has implications for growing corn with GT
sequences increased for both crops following glyphosate treatment cropping which may exacerbate the reduction of Acidobacteria taxa
(p = 0.081). The response of Sinobacteraceae to glyphosate treatment over a corn–soybean rotation. The abundance of this same subgroup
varied with crop (p = 0.003). In corn, Sinobacteraceae relative abun- of Acidobacteria was previously shown to decrease in subsurface sedi-
dance decreased (2.2% to 1.5%) following glyphosate treatment, but in ments contaminated with uranium (Barns et al., 2007), indicating that
soybean relative abundance of Sinobacteraceae increased (1.5% to 2.1%). these taxa are responsive and sensitive to environmental change and
In contrast, the relative abundance of members of the phylum may serve as useful bioindicators of environmental alteration.
Acidobacteria showed a decrease in response to glyphosate treatment Although effects of glyphosate on specific bacterial taxonomic
(p = 0.083). In corn, the average relative abundance of Acidobacteria groups were observed, there was no overall effect of glyphosate on
sequences decreased from 21.5 ± 1.1% in the control samples to bacterial community diversity. This highlights the need to examine
18.7 ± 0.8% in glyphosate-treated samples. For soybean there was also the microbial diversity response to herbicide application at a finer
a decrease in the average relative abundance of Acidobacteria se- level both taxonomically as well as functionally rather than solely
quences from 22.3 ± 0.6% in control samples to 21.5 ± 0.3% in looking at net diversity responses. In addition, bacteria within the
glyphosate-treated samples. The Acidobacteria subgroup 6, a domi- phylum Acidobacteria are somewhat recalcitrant to cultivation, espe-
nant Acidobacteria subgroup in soils with few cultured representa- cially given their high abundance in soil (George et al., 2011; Janssen,
tives, made up the majority of Acidobacteria sequences for both 2006). The high relative abundance of Acidobacteria ribotypes observed
corn (45.9%) and soybean (49.1%) and decreased in abundance follow- within the rhizosphere of this study and their reduced abundance in
ing glyphosate treatment for corn, and to a lesser extent soybean. The glyphosate-treated rhizosphere soil would not have been observed
average relative abundance of Actinobacteria also decreased following using culture-based methods but was made possible by employing
glyphosate treatment from 16.45% to 14.95% in corn and 14.35% to next-generation sequencing instrumentation.
12.6% in soybean (p = 0.445). The results of this study are specific to the soil used and alternative
results could have occurred on other soil types. This is because soil tex-
4. Discussion ture, mineralogy, pH, and organic matter have a major impact on the
microbial community structure and secondly, on the fate, decomposi-
Analysis of the corn and soybean rhizosphere microbiota indicated tion, and sorption of glyphosate and its metabolites, most notably
dominance by the members of the bacterial phyla Proteobacteria, aminomethylphosphonic acid (AMPA) as the major decomposition in-
Acidobacteria, and Actinobacteria. All of these phyla contain taxa com- termediate (Ascolani Yael et al., 2014; Franz et al., 1997). Thus, each
monly found within soil rhizospheres that are capable of having various soil type would have a variable rate of decomposition and degree of
effects on plant health including beneficial and pathogenic interactions sorption and toxicity of glyphosate or AMPA for susceptible populations.
(Berendsen et al., 2012; Lee et al., 2008; Philippot et al., 2013). For Less is known about AMPA sorption, but for glyphosate it is primarily
160 M.M. Newman et al. / Science of the Total Environment 543 (2016) 155–160

sorbed on surfaces of variable-charge clay and inorganic precipitates. Duke, S.O., Lydon, J., Koskinen, W.C., Moorman, T.B., Chaney, R.L., Hammerschmidt, R.,
2012. Glyphosate effects on plant mineral nutrition, crop rhizosphere microbiota,
Sorption increases as surface area of the minerals increases and pH de- and plant disease in glyphosate-resistant crops. J. Agric. Food Chem. 60, 10375–10397.
creases (Duke et al., 2012). Thus the amount of glyphosate (and pre- Edgar, R.C., 2010. Search and clustering orders of magnitude faster than BLAST. Bioinfor-
sumably AMPA) (Rampazzo et al., 2013) sorption capacity would be matics 26, 2460–2461.
Eichorst, S.A., Kuske, C.R., Schmidt, T.M., 2011. Influence of plant polymers on the distri-
relatively high for our study because this soil has a high clay content bution and cultivation of bacteria in the phylum Acidobacteria. Appl. Environ.
at 41% (and therefore high surface area), but from a pH perspective Microbiol. 77, 586–596.
this soil at 6.9 is less favorable for glyphosate sorption. However, rela- Faith, D.P., 1992. Conservation evaluation and phylogenetic diversity. Biol. Conserv. 61,
1–10.
tively little is known about rates of microbial toxicity for glyphosate or Figueiredo, M.D.V.B., Seldin, L., de Araujo, F.F., Mariano, R.L.R., 2011. Plant growth promot-
AMPA in soil solution or in any sorbed fraction. In part this is due to in- ing rhizobacteria: fundamentals and applications. Plant Growth and Health Promot-
ability to accurately determine what is biologically available to microor- ing Bacteria. Springer, pp. 21–43.
Franz, J.E., Mao, M.K., Sikorski, J.A., 1997. Glyphosate: A Unique Global Herbicide.
ganisms (Duke et al., 2012). Furthermore in our case, we did not have
American Chemical Society.
several soil types with variable levels of chemical and physical proper- George, I.F., Hartmann, M., Liles, M.R., Agathos, S.N., 2011. Recovery of as-yet-uncultured
ties in order to draw inferences and indirectly determine toxicity effects soil Acidobacteria on dilute solid media. Appl. Environ. Microbiol. 77, 8184–8188.
on microbial community members. Global Industry Analysts, 2011. Global Glyphosate Market to Reach 1.35 Million Metric
Tons by 2017, According to a New Report by Global Industry Analysts, Inc. (10 Octo-
Given the wide-spread and increasing use of glyphosate, the results ber) .
of our study are important because they show that shifts in subpopula- Grube, A., Donaldson, D., Kiely, T., Wu, L., 2011. Pesticides Industry Sales and Usage. US
tions (particularly Xanthamonadales, Acidobacteria) are possible under EPA, Washington, DC.
Gupta, A., Gopal, M., Tilak, K., 2000. Mechanism of plant growth promotion by
long-term GT cropping. However, this does not provide any information rhizobacteria. Indian J. Exp. Biol. 38, 856–862.
on whether this shift affects functional capability of the soil under GT Imfeld, G., Vuilleumier, S., 2012. Measuring the effects of pesticides on bacterial commu-
cropping. Thus, our research provides justification for more detailed, nities in soil: a critical review. Eur. J. Soil Biol. 49, 22–30.
Janssen, P.H., 2006. Identifying the dominant soil bacterial taxa in libraries of 16S rRNA
culture-independent investigations on function and delivery of ecosys- and 16S rRNA genes. Appl. Environ. Microbiol. 72, 1719–1728.
tem services or negative impacts such as crop pathogens. Johnsen, K., Jacobsen, C.S., Torsvik, V., Sorensen, J., 2001. Pesticide effects on bacterial di-
Supplementary data to this article can be found online at http://dx. versity in agricultural soils—a review. Biol. Fertil. Soils 33, 443–453.
Lee, S.-H., Ka, J.-O., Cho, J.-C., 2008. Members of the phylum Acidobacteria are dominant
doi.org/10.1016/j.scitotenv.2015.11.008.
and metabolically active in rhizosphere soil. FEMS Microbiol. Lett. 285, 263–269.
Lehr, P., 2014. Global markets for biopesticides. Report Code CHM029E BCC Research.
Acknowledgments Lozupone, C., Knight, R., 2005. UniFrac: a new phylogenetic method for comparing micro-
bial communities. Appl. Environ. Microbiol. 71, 8228–8235.
Lugtenberg, B., Kamilova, F., 2009. Plant-growth-promoting rhizobacteria. Annu. Rev.
This research was generously supported by Ag Spectrum Company, Microbiol. 63, 541–556.
DeWitt, Iowa. Lynch, J., Benedetti, A., Insam, H., Nuti, M., Smalla, K., Torsvik, V., et al., 2004. Microbial di-
versity in soil: ecological theories, the contribution of molecular techniques and the
impact of transgenic plants and transgenic microorganisms. Biol. Fertil. Soils 40,
References 363–385.
Masella, A.P., Bartram, A.K., Truszkowski, J.M., Brown, D.G., Neufeld, J.D., 2012. PANDAseq:
Ascolani Yael, J., Fuhr, J.D., Bocan, G.A., Daza Millone, A., Tognalli, N., dos Santos, A.M., et paired-end assembler for illumina sequences. BMC Bioinf. 13, 31.
al., 2014. Abiotic degradation of glyphosate into aminomethylphosphonic acid in McDonald, D., Price, M.N., Goodrich, J., Nawrocki, E.P., DeSantis, T.Z., Probst, A., et al., 2011.
the presence of metals. J. Agric. Food Chem. 62, 9651–9656. An improved Greengenes taxonomy with explicit ranks for ecological and evolution-
Barns, S.M., Cain, E.C., Sommerville, L., Kuske, C.R., 2007. Acidobacteria phylum sequences ary analyses of bacteria and archaea. ISME J. 6, 610–618.
in uranium-contaminated subsurface sediments greatly expand the known diversity Mijangos, I., Becerril, J.M., Albizu, I., Epelde, L., Garbisu, C., 2009. Effects of glyphosate on
within the phylum. Appl. Environ. Microbiol. 73, 3113–3116. rhizosphere soil microbial communities under two different plant compositions by
Bartram, A.K., Lynch, M.D., Stearns, J.C., Moreno-Hagelsieb, G., Neufeld, J.D., 2011. Gener- cultivation-dependent and-independent methodologies. Soil Biol. Biochem. 41,
ation of multimillion-sequence 16S rRNA gene libraries from complex microbial com- 505–513.
munities by assembling paired-end Illumina reads. Appl. Environ. Microbiol. 77, Nye, M., Hoilett, N., Ramsier, C., Renz, P., Dick, R.P., 2014. Microbial community structure
3846–3852. in soils amended with glyphosate-tolerant soybean residue. Appl. Ecol. Environ. Sci.
Baute, T., Hayes, A., McDonald, I., Reid, K., 2002. Agronomy guide for field crops. Publica- 2, 74–81.
tion 811. The Ontario Ministry of Agriculture. Food and Rural Affairs. Philippot, L., Raaijmakers, J.M., Lemanceau, P., van der Putten, W.H., 2013. Going back to
Bending, G.D., Rodriguez-Cruz, M.S., Lincoln, S.D., 2007. Fungicide impacts on microbial the roots: the microbial ecology of the rhizosphere. Nat. Rev. Microbiol. 11, 789–799.
communities in soils with contrasting management histories. Chemosphere 69, Prashar, P., Kapoor, N., Sachdeva, S., 2014. Rhizosphere: its structure, bacterial diversity
82–88. and significance. Rev. Environ. Sci. Bio/Technol. 13, 63–77.
Berendsen, R.L., Pieterse, C.M., Bakker, P.A., 2012. The rhizosphere microbiome and plant Rampazzo, N., Rampazzo Todorovic, G., Mentler, A., Blum, W., 2013. Adsorption of glyph-
health. Trends Plant Sci. 17, 478–486. osate and aminomethylphosphonic acid in soils. Int. Agrophys. 27, 203–209.
Bott, S., Tesfamariam, T., Candan, H., Cakmak, I., Römheld, V., Neumann, G., 2008. Glyphosate- Shannon, C.E., 1948. A mathematical theory of communication. Bell Syst. Tech. J. 27,
induced impairment of plant growth and micronutrient status in Glyphosate-resistant 379–423.
soybean (Glycine max L.). Plant Soil 312, 185–194. Sihtmäe, M., Blinova, I., Künnis-Beres, K., Kanarbik, L., Heinlaan, M., Kahru, A., 2013. Eco-
Bünemann, E., Schwenke, G., Van Zwieten, L., 2006. Impact of agricultural inputs on soil toxicological effects of different glyphosate formulations. Appl. Soil Ecol. 72, 215–224.
organisms—a review. Soil Res. 44, 379–406. Štursová, M., Žifčáková, L., Leigh, M.B., Burgess, R., Baldrian, P., 2012. Cellulose utilization
Campbell, B.J., Polson, S.W., Hanson, T.E., Mack, M.C., Schuur, E.A.G., 2010. The effect of nu- in forest litter and soil: identification of bacterial and fungal decomposers. FEMS
trient deposition on bacterial communities in Arctic tundra soil. Environ. Microbiol. Microbiol. Ecol. 80, 735–746.
12, 1842–1854. Topp, E., 2003. Bacteria in agricultural soils: diversity, role and future perspectives. Can.
Caporaso, J.G., Bittinger, K., Bushman, F.D., DeSantis, T.Z., Andersen, G.L., Knight, R., 2010a. J. Soil Sci. 83, 303–309.
PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformat- Tsui, M.T.K., Chu, L.M., 2003. Aquatic toxicity of glyphosate-based formulations: compar-
ics 26, 266–267. ison between different organisms and the effects of environmental factors.
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., et al., Chemosphere 52, 1189–1197.
2010b. QIIME allows analysis of high-throughput community sequencing data. Nat. Widenfalk, A., Bertilsson, S., Sundh, I., Goedkoop, W., 2008. Effects of pesticides on com-
Methods 7, 335–336. munity composition and activity of sediment microbes—responses at various levels
Caporaso, J.G., Lauber, C.L., Walters, W.A., Berg-Lyons, D., Huntley, J., Fierer, N., et al., 2012. of microbial community organization. Environ. Pollut. 152, 576–584.
Ultra-high-throughput microbial community analysis on the Illumina HiSeq and Zhou, Y., Zhang, Y.-Q., Zhi, X.-Y., Wang, X., Dong, J., Chen, Y., et al., 2008. Description of
MiSeq platforms. ISME J. 6, 1621–1624. Sinobacter flavus gen. nov., sp. nov., and proposal of Sinobacteraceae fam. nov. Int.
Chao, A., 1984. Nonparametric-estimation of the number of classes in a population. Scand. J. Syst. Evol. Microbiol. 58, 184–189.
J. Stat. 11, 265–270. Zobiole, L., Kremer, R., Oliveira, R., Constantin, J., 2011. Glyphosate affects micro-
DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K., et al., 2006. organisms in rhizospheres of glyphosate-resistant soybeans. J. Appl. Microbiol. 110,
Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible 118–127.
with ARB. Appl. Environ. Microbiol. 72, 5069–5072.
Dontsova, K.M., Norton, L.D., 2002. Clay dispersion, infiltration, and erosion as influenced
by exchangeable Ca and Mg. Soil Sci. 167, 184–193.