Journal of Applied Microbiology, 2023, 134, 1–12
https://doi.org/10.1093/jambio/lxad006
Advance access publication date: 13 January 2023
Research Article
Glyphosate-based herbicide use affects individual
microbial taxa in strawberry endosphere but not the
microbial community composition
Suni Anie Mathew1,*, Benjamin Fuchs2, Riitta Nissinen3, Marjo Helander1, Pere Puigbò1,4,5,
Kari Saikkonen2, Anne Muola2,6
1Department of Biology, University of Turku, 20014 Turku, Finland
2Biodiversity Unit, University of Turku, 20014 Turku, Finland
3Department of Biological and Environmental Science, University of Jyväskylä, 40014 Jyväskylä, Finland
4Nutrition and Health Unit, Eurecat Technology Centre of Catalonia, 43204 Reus, Catalonia
5Department of Biochemistry and Biotechnology, Rovira i Virgili University, 43007 Tarragona, Catalonia
6Division of Biotechnology and Plant Health, Norwegian Institute of Bioeconomy Research, 9016 Tromsø, Norway
∗Corresponding author. Department of Biology, University of Turku, 20014 Turku, Finland. E-mail: suni.mathew@utu.fi
Abstract
Aims: In a field study, the effects of treatments of glyphosate-based herbicides (GBHs) in soil, alone and in combination with phosphate fertilizer,
were examined on the performance and endophytic microbiota of garden strawberry.
Methods and results: The root and leaf endophytic microbiota of garden strawberries grown in GBH-treated and untreated soil, with and
without phosphate fertilizer, were analyzed. Next, bioinformatics analysis on the type of 5-enolpyruvylshikimate-3-phosphate synthase enzyme
was conducted to assess the potential sensitivity of strawberry-associated bacteria and fungi to glyphosate, and to compare the results with
field observations. GBH treatments altered the abundance and/or frequency of several operational taxonomic units (OTUs), especially those of
root-associated fungi and bacteria. These changes were partly related to their sensitivity to glyphosate. Still, GBH treatments did not shape
the overall community structure of strawberry microbiota or affect plant performance. Phosphate fertilizer increased the abundance of both
glyphosate-resistant and glyphosate-sensitive bacterial OTUs, regardless of the GBH treatments.
Conclusions: These findings demonstrate that although the overall community structure of strawberry endophytic microbes is not affected by
GBH use, some individual taxa are.
Significance and impact of study
Agrochemical residues in soil can shape the endophytic microbiota in crop plants.
Keywords: EPSPS, Fragaria× ananassa, herbicide, roundup, endophyte, microbiota, phosphate fertilizer, plant performance
Introduction
Glyphosate (N-phosphonomethyl glycine)-based herbicides
(GBHs) are currently the most commonly used pesticides
globally (Benbrook 2016). Glyphosate has been considered
safe for nontarget organisms for two reasons. First, the
effect of glyphosate is based on inactivation of the en-
zyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS)
in the shikimate pathway inhibiting the production of aro-
matic amino acids phenylalanine, tyrosine, and tryptophan
(Parthasarathy et al. 2018), which is an essential metabolic
route in plants (Duke and Powles 2008) but is not present
in animal cells (Williams et al. 2000, Helander et al. 2012).
Second, glyphosate typically degrades in a few weeks or is ad-
sorbed by soil solids; this limits its bioavailability and, thus,
phytotoxicity (Sprankle et al. 1975). However, mounting ev-
idence has revealed that there are ecological, environmental,
and health risks of intensive GBH use that are largely related
to the presence of glyphosate residues and degradation prod-
ucts in diverse agricultural habitats (Gill et al. 2018, Maggi
et al. 2020, Leino et al. 2021). Still, it can be difficult to find
a clear link between these risks and the use of GBHs because
glyphosate and its degradation products can be retained and
transported in ecosystems and their effects are often context
dependent (Muola et al. 2021, Fuchs et al. 2022a).
Despite the relatively fast degradation of glyphosate un-
der optimal conditions, glyphosate itself and its degradation
products have been reported to persist in ecosystems even for
years (Laitinen et al. 2009), especially in colder climates (He-
lander et al. 2012). In fact, an increasing number of studies
have reported that, following the use of GBHs, residues of
glyphosate and its degradation products are found in soil, wa-
ter, crop plants, and even animal tissues (Bøhn et al. 2014, Bai
and Ogbourne 2016, Helander et al. 2018, Silva et al. 2019,
Ruuskanen et al. 2020). Further, even the use of field-realistic
doses of GBH with a 2-week safety period has been shown
to negatively affect the germination and growth of different
crop plants (Helander et al. 2019). However, in other stud-
ies, low residues of glyphosate in soil have shown growth-
promoting effects in some crop species (Cedergreen 2008,
Brito et al. 2018,Helander et al. 2019, Fuchs et al. 2022b). The
Received: October 10, 2022. Accepted: January 12, 2023
C© The Author(s) 2023. Published by Oxford University Press on behalf of Applied Microbiology International. This is an Open Access article distributed
under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution,
and reproduction in any medium, provided the original work is properly cited.
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
2 Mathew et al.
consequences of glyphosate residues for plants have shown
to depend on plant species and weather conditions (i.e. via
the degradation of GBHs) in addition to soil chemistry, es-
pecially in connection to soil phosphorus concentration (He-
lander et al. 2012). Glyphosate and inorganic phosphate are
known to compete for sorption sites in a wide variety of soils
(for instance, Sprankle et al. 1975, de Jonge and de Jonge
1999, de Jonge et al. 2001, Kanissery et al. 2015, Hébert et al.
2019). Depending on weather conditions and soil chemistry,
inorganic phosphate can outcompete glyphosate for sorption
sites in soil particles; this might affect the exposure of plants
to glyphosate (Wang et al. 2005, Padilla and Selim 2019).
Given the extensive use of both GBHs and inorganic phos-
phate fertilizers, it is essential to comprehensively understand
how glyphosate residues—alone or in combination with phos-
phate fertilizers—affect factors associated with the health and
fitness of crop plants.
In addition, recent studies have indicated that the risks of
GBHs for nontarget organisms are not limited to the effects
of active ingredient glyphosate alone. GBHs contain various
co-formulants, in particular surfactants that enhance the up-
take and translocation of active ingredients in weeds and pos-
sibly other organisms. Some of the co-formulants may bemore
toxic than glyphosate (Mesnage et al. 2019, Straw et al. 2021).
The list of co-formulants added to commercial herbicide for-
mulations is increasing, but the current safety regulations and
laws specify that only the active ingredients are required to
be tested for their toxicity to nontarget organisms. Informa-
tion on the final composition of commercial products is de-
fective because the information of co-formulants is treated as
confidential trade secrets (Mesnage et al. 2019). Researchers
can analyze the glyphosate residues in soil and nontarget or-
ganisms, but the residues and degradation time of the co-
formulants remain unknown. This complicates the research
on GBH effects on nontarget organisms.
Recent studies suggest that microorganisms may be one of
the nontarget organisms at risk with GBH use, since the shiki-
mate pathway is present in most fungi and bacteria (Bent-
ley and Haslam 1990, Shehata et al. 2013, Leino et al. 2021,
Rainio et al. 2021). This is corroborated by studies on the
indirect negative effects of glyphosate residues on plants via
soil microbiota, as well as microbes associated with plants
and animals (Druille et al. 2016, Helander et al. 2018, 2019,
Motta et al. 2018, Székács and Darvas 2018, van Bruggen
et al. 2018, Ruuskanen et al. 2020, 2023). Plants and their
associated microbiomes are increasingly being considered as
co-evolving ecological entities, or holobionts, which can form
a unit of selection in evolutionary processes (Saikkonen et
al. 2020). The endophytic microbiota is tightly linked to the
host plant, and changes in its composition and the abun-
dance of individual microbial species can determine plant
health (Nissinen et al. 2012). The intrinsic susceptibility of
microbes to glyphosate can be determined based on the struc-
tural characterization of the EPSPS enzyme (Mathew et al.
2022). In fact, recent bioinformatics analyses have revealed
that a large proportion of prokaryotes (including bacteria
and archaea) are susceptible to glyphosate (Leino et al. 2021,
Rainio et al. 2021), but no study has so far reported the pro-
portion of glyphosate-sensitive and glyphosate-resistant mi-
crobes for any plant microbiome. Furthermore, in order to
understand the potential risks of GBH use to the health and
fitness of the plant holobiont, it is crucial to determine whether
plant-associated glyphosate-sensitive and glyphosate-resistant
endophytic microbes respond differently when exposed to
glyphosate residues following the GBH applications.
Here, the effects of GBH application and phosphate fertil-
izer individually and in combination were studied on plant
performance and endophytic microbiomes in garden straw-
berry (Fragaria × ananassa). Endophytic bacteria and fungi
of commercially important strawberry varieties have been
shown to promote plant growth, as well as fruit quantity
and quality. For instance, bacterial endophytes isolated from
strawberry meristematic tissues (Dias et al. 2009), strawberry
fruits (de Melo Pereira et al. 2012), and wild strawberry
varieties (de Andrade et al. 2019) were found to produce
growth hormones, solubilize phosphate, and fix nitrogen—
all of which are important for improving plant performance.
However, there is no information about the potential impacts
GBH use might have for strawberry microbiota. Therefore,
an experimental study was conducted to determine whether
glyphosate residues in soil following the application of field-
realistic doses of GBH alone, or in combination with phos-
phate fertilizer, alter strawberry performance and endophytic
microbiota community composition of the roots and leaves.
High-throughput sequencing of 16S rRNA gene and inter-
nal transcribed spacer (ITS) region was performed to charac-
terize the taxonomic community composition of endophytic
bacteria and fungi, respectively, in garden strawberries grown
in GBH-treated and untreated soil, with and without phos-
phate fertilizer. In addition to analyzing the potential dif-
ferences in structure and composition of endophytic bacte-
rial and fungal communities in response to exposure to GBH
alone and in combination with phosphate fertilizer, the indi-
cator species analysis was used to identify “indicator taxa”
characteristic to these treatments (Mouillot et al. 2002, Leg-
endre 2013). Further, a novel bioinformatics tool, based on
the type of EPSPS enzyme produced in the identified endo-
phytic strawberry microbes, was utilized to determine the
predisposition of the microbes to glyphosate. These results
were compared with the alterations in microbiome detected
in the field study to understand whether glyphosate-sensitive
and glyphosate-resistant microbes respond differently when
they are exposed to glyphosate residues. More specifically,
the following hypotheses were tested: (1) GBH application
affects the strawberry microbiome and decreases the abun-
dance of glyphosate-sensitive microbes. (2) The effects of GBH
application are systemic and affect the entire plant vascular
system, but the root microbiome is more strongly affected
than the leaf microbiome because roots are directly in con-
tact with soil. In addition, we examined whether the ap-
plication of inorganic phosphate fertilizer alters the effects
of GBH application on plant performance and endophytic
microbiota.
Materials and methods
Field experiment
The garden strawberry (Fragaria × ananassa) cultivar
“Bounty” (Canada) was used to study the combined ef-
fects of GBH application and phosphate fertilizer on straw-
berry performance and the associated endophytic micro-
biota. To this end, a field experiment was conducted at
Ruissalo Botanical Garden (University of Turku) in south-
western Finland (60◦26′N, 22◦10′E). In 2019, the mean
annual temperature and precipitation in the area were
7.4◦C and 741mm, respectively (http://www.fmi.fi/en). The
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
GBH effect on strawberry microbiota 3
experimental field (25m×50m) was established in 2013 by
mixing sand and peat with existing clay soil (pH 7.1) before
tilling at a depth of 15 cm and fencing to exclude larger ver-
tebrates from the area. The field was divided into alternat-
ing control and GBH treatment plots (10 plots each, mea-
suring 23m×1.5m), with 1.5-m buffer strips between the
plots. Twice a year, the plots were tilled to a depth of 5 cm
with a hand-held rotary tiller. Following this, the control plots
were treated with tap water (5 L per plot), and the GBH plots
were treated with Roundup Gold (glyphosate concentration:
450 gL–1, CAS: 3864-194-0, application rate: 6.4 L ha−1 in
5 L of tap water per plot) using a hand-operated pressure
tank with a plastic hood over the sprinkler tip to prevent the
glyphosate from spreading outside the treatment plots. For the
experiment, each of the 20 plots were divided into two: half
of the plot was treated with phosphate fertilizer (Yara Ferti-
care; application rate: 80 g in 10L of tap water), and the other
half was treated with tap water. This resulted in four differ-
ent treatments: control (C), phosphate fertilizer treatment (P),
GBH treatment (G), and combined GBH and phosphate fertil-
izer treatment (G+ P). All the plots were hand-weeded several
times during the growing season to prevent plant competition.
The plots were watered with a sprinkler located in the middle
of the experimental field when needed throughout the grow-
ing season. The buffer strips between the plots were mowed
several times during each field season to minimize weed in-
vasion. A more detailed description about the establishment
of the experimental field and its management, including GBH
treatments, can be found in Helander et al. (2019).
The plant material used in this experiment was propagated
from strawberry plants that were growing in the same exper-
imental field in 2018. In September 2018, runners were col-
lected and planted in small pots containing potting substrate
(Kekkilä Taimimulta). After the runners had rooted, they were
kept in the greenhouse until spring 2019 at a maximum tem-
perature of +10◦C under ambient light. Plantlets propagated
from the runners originating from each treatment (C, P,G, and
G + P) were randomly planted back under the same treatment
conditions. In early June 2019, 3weeks after GBH applica-
tion and 1month after phosphate application, 12 equal-sized
plantlets were planted in each plot, i.e. 6 plantlets for each
treatment. Out of 240 strawberry plantlets, 200 survived un-
til the end of the growing season.
To study the combined effects of GBH application and
phosphate fertilizer on strawberry growth and reproduction,
the number of leaves was counted and considered as an indi-
cator of plant growth. Since very few strawberries were flow-
ering, but almost 90% were producing runners, vegetative re-
production was estimated by measuring the length of the pro-
duced runners in mid-August 2019. At the same time, soil was
sampled from each subplot for analysis of glyphosate, and its
degradation product aminomethylphosphonic acid (AMPA).
Samples consisted of soil material that was ∼2.5 cm in diame-
ter and 5 cm in depth and was air-dried before extraction and
pooled according to treatment. Glyphosate and AMPA were
extracted with aqueous acidified methanol followed by anal-
ysis via liquid chromatography coupled to mass spectrometry
at GroenAgro (agrocontrol.nl).
Sampling for microbiota analysis
For microbiome analysis, one fully expanded, healthy leaf was
collected and ∼100mg of root material was dug up from one
randomly selected strawberry individual in each treatment
plot in early August 2019. Altogether, 10 replicates of each
of the leaf and root per treatment was collected; thus, alto-
gether 40 root and 40 leaf samples were collected. In the field,
samples were placed in sterile plastic bags, stored on ice, and
brought immediately to the laboratory for further processing.
The samples were washed with tap water, air-dried,
weighed, and surface sterilized in a laminar air flow hood.
Approximately 100mg of root and leaf tissue was washed in
70% ethanol (1min) and then 3% sodium hypochlorite so-
lution (3min), rinsed thrice with sterile distilled water (1min
per rinse), and air-dried. The samples were then transferred
to 2-mL microcentrifuge tubes and stored at −80◦C until fur-
ther processing. A negative control was prepared by plating
100μL of the water from the last rinse on Reasoner’s 2 Agar
(R2A) plate. Plates were kept at room temperature to monitor
microbial growth.
DNA extraction
Frozen samples were homogenized twice on a bead mill ho-
mogenizer (Bead Ruptor 96 Well Plate Homogenizer, OMNI
International US) for 30 s. DNA extraction was carried out
using the Invisorb Spin Plant Mini Kit (STRATEC Biomedical
AG, Germany) according to the manufacturer’s instructions.
The DNA concentration was adjusted to 30 ngμL−1.
16S rRNA gene-targeted PCR for bacterial
community analysis
The variable regions V6–V8 of the bacterial 16S rRNA gene
were amplified by the nested PCR approach. In the first
round of PCR, the primers 799F (AACMGGATTAGATAC-
CCKG; Chelius and Triplett 2001) and 1492R [GGYTAC-
CTTGTTACGACTT; modified from Lane (1991)] were used
to eliminate chloroplast amplification. The PCR reaction so-
lution contained 30 ng DNA, 1× PCR buffer, 0.2mM dNTPs,
0.3μM of each primer, and 2000UmL−1 of GoTaq DNA
polymerase (Promega, WI, USA) in a 30-μL reaction volume.
This was followed by PCR with the primers M13-1062F (TG-
TAAACGACGGCCAGTGTCAGCTCGTGYYGTGA) (Ghy-
selinck et al. 2013,Mäki et al. 2016) and 1390R (ACGGGCG-
GTGTGTRCAA) (Zheng et al. 1996) for sample bar-
coding. The PCR components were the same as in the
first PCR, except for the template, which was 1:10 di-
lution of the first PCR product. This was followed by
a third round of PCR with IonA-barcode-M13 primers
(Mäki et al. 2016) 1390R-P1 for sample tagging. The PCR
reactions were the same as those for the first two PCRs, ex-
cept that a 1:1 dilution of the second PCR product was used as
the template. The protocol for the first PCR were as follows:
one cycle of 3min of initial denaturation at 95◦C; 35 cycles
each of denaturation at 95◦C for 45 s; annealing at 54◦C for
45 s; extension at 72◦C for 1min; and final extension at 72◦C
for 5min. The same protocol was followed for the second and
third PCRs, except the number of cycles were reduced to 25
and 8 cycles, respectively. The amplicons were analyzed on
1.5% agarose gel.
ITS region-targeted PCR for fungal community
analysis
The ITS regions of fungal endophytes were amplified us-
ing the primers M13-ITS7F (TGTAAAACGACGGCCAGT-
GTGARTCATCGAATCTTTG) and ITS4R (TCCTCCGCT-
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
4 Mathew et al.
TATTGATATGC) (Ihrmark et al. 2012). The 30-μL PCR reac-
tion mixture contained 30 ng of sample DNA, 1× PCR buffer,
0.2mM dNTPs, 0.3μM of each primer, and 1250UmL−1
GoTaq DNA Polymerase (Promega, WI, USA). The ampli-
fication was initial denaturation at 95◦C for 5min, 35 cy-
cles of denaturation, annealing at 55◦C for 30 s, and ex-
tension at 72◦C for 1min, and followed by a final exten-
sion at 72◦C for 7min. The second round of PCR was per-
formed using a 1:10 dilution of the first PCR product as
template with IonA-barcode-M13 as the forward primer and
ITS4-P1 (CCTCTCTATGGGCAGTCGGTGATTCCTCGCT-
TATTGATATGC) as the reverse primer. All the other PCR
components were the same as those used in the first round of
PCR, but denaturation, annealing, and extension steps were
repeated for only 8 cycles instead of 35. The amplicons were
analyzed on 1.5% agarose gel.
Library preparation and sequencing
The PCR products were quantified on the Agilent 2100 Bio-
analyzer system, and 30 ng of target amplicons (16S rRNA
gene/ITS region) of each sample were pooled to prepare the
library. To eliminate plant mitochondrial amplicons and small
fragments, amplicons of size 350–550 bp were collected by
size fractionation in Pippin Prep (Sage Science, MA, USA)
using a 2% agarose gel cassette (Marker B). The purified li-
braries were subjected to emulsion PCR (Ion OneTouch™ 2
System, ThermoFisher Scientific Ltd) and sequenced on Ion
314™ Chip v2 in the Ion Personal Genome Machine™ (Ther-
moFisher Scientific Ltd).
Bioinformatics and statistical analyses
The sequence reads were processed using CLC Genomics
Workbench 11.0 with a Microbial Genomics Module (Qi-
agen, Denmark). Low-quality sequences were filtered, and
high-quality reads were trimmed to 250 bp and aligned. Reads
with 97% sequence identity were clustered into operational
taxonomic units (OTUs). Taxonomic classification of OTUs
was done using the reference databases RDP 16S rRNA train-
ing set 16 for bacteria and UNITE Fungal ITS trainset 7.1
for fungi (https://rdp.cme.msu.edu) (Wang et al. 2007). To
reduce noise from randomly occurring OTUs, low-abundant
OTUs with combined abundance of <10 reads were elimi-
nated. All statistical analyses were conducted separately for
bacterial and fungal OTU datasets.
The effect of tissue type (root or leaf) and treatment (con-
trol, phosphate, GBH, or GBH with phosphate) on overall
structure and composition of bacterial and fungal commu-
nities was analyzed using permutational multivariate analy-
sis of variance (PERMANOVA) (Anderson 2017) based on
Bray–Curtis dissimilarity matrix of square-root transformed
data. The results were visualized using principal coordinate
analysis (PCoA). To evaluate the diversity of bacterial and
fungal communities, species richness and Shannon diversity
index (H′) were calculated using Univariate Diversity Indices
(DIVERSE). The above statistical analyses were conducted us-
ing the PRIMER 7+ PERMANOVA software (primer-e.com).
To assess differences in microbial diversity based on treat-
ment and tissue type, we performed ANOVA and glht post-
hoc tests on the obtained Shannon values using the soft-
ware R 3.6.1 and the package “multcomp” (Hothorn et al.
2020).
Estimation of potential sensitivity to glyphosate
To determine the predisposition of the identified endophytic
strawberry microbes to glyphosate, we utilized a novel bioin-
formatics tool that is based on the type of EPSPS enzyme pro-
duced in microbes (Mathew et al. 2022). EPSPS can be classi-
fied as potentially sensitive or resistant to glyphosate based on
amino acidmarkers present at its active site (Leino et al. 2021).
However, in this study, the microbiome was studied based on
an analysis of the 16S rRNA gene (bacteria) and ITS region
(fungi). Thus, there was no information about the type of EP-
SPS sequence or the exact bacterial or fungal strain or species.
Therefore, a probabilistic approach, based on the method of
Mathew et al. (2022), was used to estimate the potential sen-
sitivity of the strawberry endophytic microbe to glyphosate.
First, bacterial species were mapped onto two precomputed
datasets of EPSPS sequences available at EPSPSClass web-
server at this link: https://ppuigbo.me/programs/EPSPSClass
(Leino et al. 2021). Previous studies have demonstrated high
taxonomic conservation of the EPSPS sequence and potential
sensitivity to glyphosate (Rainio et al. 2021).Accordingly, bac-
terial and fungal species were mapped onto EPSPS sequences
from the ATGC (Kristensen et al. 2017) and PFAM (El-Gebali
et al. 2019) databases, respectively. A probabilistic score of
sensitivity to glyphosate ranging between 0 (resistant: no sen-
sitive sequences to glyphosate were found) and 1 (sensitive:
all known sequences in the taxonomic group were sensitive to
glyphosate) was calculated. Evidently, there is a range of in-
between values, i.e. taxonomic groups that contain sequences
that are sensitive, resistant, and unknown. In order to account
for this, a cut-off of <0.2 and >0.8 was used to indicate re-
sistance and sensitivity, respectively.
The distribution of potentially glyphosate-sensitive and
glyphosate-resistant bacterial and fungal OTUs under differ-
ent treatment conditions compared to the control conditions
was estimated using the z-test and kernel density distribution
with Excel and the R package “density,” respectively.
Indicator species analysis
Indicator species analysis was used to test the differences in
abundance and frequency of each OTU in the samples col-
lected from different treatments. While PERMANOVA fo-
cuses on the entire microbial community structure, indicator
species analysis provides a tool to analyze the potential effects
of glyphosate residues alone and in combination with phos-
phate fertilizer on individual OTUs in order to identify “indi-
cator taxa”for different treatments (e.g. Fortunato et al. 2013,
Tedjo et al. 2016). The filtered OTU dataset was analyzed with
indicator species analysis with the R package “labdsv”and the
“ìndval” test (Dufrêne and Legendre 1997) to identify OTUs
specific to each of the treatments (control, phosphate, GBH,
and GBH with phosphate). The indicator values range from
0 to 1, with better indicators having higher values for the re-
spective treatment and P-value < 0.05. Both abundance and
frequency are used to determine whether an OTU is charac-
teristic for strawberry root or leaf under a given treatment
condition.
Plant growth and vegetative reproduction
The combined effects of GBH application and phosphate
fertilizer on strawberry growth and vegetative reproduction
were analyzed by using generalized and general linear mixed
models, respectively (PROC GLIMMIX, SAS 9.4). Poisson
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
GBH effect on strawberry microbiota 5
Figure 1. Order-level taxonomic distribution of endophytic microbial communities in the roots and leaves of garden strawberry (Fragaria × ananassa).
Bacterial communities (a) and fungal communities (b) in the control (C), GBH (G), phosphate fertilizer (P), and GBH with phosphate fertilizer (G + P)
treatment groups. Highly abundant orders are presented in the figure, and minor taxa are grouped together as “Others.”
distribution and log link function were used in the strawberry
growth model. Strawberry growth was measured in terms of
the number of leaves at the end of the growing season, and
vegetative reproductionwasmeasured as the length of the run-
ners produced. In the models, treatment was used as a fixed
factor and plot as a random factor to eliminate the potential
effects of environmental variation in the experimental field. In
addition, to control for the effect of plant size on vegetative re-
production, the number of leaves was included as a covariate
in the model for analyzing the combined effects of glyphosate
residues and phosphate fertilizer on vegetative reproduction.
Finally, the normality and equality of variances of the resid-
uals were assessed by visual examination and Levene’s test,
respectively.
Results
Glyphosate residues in soil
The levels of glyphosate and its degradation product AMPA
were markedly higher in soil samples from the GBH treat-
ment than in the control treatment samples, irrespective of
whether phosphate was added. In soil samples from the con-
trol and phosphate treatments, glyphosate and AMPA con-
centrations were below the quantification limit (0.01mgkg−1
for glyphosate and 0.05mgkg−1 for AMPA). The glyphosate
concentration was 0.06mgkg−1 in the GBH treatment sam-
ples and 0.07mgkg−1 in the GBH with phosphate treatment
samples, and the AMPA concentration was 1.9mgkg−1 in the
GBH treatment samples and 1.8mgkg−1 in the GBH with
phosphate treatment samples.
Taxonomic distribution of microbial communities
Regarding taxonomic distribution of the strawberry bacterial
community, 140 912 bacterial sequence reads were clustered
into 911OTUs belonging to 17 phyla, 65 orders, and 106 fam-
ilies. The main bacterial phyla found in the strawberry endo-
phytic community were Proteobacteria and Bacteroidetes. The
major orders in the strawberry root samples were Burkholde-
riales, Xanthomonadales, Rhizobiales, and Pseudomonadales,
with Burkholderiales showing higher relative abundance in
root samples from the phosphate and GBH with phosphate
treatments (Fig. 1a). In the leaf bacterial community, 95%was
distributed among the orders Burkholderiales, Sphingomon-
adales, Rhizobiales, and Cytophagales (Fig. 1a).
Regarding taxonomic distribution of the fungal community,
196 410 sequence reads were clustered into 363 OTUs repre-
senting 5 phyla, 52 orders, and 90 families. The root fungal
communities belonged to the phyla Ascomycota, Basidiomy-
cota, and Glomeromycota, while the leaf communities were
dominated by the phyla Ascomycota and Basidiomycota. Fur-
ther, root fungal communities mainly belonged to the Pleospo-
rales, Helotiales, Sebacinales, Auriculariales, and Glomerales
orders, while the orders Capnodiales, Pleosporales, Tremel-
lales, Sporidiobolales, and Filobasidiales were dominant in the
leaf communities (Fig. 1b). The relative abundance of Sebaci-
nales were lower and that of Auriculariales were higher in the
root communities as a consequence of GBH treatments in soil
(Fig. 1b).
Impact of glyphosate and phosphate on the
diversity and composition of microbial
communities
The Shannon diversity index for measuring the richness and
uniformity of bacterial and fungal communities showed that
bacterial communities present in the roots were more diverse
than those in the leaves. Phosphate treatment resulted in an
increase in the diversity of bacterial communities in the roots
(Table 1, Supplementary Fig. S1), while neither tissue type nor
treatment had an effect on the diversity of fungal communities
(Table 1). The structures of both bacterial and fungal com-
munities were significantly different between the roots and
leaves (P= 0.001 according to PERMANOVA,Table 2).Com-
parison of root bacterial communities between the treatment
(Table 3) with PCoA showed that phosphate fertilizer treat-
ment significantly shaped bacterial communities in the roots
(Table 3, Fig. 2a) but not in the leaves (Table 3, Fig 2b). In
contrast, GBH treatment did not have a significant impact on
bacterial community structure in either root or leaf tissues (Ta-
ble 2). None of the treatments impacted the structure of the
fungal communities (Table 2, Fig. 2c and d).
In silico analysis of the glyphosate sensitivity of
microbial community members
The dataset contains 389 bacterial OTUs that are potentially
sensitive to glyphosate, 376 OTUs that are potentially resis-
tant, and 147 OTUs with unknown sensitivity status. The ra-
tio of potentially glyphosate-sensitive bacteria to potentially
glyphosate-resistant bacteria indicate higher abundance of
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
6 Mathew et al.
Table 1. Results of a linear model of the effects of tissue type (root or leaf) and treatment (control, GBH, phosphate fertilizer, or GBH with phosphate
fertilizer) on the Shannon diversity of endophytic bacterial and fungal communities of garden strawberry (Fragaria × ananassa).
Response variable Explanatory variable F Df P
Bacteria GBH 0.267 1/68 0.607
Phosphate 1.260 1/68 0.266
Tissue (root/leaf) 344.93 1/68 <0.001
GBH × phosphate 0.023 1/68 0.558
GBH × tissue 0.044 1/68 0.477
Phosphate × tissue 8.265 1/68 0.005
GBH × phosphate × tissue 0.017 1/68 0.656
Fungi GBH 0.092 1/68 0.408
Phosphate 0.039 1/68 0.473
Tissue (root/leaf) 0.051 1/68 0.342
GBH × phosphate 0.027 1/68 0.488
GBH × tissue 0.002 1/68 0.863
Phosphate × tissue 0.014 1/68 0.617
GBH × phosphate × tissue 0.056 1/68 0.320
Table 2.Results of PERMANOVA analysis of the effect of plant tissue (root or leaf) and treatment (control, GBH, phosphate fertilizer, or GBHwith phosphate
fertilizer) on the composition of endophytic bacterial and fungal communities in garden strawberry (Fragaria × ananassa).
Response variable Source df Pseudo-F P Unique perms
Bacteria Tissue 1 42.92 0.001 998
Treatment 3 1.54 0.001 996
Tissue × treatment 3 1.49 0.001 995
Fungi Tissue 1 65.46 0.001 999
Treatment 3 0.92 0.707 998
Tissue × treatment 3 1.03 0.370 997
Table 3. Results of pairwise t-test for analysis of the difference in the
composition of endophytic bacterial communities between different treat-
ments in the roots and leaves of garden strawberry (Fragaria × ananassa).
Tissue Treatment t P Unique perms
Root C vs. G 1.020 0.357 993
C vs. P 1.484 0.001 989
C vs. G + P 1.629 0.001 998
P vs. G 1.484 0.001 988
P vs. G + P 1.064 0.166 988
G + P vs. G 1.607 0.001 994
Leaf C vs. G 1.008 0.358 986
C vs. P 1.076 0.181 990
C vs. G + P 0.999 0.39 990
P vs. G 1.078 0.187 991
P vs. G + P 0.936 0.667 991
G + P vs. G 1.011 0.365 992
C = control, G = GBH, P = phosphate fertilizer, and GBH + P = GBH with
phosphate fertilizer.
potentially glyphosate-sensitive bacteria in the roots than in
the leaves (Supplementary Table S2). The Kernel density plot
indicated that the addition of phosphate fertilizer tended to
affect the abundance of bacterial OTUs belonging to both
potentially glyphosate-sensitive and glyphosate-resistant com-
munities in the roots (Fig. 3a). A z-test confirmed a signif-
icant increase in the abundance of potentially glyphosate-
resistant and glyphosate-sensitive bacterial OTUs in root sam-
ples from both phosphate fertilizer treatments compared to
the control treatment. In the leaves, no significant difference
in the glyphosate sensitivity of OTUs was observed between
the treatments (Fig. 3b, Supplementary Table S1).
Analysis of the EPSPS enzyme in fungal communi-
ties showed that 212 OTUs were potentially sensitive to
glyphosate, while the remaining 156 OTUs were unclassified.
Thus, the findings do not indicate any impact of GBH or phos-
phate fertilizer on glyphosate sensitivity of the fungal commu-
nity.
Impact of phosphate and glyphosate on bacterial
and fungal community members
To identify OTUs associated with the effects of glyphosate
residues in soil, indicator species analysis was used to test for
tissue-specific differences in the abundance and/or frequency
of bacterial and fungal OTUs between the control and GBH
treatment samples, and between the phosphate and GBHwith
phosphate treatment samples (Table 4). In the roots, eight bac-
terial OTUswere enriched in the control samples,where five of
these OTUs were classified as potentially glyphosate resistant.
Six bacterial OTUs were enriched in the GBH treatment sam-
ples, out of which five were classified as potentially glyphosate
resistant (Table 4). In the leaves, seven bacterial OTUs were
enriched in the control treatment samples, where three were
classified as potentially glyphosate sensitive represent the bac-
terial genus Hymenobacter. None of the leaf bacterial OTUs
were enriched in the GBH treatment samples (Table 4).
In the comparison between phosphate and GBHwith phos-
phate treatments, 11 bacterial OTUs were enriched in root
samples from phosphate treatment. A total of 18 bacterial
OTUs were enriched in root samples from the GBH with
phosphate treatment: 10 were classified as glyphosate resis-
tant, 5 were classified as glyphosate sensitive, and 3 were un-
classified (Table 4). OTUs belonging to Burkholderiales were
enriched in both GBH and GBH with phosphate treatment
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
GBH effect on strawberry microbiota 7
Figure 2. Results of PCoA showing the impact of different treatments on the structure of microbial communities in the endosphere of garden
strawberry (Fragaria × ananassa). The GBH (G) (), phosphate fertilizer (P) (), and GBH with phosphate fertilizer (G + P) () treatments are compared
with control treatment () in terms of their effects on bacterial communities in the roots (a), bacterial communities in the leaves (b), fungal communities
in the roots (c), and fungal communities in the leaves (d).
(Supplementary Table S3). In the leaves, one glyphosate-
resistant and one glyphosate-sensitive bacterial OTU were
enriched in the phosphate treatment samples, while two
glyphosate-sensitive bacterial OTUswere enriched in the GBH
with phosphate treatment samples (Table 4). Further details
on taxonomic classification and potential glyphosate sensitiv-
ity of bacterial indicator OTUs for each treatment is provided
in Supplementary Table S3.
In the comparison of root fungal OTUs between con-
trol and GBH treatment, 11 fungal OTUs were enriched
in the control treatment samples. A total of 6 out of the
11 OTUs enriched in the control treatment samples be-
longed to the order Sebacinales, five of them were clas-
sified as potentially glyphosate sensitive. Two potentially
glyphosate-sensitive fungal OTUs representing Tetracladium
(Pleosporales) and Paraphoma (Agaricales) genera were en-
riched in the GBH treatment samples (Table 4). In the
leaves, two of the three fungal OTUs enriched in the con-
trol samples belonged to genus Leucosporidium. Regard-
ing the comparison between phosphate treatment and GBH
with phosphate treatments in roots, three fungal OTUs
were enriched in the phosphate treatment samples. Two
of the OTUs were classified as potentially glyphosate sen-
sitive and one had unknown sensitivity status. Two fun-
gal OTUs of unknown sensitivity status were enriched in
the GBH with phosphate treatment samples (Table 4). No
indicator species were identified for leaf fungal OTUs in
any of these treatments (Table 4). Details on fungal indica-
tor taxa for each treatment are provided in Supplementary
Table S4.
Plant growth and vegetative reproduction
Overall, experimental plants had 10 ± 0.3 leaves and
68 ± 4 cm runners. None of the treatments affected straw-
berry size, as indicated by the number of leaves, or vegetative
reproduction (plant size: F3, 178 = 0.22, P = 0.8814; vege-
tative reproduction: F3, 173 = 1.76, P = 0.1561). The effect
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
8 Mathew et al.
Figure 3. Kernel density distribution of the abundances of glyphosate-sensitive and glyphosate-resistant endophytic bacterial OTUs according to
treatment. The abundances of glyphosate-sensitive (S) and glyphosate-resistant (R) endophytic bacterial OTUs in the roots (a) and leaves (b) of garden
strawberry (Fragaria × ananassa) are shown for the control (C), GBH (G), phosphate fertilizer (P), and GBH with phosphate fertilizer (G + P) treatments.
Table 4. Summary of indicator species analyses indicating the number of
OTUs with potential glyphosate sensitivity (S), potential glyphosate resis-
tance (R), and unknown glyphosate sensitivity status (U) for different treat-
ments and tissues.
Number of indicator species
C vs. G P vs. G + P
Bacteria Roots S 2 1 7 5
U 1 0 1 3
R 5 5 3 10
Leaves S 4 0 1 2
U 0 0 0 0
R 3 0 1 0
Fungi Roots S 7 2 2 0
U 4 0 1 2
R 0 0 0 0
Leaves S 1 0 0 0
U 2 0 0 0
R 0 0 0 0
The detailed results of indicator species analysis are provided in Supplemen-
tary Tables S3 and S4. Comparisons were performed between control (C)
and GBH (G), and phosphate fertilizer (P) and GBH with phosphate fertil-
izer (G + P) treatments.
of plant size on runner production differed across treatments,
as indicated by the significant treatment–plant size interaction
(F3, 173 = 3.07, P= 0.0291). Plant size and runner production
were not correlated with each other in the GBH treatment,
but for all the other treatments, they were positively corre-
lated (control: r = 0.49, P = 0.0003, n = 50; GBH treatment:
r = 0.01, P = 0.9235, n = 49; phosphate treatment: r = 0.55,
P< 0.0001, n= 50; GBHwith phosphate treatment: r= 0.43,
P= 0.0017, n= 50). These findings indicate that larger plants
produced more runners under all treatment conditions, except
for GBH treatment.
Discussion
GBH application in soil did not cause shifts in the diver-
sity or the overall community structure of strawberry endo-
phytic bacteria or fungi and did not affect plant performance
or vegetative reproduction. Similar to these findings, GBH
application had no effect on the richness of endophytic mi-
crobial communities or the composition of the endophytic
fungal community of a perennial weed (Ramula et al. 2022)
or root-associated soil microbial communities of corn and
soybean (Kepler et al. 2020), although other studies have
reported shifts in soil- and plant-associated microbes in re-
sponse to glyphosate residues and/or GBH applications [re-
viewed by van Bruggen et al. (2018)]. Despite the finding that
GBH application did not have effect on the overall community
structure of strawberry microbiota, GBH application shifted
the relative abundances and/or frequencies of several fungal
and bacterial OTUs in garden strawberry, especially in straw-
berry roots. This supports the hypothesis that the root mi-
crobiome is more affected than the leaf microbiome because
roots are directly in contact with soil and, thus, glyphosate
residues. A large proportion of the observed shifts in the abun-
dance of bacterial OTUs were related to the potential resis-
tance/sensitivity of their taxon to glyphosate. Interestingly, the
ratio of sensitive-to-resistant bacteria in different plant tissues
indicates that there are more bacterial taxa potentially sen-
sitive to glyphosate in strawberry roots than in strawberry
leaves. Although this might further explain the observed shifts
in abundance and/or frequency of endophytic bacteria in the
root, it is contrary to the prediction of Rainio et al. (2021) that
bacteria that are more exposed to glyphosate are intrinsically
more resistant to it. However, the lifestyle of plant-associated
endophytic bacteria might protect them from direct exposure
(Rainio et al. 2021). Lastly, results of this study show that the
use of phosphate fertilizer significantly shaped the root but not
leaf endophytic bacterial communities and increased the abun-
dance of both glyphosate-resistant and glyphosate-sensitive
OTUs, regardless of the presence of glyphosate residues in soil.
This agrees with previous findings that showed soil phospho-
rus levels affected rhizo- and endospheric microbiota in the
roots but did not affect the community structure of shoot bac-
teria (Finkel et al. 2019).
Indicator species analysis revealed that there was a shift in
the abundance and/or frequency of certain bacterial and fun-
gal OTUs in response to GBH treatments. Bacterial OTUs in
leaves that were enriched in the control and phosphate (i.e.
glyphosate residue free) treatment samples were mainly from
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
GBH effect on strawberry microbiota 9
the orders Rhizobiales (Alphaproteobacteria) and Cytopha-
gales (Bacteroidetes). Most of the Cytophagales OTUs were
classified under the genus Hymenobacter, which is known to
be a consistent member of the leaf core microbiome (Grady
et al. 2019, Ares et al. 2021). A clear taxonomical pattern
in the root bacterial OTUs in response to the GBH treat-
ment was not found. For instance, OTUs enriched in both the
GBH treatment and control treatment samples belonged to
the order Burkholderiales. The explanation behind this might
be that the sensitivity or resistance of EPSPS to glyphosate
can vary within the same bacterial taxon, even at the species
level, as reported previously (Leino et al. 2021). Remark-
ably, 83% and 56% of the root bacterial OTUs enriched in
the GBH treatment and GBH with phosphate fertilizer treat-
ment samples, respectively, were characterized as being po-
tentially glyphosate resistant, and only 17% and 28%, re-
spectively, were potentially glyphosate sensitive. This is in
line with the hypothesis that the abundance of glyphosate-
sensitive microbes is decreasing as a response to glyphosate
residues in soil. The observed loss in abundance of these po-
tentially glyphosate-sensitive bacterial OTUs might be caused
by the mechanism of action of glyphosate residues targeting
the EPSPS enzyme in the shikimate pathway and, thereby,
inhibiting the growth and/or reproduction of these bacterial
OTUs. On the other hand, bacteria may easily become resis-
tant to glyphosate as a result of a single mutation in the EP-
SPS active site, and this would result in a higher proportion of
glyphosate-resistant bacteria in glyphosate-exposed environ-
ments (Rainio et al. 2021).
Contrary to the low sensitivity of bacterial communities to
glyphosate, most fungi (92%) are known to be sensitive to
glyphosate (Leino et al. 2021). In line with this, 212 out of
369 fungal OTUs in this study were found to be potentially
glyphosate sensitive, while the remaining could not be classi-
fied.However, the response of the EPSPS protein to glyphosate
has been determined mainly in plants and bacteria, as both
have a unidomain EPSPS protein; in contrast, fungal EPSPS
contains multiple domains that can lead to variable responses
to glyphosate (Mathew et al. 2022). Accordingly, we found
potentially sensitive fungal OTUs that were enriched in the
GBH treatment samples, while other potentially sensitive fun-
gal OTUs were also enriched in other treatments.
Indicator species analysis of strawberry fungal OTUs
showed a consistent taxonomic trend at the order and family
levels in response to GBH treatments.Most root fungal OTUs
enriched in the control treatment samples represented the fun-
gal families Sebacinaceae and the arbuscular mycorrhizal fun-
gal (AMF) family Glomeraceae. This is in line with previous
studies that have reported that GBH application can decrease
AMF abundance as well as root colonization in different sys-
tems (Zaller et al. 2015, Druille et al. 2016, Helander et al.
2018). In leaves, two out of three fungal OTUs enriched in the
control treatment samples were from the order Leucosporid-
ium, which has been reported to be part of the phyllosphere
microflora of several plant species (Bálint et al. 2013,Wang et
al. 2016, Suryanarayanan and Shaanker 2021). Reduction in
the abundance of important plant-associated fungal taxa fol-
lowing GBH treatment may affect plant resilience to a multi-
tude of stressors (van Bruggen et al. 2021), but more directed
studies are needed to test this. However, since the exact EPSPS
sequences and, thus, potential sensitivity of several respond-
ing OTUs from our samples are not known, further empirical
and theoretical studies are needed to fine-tune the classifica-
tion of the EPSPS enzyme in fungi in order to determine more
precisely their susceptibility to glyphosate.
Phosphate and glyphosate are known to compete for the
same adsorption sites in soil. As a result, phosphate fertil-
izer applications have been demonstrated to increase the mo-
bility of glyphosate in soil and lead to increased uptake of
glyphosate by plant roots (Bott et al. 2011, Gomes et al.
2015). This has the potential to increase the harmful effects
of glyphosate on nontarget organisms. However, in the cur-
rent study, phosphate was not found to mediate the effects of
GBH residues on plant microbiota or performance. Phosphate
fertilizer treatment did not have effect on strawberry perfor-
mance, but it affected the diversity, community composition,
and abundance of bacterial OTUs in strawberry roots. In ac-
cordance with these findings, phosphate addition has been
shown to increase root bacterial diversity in two model plant
species (Bodenhausen et al. 2019). Furthermore, recent stud-
ies on Arabidopsis indicate that phosphate availability has an
effect on the colonization of roots by soil bacteria (Zuccaro
2020).
Given the extensive use of GBH in agriculture and horti-
culture, as well as glyphosate persistence in ecosystems (He-
lander et al. 2012, Maggi et al. 2020), both wild and culti-
vated plants are likely to be exposed to glyphosate residues.
In addition, the effects of various co-formulants on plants and
other nontarget organisms after GBH use cannot be ruled
out (Mesnage et al. 2019). An increasing number of stud-
ies are showing that the effects of commercial herbicide for-
mulations on nontarget organisms can be stronger than the
effects of the active ingredient alone (Helander et al. 2019,
Straw et al. 2021). Although our results fail to differentiate
whether the results are the outcome of the effect of the ac-
tive ingredient or co-formulants, the use of GBHs is justified
in experiments because it corresponds to the actual weed con-
trol. Changes in endophytic microbiota may function as a re-
liable indicator of persistent, environmental stressor-mediated
changes in plants. However, the results of this study did not
indicate any major changes in the overall endophytic micro-
biome mediated by soil GBH treatment. Still, certain bac-
terial and fungal taxa, especially in strawberry roots, re-
sponded to GBH treatments. Although these changes were
not reflected in the measured plant performance traits in this
short-term experiment, it remains to be answered whether
these GBH-driven alterations in strawberry microbiota may
affect other plant functions and, eventually, plant resilience.
This study is the first one to report the proportion of poten-
tially glyphosate-resistant and glyphosate-sensitive bacteria in
any plant species. Further studies are needed to understand
the glyphosate-resistance mechanisms for fungi and how dif-
ferences in sensitivity/resistance to glyphosate, together with
long-term exposure to GBH application, affect the compo-
sition and diversity of endophytic microbial communities.
Strong selection pressures due to the intensive GBH use might
cause rapid evolution of glyphosate-sensitive bacteria possibly
having consequences on plant–microbe interactions and thus
on plant health.
Acknowledgments
We thank Ida Palmroos, Lyydia Leino, and Lauri Heikkonen
for their assistance in the field experiment. This work was
supported by the Academy of Finland (Grant No. 324523 to
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
10 Mathew et al.
B.F. and Grant No. 311077 to M.H.) and the Maj and Tor
Nessling Foundation (Grant No. 201800048 to A.M.).
Supplementary data
Supplementary data is available at JAMBIO online.
Conflict of interest
No conflict of interest declared.
Author contributions
A.M. andM.H. conceived and designed the study. S.A.M. con-
ducted microbiome experiments and A.M. conducted plant
performance experiments. R.N., S.A.M., B.F., P.P., and A.M.
conducted data analysis. S.A.M., A.M., and B.F. drafted the
manuscript. All authors contributed to editing and proofread-
ing the manuscript.
Data availability
The datasets generated during the study are available from
the Dryad Digital Repository (https://doi.org/10.5061/dryad.
cvdncjt7k).
References
Anderson MJ. Permutational Multivariate Analysis of Variance (PER-
MANOVA). In: Balakrishnan N, Colton T, Everitt B, Piegorsch
W, Ruggeri F, Teugels JL (eds.), Wiley StatsRef: Statistics Refer-
ence Online. Chichester, UK: John Wiley & Sons Ltd., 2017, 1–15.
https://doi.org/10.1002/9781118445112.stat07841
Ares A, Pereira J, Garcia E et al. The leaf bacterial microbiota of female
and male kiwifruit plants in distinct seasons: assessing the impact of
Pseudomonas syringae pv. actinidiae. Phytobiomes J 2021;5:275–
87. https://doi.org/10.1094/PBIOMES-09-20-0070-R
Bai SH, Ogbourne SM. Glyphosate: environmental contamina-
tion, toxicity and potential risks to human health via food
contamination. Environ Sci Pollut Res 2016;23:18988–9001.
https://doi.org/10.1007/s11356-016-7425-3
Bálint M, Tiffin P, Hallström B et al. Host genotype
shapes the foliar fungal microbiome of balsam poplar
(Populus balsamifera). PLoS One 2013;8:e53987.
https://doi.org/10.1371/journal.pone.0053987
Benbrook CM. Trends in glyphosate herbicide use in the
United States and globally. Environ Sci Eur 2016;28:3.
https://doi.org/10.1186/s12302-016-0070-0
Bentley R, Haslam E. The shikimate pathway—a metabolic tree
with many branche. Crit Rev Biochem Mol Biol 1990;25:307–84.
https://doi.org/10.3109/10409239009090615
Bodenhausen N, Somerville V,Desirò A et al. Petunia- and Arabidopsis-
specific root microbiota responses to phosphate supplementation.
Phytobiomes J 2019;3:112–24. https://doi.org/10.1094/PBIOMES-
12-18-0057-R
Bøhn T, Cuhra M, Traavik T et al. Compositional differ-
ences in soybeans on the market: glyphosate accumulates in
Roundup Ready GM soybeans. Food Chem 2014;153:207–15.
https://doi.org/10.1016/j.foodchem.2013.12.054
Bott S, Tesfamariam T, Kania A et al. Phytotoxicity of glyphosate
soil residues re-mobilised by phosphate fertilisation. Plant Soil
2011;342:249–63. https://doi.org/10.1007/s11104-010-0689-3
Brito IP, Tropaldi L, Carbonari CA et al. Hormetic effects of glyphosate
on plants: glyphosate hormesis. Pest Manag Sci 2018;74:1064–70.
https://doi.org/10.1002/ps.4523
Cedergreen N. Is the growth stimulation by low doses of glyphosate
sustained over time? Environ Pollut 2008;156:1099–104.
https://doi.org/10.1016/j.envpol.2008.04.016
Chelius MK, Triplett EW. The diversity of archaea and bacteria in asso-
ciation with the roots of Zea mays L.Microb Ecol 2001;41:252–63.
https://doi.org/10.1007/s002480000087
de Andrade FM, de Assis Pereira T, Souza TP et al. Ben-
eficial effects of inoculation of growth-promoting bac-
teria in strawberry. Microbiol Res 2019;223-225:120–8.
https://doi.org/10.1016/j.micres.2019.04.005
de Jonge H, de Jonge LW, Jacobsen OH et al. Glyphosate sorp-
tion in soils of different pH and phosphorus content. Soil Sci
2001;166:230–8. https://doi.org/10.1097/00010694-200104000-
00002
de Jonge H, de Jonge LW. Influence of pH and solution composition
on the sorption of glyphosate and prochloraz to a sandy loam soil.
Chemosphere 1999;39:753–63. https://doi.org/10.1016/S0045-
6535(99)00011-9
de Melo Pereira GV, Magalhães KT, Lorenzetii ER et al. A multiphasic
approach for the identification of endophytic bacterial in strawberry
fruit and their potential for plant growth promotion. Microb Ecol
2012;63:405–17. https://doi.org/10.1007/s00248-011-9919-3
Dias ACF, Costa FEC, Andreote FD et al. Isolation of micropropa-
gated strawberry endophytic bacteria and assessment of their po-
tential for plant growth promotion. World J Microbiol Biotechnol
2009;25:189–95. https://doi.org/10.1007/s11274-008-9878-0
Druille M, García-Parisi PA, Golluscio RA et al. Repeated annual
glyphosate applications may impair beneficial soil microorganisms
in temperate grassland. Agric Ecosyst Environ 2016;230:184–90.
https://doi.org/10.1016/j.agee.2016.06.011
Dufrêne M, Legendre P. Species assemblages and indicator
species: the need for a flexible asymmetrical approach.
Ecol Monogr 1997;67:345–66. https://doi.org/10.1890/0012-
9615(1997)067[0345:SAAIST]2.0.CO;2
Duke SO, Powles SB. Glyphosate: a once-in-a-century herbicide. Pest
Manag Sci 2008;64:319–25.
El-Gebali S, Mistry J, Bateman A et al. The Pfam protein fam-
ilies database in 2019. Nucleic Acids Res 2019;47:D427–32.
https://doi.org/10.1093/nar/gky995
Finkel OM, Salas-González I, Castrillo G et al. The effects of
soil phosphorus content on plant microbiota are driven by the
plant phosphate starvation response. PLoS Biol 2019;17:e3000534.
https://doi.org/10.1371/journal.pbio.3000534
Fortunato C, Eiler A,Herfort L et al. . Determining indicator taxa across
spatial and seasonal gradients in the Columbia River coastal margin.
ISME J 2013;7:1899–911.
Fuchs B, Laihonen M, Muola A et al. A glyphosate-based herbi-
cide in soil differentially affects hormonal homeostasis and perfor-
mance of non-target crop plants. Front Plant Sci 2022b;12:787958.
https://doi.org/10.3389/fpls.2021.787958
Fuchs B, Saikkonen K,Helander M et al. Legacy of agrochemicals in the
circular food economy: glyphosate-based herbicides introduced via
manure fertilizer affect the yield and biochemistry of perennial crop
plants during the following year. Chemosphere 2022a;308:136366.
https://doi.org/10.1016/j.chemosphere.2022.136366
Ghyselinck J, Pfeiffer S, Heylen K et al. The effect of primer
choice and short read sequences on the outcome of 16S
rRNA gene based diversity studies. PLoS One 2013;8:e71360.
https://doi.org/10.1371/journal.pone.0071360
Gill JPK, Sethi N, Mohan A et al. Glyphosate toxicity for animals. En-
viron Chem Lett 2018;16:401–26. https://doi.org/10.1007/s10311-
017-0689-0
Gomes MP, Maccario S, Lucotte M et al. Consequences of phosphate
application on glyphosate uptake by roots: impacts for environ-
mental management practices. Sci Total Environ 2015;537:115–9.
https://doi.org/10.1016/j.scitotenv.2015.07.054
Grady KL, Sorensen JW, Stopnisek N et al. Assembly and seasonality of
core phyllosphere microbiota on perennial biofuel crops.Nat Com-
mun 2019;10:4135. https://doi.org/10.1038/s41467-019-11974-4
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
GBH effect on strawberry microbiota 11
Hébert MP, Fugère V, Gonzalez A. The overlooked impact
of rising glyphosate use on phosphorus loading in agri-
cultural watersheds. Front Ecol Environ 2019;17:48–56.
https://doi.org/10.1002/fee.1985
Helander M, Pauna A, Saikkonen K et al. Glyphosate residues in soil
affect crop plant germination and growth. Sci Rep 2019;9:19653.
https://doi.org/10.1038/s41598-019-56195-3
HelanderM, Saloniemi I,OmaciniM et al. Glyphosate decreases mycor-
rhizal colonization and affects plant-soil feedback. Sci Total Environ
2018;642:285–91. https://doi.org/10.1016/j.scitotenv.2018.05.377
Helander M, Saloniemi I, Saikkonen K. Glyphosate in
northern ecosystems. Trends Plant Sci 2012;17:569–74.
https://doi.org/10.1016/j.tplants.2012.05.008
Hothorn T, Bretz F, Westfall P et al. Multcomp: simultaneous inference
in general parametric models; 1.4–14. Vienna: CRAN, 2020.
Ihrmark K, Bödeker ITM, Cruz-Martinez K et al. New primers to am-
plify the fungal ITS2 region—evaluation by 454-sequencing of arti-
ficial and natural communities. FEMSMicrobiol Ecol 2012;82:666–
77. https://doi.org/10.1111/j.1574-6941.2012.01437.x
Kanissery RG, Welsh A, Sims GK. Effect of soil aeration
and phosphate addition on the microbial bioavailability
of carbon-14-glyphosate. J Environ Qual 2015;44:137–44.
https://doi.org/10.2134/jeq2014.08.0331
Kepler RM, Epp Schmidt DJ, Yarwood SA et al. Soil microbial
communities in diverse agroecosystems exposed to the herbi-
cide glyphosate. Appl Environ Microbiol 2020;86:e01744–19.
https://doi.org/10.1128/AEM.01744-19
Kristensen DM, Wolf YI, Koonin EV. ATGC database and
ATGC-COGs: an updated resource for micro- and macro-
evolutionary studies of prokaryotic genomes and protein
family annotation. Nucleic Acids Res 2017;45:D210–8.
https://doi.org/10.1093/nar/gkw934
Laitinen P, Rämö S,Nikunen U et al. Glyphosate and phosphorus leach-
ing and residues in boreal sandy soil. Plant Soil 2009;323:267–83.
https://doi.org/10.1007/s11104-009-9935-y
Lane DJ. 16S/23S rRNA sequencing. In: Stackebrandt E, GoodfellowM
(eds.),Nucleic Acid Techniques in Bacterial Systematics. New York:
Wiley, 1991, 115–75.
Legendre P. Indicator species: computation. In: Levin S (ed.), Encyclo-
pedia of Biodiversity. San Diego: Academic Press, 2013, 264–8.
Leino L, Tall T, Helander M et al. Classification of the glyphosate tar-
get enzyme (5-enolpyruvylshikimate-3-phosphate synthase) for as-
sessing sensitivity of organisms to the herbicide. J Hazard Mater
2021;408:124556. https://doi.org/10.1016/j.jhazmat.2020.124556
Maggi F, la Cecilia D, Tang FHM et al. The global environmen-
tal hazard of glyphosate use. Sci Total Environ 2020;717:137167.
https://doi.org/10.1016/j.scitotenv.2020.137167
Mäki A, Rissanen AJ, Tiirola M. A practical method for barcoding and
size-trimming PCR templates for amplicon sequencing. BioTech-
niques 2016;60:88–90. https://doi.org/10.2144/000114380
Mathew SA, Muola A, Saikkonen K et al. Quantification of the poten-
tial impact of glyphosate-based products on microbiomes. J Vis Exp
2022;179:63109. https://doi.org/10.3791/63109
Mesnage R, Benbrook C, Antoniou MN. Insight into the
confusion over surfactant co-formulants in glyphosate-
based herbicides. Food Chem Toxicol 2019;128:137–45.
https://doi.org/10.1016/j.fct.2019.03.053
Motta EVS, Raymann K, Moran NA. Glyphosate perturbs the gut mi-
crobiota of honeybees. Proc Natl Acad Sci 2018;115:10305–10.
https://doi.org/10.1073/pnas.1803880115
Mouillot D, Culioli J-M, Do Chi T. Indicator species analysis as a
test of non-random distribution of species in the context of marine
protected areas. Environ Conserv 2002;29:385–90.
Muola A, Fuchs B, Laihonen M et al. Risk in the circular food
economy: glyphosate-based herbicide residues in manure fertil-
izers decrease crop yield. Sci Total Environ 2021;750:141422.
https://doi.org/10.1016/j.scitotenv.2020.141422
Nissinen RM, Männistö MK, van Elsas JD. Endophytic bacte-
rial communities in three arctic plants from low arctic fell
tundra are cold-adapted and host-plant specific. FEMS Mi-
crobiol Ecol 2012;82:510–22. https://doi.org/10.1111/j.1574-
6941.2012.01464.x
Padilla JT, Selim HM. Interactions among glyphosate and phosphate
in soils: laboratory retention and transport studies. J Environ Qual
2019;48:156–63. https://doi.org/10.2134/jeq2018.06.0252
Parthasarathy A, Cross PJ, Dobson RCJ et al. A three-ring circus:
metabolism of the three proteogenic aromatic amino acids and their
role in the health of plants and animals. FrontMol Biosci 2018;5:29.
https://doi.org/10.3389/fmolb.2018.00029
Rainio MJ, Ruuskanen S, Helander M et al. Adaptation of bacte-
ria to glyphosate: a microevolutionary perspective of the enzyme
5-enolpyruvylshikimate-3-phosphate synthase. Environ Microbiol
Rep 2021;13:309–16. https://doi.org/10.1111/1758-2229.12931
Ramula S, Mathew SA, Kalske A et al. Glyphosate residues al-
ter the microbiota of a perennial weed with a minimal indi-
rect impact on plant performance. Plant Soil 2022;472:161–74.
https://doi.org/10.1007/s11104-021-05196-1
Ruuskanen S, Fuchs B, Nissinen R et al. Ecosystem consequences of
herbicides: the role of microbiome. Trends Ecol Evol 2023;38:35–
43. https://doi.org/10.1016/j.envpol.2020.115108
Ruuskanen S, Rainio MJ, Gómez-Gallego C et al. Glyphosate-
based herbicides influence antioxidants, reproductive hor-
mones and gut microbiome but not reproduction: a long-
term experiment in an avian model. Environ Pollut 2020;266:
115108.
Saikkonen K, Nissinen R, Helander M. Toward comprehen-
sive plant microbiome research. Front Ecol Evol 2020;8:61.
https://doi.org/10.3389/fevo.2020.00061
Shehata AA, Schrödl W, Aldin Alaa A et al. The Effect of
glyphosate on potential pathogens and beneficial members of
poultry microbiota in vitro. Curr Microbiol 2013;66:350–8.
https://doi.org/10.1007/s00284-012-0277-2
Silva V, Mol HGJ, Zomer P et al. Pesticide residues
in European agricultural soils—a hidden real-
ity unfolded. Sci Total Environ 2019;653:1532–45.
https://doi.org/10.1016/j.scitotenv.2018.10.441
Sprankle P, Meggitt WF, Penner D. Rapid Inactivation
of glyphosate in the soil. Weed Sci 1975;23:224–8.
https://doi.org/10.1017/S0043174500052917
Straw EA, Carpentier EN, Brown MJ. Roundup causes high levels of
mortality following contact exposure in bumble bees. J Appl Ecol
2021;58:1167–76. https://doi.org/10.1111/1365-2664.13867
Suryanarayanan TS, Shaanker RU. Can fungal endophytes fast-track
plant adaptations to climate change? Fungal Ecol 2021;50:101039.
https://doi.org/10.1016/j.funeco.2021.101039
Székács A, Darvas B. Re-registration challenges of glyphosate
in the European Union. Front Environ Sci 2018;6:78.
https://doi.org/10.3389/fenvs.2018.00078
Tedjo D, Smolinska A, Savelkoul P et al. The fecal microbiota
as a biomarker for disease activity in Crohn’s disease. Sci Rep
2016;6:35216.
van Bruggen AHC, Finckh MR, He M et al. Indirect effects of the her-
bicide glyphosate on plant, animal and human health through its ef-
fects on microbial communities. Front Environ Sci 2021;9:763917.
https://doi.org/10.3389/fenvs.2021.763917
van Bruggen AHC, He MM, Shin K et al. Environmental and health
effects of the herbicide glyphosate. Sci Total Environ 2018;616-
617:255–68. https://doi.org/10.1016/j.scitotenv.2017.10.309
Wang K, Sipilä TP, Overmyer K. The isolation and characterization of
resident yeasts from the phylloplane ofArabidopsis thaliana. Sci Rep
2016;6:39403. https://doi.org/10.1038/srep39403
Wang Q, Garrity GM, Tiedje JM et al. Naïve Bayesian classi-
fier for rapid assignment of rRNA sequences into the new
bacterial taxonomy. Appl Environ Microbiol 2007;73:5261–7.
https://doi.org/10.1128/AEM.00062-07
Wang Y, Zhou D, Sun R. Effects of phosphate on the adsorption of
glyphosate on three different types of Chinese soils. J Environ Sci
2005;17:711–5.
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023
12 Mathew et al.
Williams GM, Kroes R, Munro IC. Safety evaluation and risk as-
sessment of the herbicide roundup and its active ingredient,
glyphosate, for humans. Regul Toxicol Pharm 2000;31:117–65.
https://doi.org/10.1006/rtph.1999.1371
Zaller JG, Heigl F, Ruess L et al. Glyphosate herbicide affects be-
lowground interactions between earthworms and symbiotic myc-
orrhizal fungi in a model ecosystem. Sci Rep 2015;4:5634. https:
//doi.org/10.1038/srep05634
Zheng D, Alm EW, Stahl DA et al. Characterization of univer-
sal small-subunit rRNA hybridization probes for quantitative
molecular microbial ecology studies. Appl Environ Microbiol
1996;62:4504–13. https://doi.org/10.1128/aem.62.12.4504-4513.
1996
Zuccaro A. Plant phosphate status drives host microbial prefer-
ences: a trade-off between fungi and bacteria. EMBO J 2020;39:
e104144.
D
ow
nloaded from
 https://academ
ic.oup.com
/jam
bio/article/134/2/lxad006/6987274 by Library of M
edical Faculty user on 24 M
arch 2023