Wheat varietal diversity affects arbuscular mycorrhizal symbiosis and soil enzymatic activities in the root zone

Abstract

High-input agriculture has been associated with a drastic reduction of within-field crop genetic diversity, while plant (mostly functional) diversity in natural ecosystems has been shown to promote ecosystem functioning. Increasing intra-specific diversity in agroecosystems is a promising strategy to stabilize crop productivity and promote the associated diversity of fauna and microbiota. We investigated the effect of the within-field diversity of bread wheat varieties on arbuscular mycorrhizal fungi (AMF) and two enzymatic activities involved in organic nitrogen and phosphorus mineralization. A field experiment was designed to test whether the number of mixed wheat varieties in a plot, or their functional diversity (previously assessed), influence the abundance and diversity of AMF and the activity of leucine aminopeptidases and phosphatases in the root zone. The AMF abundance was measured by quantitative polymerase chain reaction, community composition was analyzed by Illumina metabarcoding on two AMF specific markers, and potential microbial activities were quantified by biochemical assays. Wheat traits related to root morphology and susceptibility to fungal diseases previously quantified for each variety were also used. Number of varieties significantly increased AMF abundance in roots, whereas functional cluster number did not, with no impact of root morphology. Functional cluster number influenced AMF diversity, though weakly and not linearly, responding most to binary mixtures. Both wheat variety and functional group number increased the potential leucine amino-peptidase activities in the root zone, while no effect was observed for phosphatase activities. Our results highlight that increasing crop intra-specific diversity triggered changes in key processes involved in nutrient acquisition.

Full text

Page 1/20
Wheat varietal diversity affects arbuscular mycorrhizal
symbiosis and soil enzymatic activities in the root zone
Elisa Taschen (  elisa.taschen@inra.fr )
https://orcid.org/0000-0002-5277-2876
Esther Guillot
Damien Dezette
Josiane Abadie
Didier Arnal
Claude Plassard
Adrien Taudière
Jérôme Enjalbert
Xavier Le Roux
Philippe Hinsinger
Research Article
Keywords: Triticum aestivum L., intra-specic, metabarcoding, phosphatase activities, leucine-aminopeptidase activities
Posted Date: April 10th, 2023
DOI: https://doi.org/10.21203/rs.3.rs-2756901/v1
License:   This work is licensed under a Creative Commons Attribution 4.0 International License.  Read Full License

Page 2/20
Abstract
High-input agriculture has been associated with a drastic reduction of within-eld crop genetic diversity, while plant (mostly
functional) diversity in natural ecosystems has been shown to promote ecosystem functioning. Increasing intra-specic
diversity in agroecosystems is a promising strategy to stabilize crop productivity and promote the associated diversity of
fauna and microbiota. We investigated the effect of the within-eld diversity of bread wheat varieties on arbuscular
mycorrhizal fungi (AMF) and two enzymatic activities involved in organic nitrogen and phosphorus mineralization. A eld
experiment was designed to test whether the number of mixed wheat varieties in a plot, or their functional diversity
(previously assessed), inuence the abundance and diversity of AMF and the activity of leucine aminopeptidases and
phosphatases in the root zone. The AMF abundance was measured by quantitative polymerase chain reaction, community
composition was analyzed by Illumina metabarcoding on two AMF specic markers, and potential microbial activities were
quantied by biochemical assays. Wheat traits related to root morphology and susceptibility to fungal diseases previously
quantied for each variety were also used. Number of varieties signicantly increased AMF abundance in roots, whereas
functional cluster number did not, with no impact of root morphology. Functional cluster number inuenced AMF diversity,
though weakly and not linearly, responding most to binary mixtures. Both wheat variety and functional group number
increased the potential leucine amino-peptidase activities in the root zone, while no effect was observed for phosphatase
activities. Our results highlight that increasing crop intra-specic diversity triggered changes in key processes involved in
nutrient acquisition.
Introduction
Agroecosystems need to be managed more sustainably, in particular by reducing chemical inputs and their environmental
impacts. Increasing the diversity of agricultural systems at both eld and landscape scales is a key principle of ecological
intensication proposed to improve agroecosystem performance and resilience and minimize the need for external inputs
(Gaba et al., 2015). Studies in natural and manipulated ecosystems have shown that increasing plant species richness
enhances ecosystem functioning, mainly in terms of productivity and stability (Tilman et al., 1997; Weisser et al., 2017; Hong
et al., 2022) but also in terms of soil conditions and functions (El Moujahid et al., 2017; Le Roux et al 2013). This type of
positive relationship relies on two non-exclusive mechanisms, namely the selection/sampling effect and the
complementarity effect (Loreau 1998). Functional trait diversity among plant species was shown to be a good predictor of
the ecosystem functioning (Tilman et al., 1997). Although this diversity might be lower at the intra-specic level, intra-specic
diversity is not negligible (for wheat, see Cantarel et al. 2021) and variety mixture can lead to substantial overyielding (Litrico
& Violle 2015, Wuest et al., 2021). In their meta-analyses including 91 studies on cereals and legumes, Reiss & Drinkwater
(2018) found an overall yield increase of 2.2% for varietal mixtures relative to their single-variety components. A meta-
analysis focusing on bread wheat varietal mixtures showed a mean overyielding of 3.9%, reaching 6.2% in conditions of high
disease pressure (Borg et al., 2018). In the case of intraspecic mixtures of durum wheat, Montazeaud et al (2020) found
that root traits such as the root tissue density and root angle were among variables explaining the best the productivity (plant
biomass and yield) and grain quality of durum wheat grown in binary mixtures. This makes sense as belowground traits
related to nutrient acquisition strategies have been shown to be of major importance for the positive outcomes of crop
species or variety mixtures (Barot et al. 2017; Hinsinger et al. 2011; Dubs et al. 2023).
To cope with varying or limiting nutrient resources in soil, plants have evolved different nutrient acquisition strategies, and
related traits (e.g. Erel et al. 2017). These strategies involve (i) root morphology (specic root length, branching) to optimize
soil exploration (foraging strategy) and (ii) root physiology (release of protons, carboxylates and extracellular enzymes in
particular) to mobilize inorganic nutrients or mineralize organic resources (mining strategy). Tight interactions with
rhizospheric microorganisms (especially bacteria and fungi) are recognized for their contribution to both foraging and
mining mechanisms. Amongst these, external hyphae of arbuscular mycorrhizal fungi (AMF) considerably expand the
rhizosphere volume (extending up to several centimeters away from the root surface; Thonar et al., 2011) and thus make a
major contribution to the foraging strategy of mycorrhizal plants. The AMF may also enhance nutrient mining from less

Page 3/20
available pools through stimulating phosphorus (P) solubilizing bacteria (Wang et al. 2017) and bacterial communities
involved in organic P and nitrogen (N) mineralization (Wang et al., 2022, Nuccio et al. 2012) as related to specic enzymatic
activities (Ezawa & Saito 2018). The benet of the mycorrhizal symbiosis is however variable and highly dependent on the
environment (e.g., nutrient availability; Ingraa et al., 2020) but also on the involved plant and fungal partners. Despite the
lack of host specicity, AMF display host preferences, and AMF community varies between plant species among a same
genus (Pivato et al., 2007) or between genotypes of the same species (Ercoli et al 2017, Ellouze et al., 2018). If different
genotypes harbor different AMF communities, increasing genotypic diversity within a eld should enrich the total AMF
diversity. This vertical effect of plant diversity on AMF communities has been shown at the inter-specic level (van der
Heijden et al. 1998; Neuenkamp et al. 2018). However, the role of genetic and phenotypic diversity within the population of
host plant species on AMF communities and functioning has received little attention so far, while being a critical leverage for
a more sustainable agriculture.
It has been shown that wheat genotypes varied in their colonization rates, carbon investment into AMF and growth or nutrient
uptake response (Garcia de Leon et al., 2020; Elliott et al., 2021). Wheat has long been recognized as a crop with divergent
responses to the AMF symbiosis, from negative, neutral to positive effects. Plant dependency to AMF association was shown
to be linked to root morphology, with lower dependency for plant species with coarser roots, and higher specic root length
(ratio of root length to dry mass of roots; SRL) (Bergmann et al., 2020). At the plant intra-specic species level, this
hypothesis (Hetrick et al., 1991) has been poorly studied so far. A side effect of modern breeding might comes from the
selection of fungal disease resistant varieties. Because of similarities of the plant immune system responses to
infection/colonization processes for AMF and biotrophic pathogens, a trade-off might exist as discussed by Jacott et al
(2017). Thus, breeding crops for high resistance against fungal diseases, such as yellow rust and septoriose for wheat, might
have reduced AMF colonization potential. Moreover, there is an unclosed debate on the impact of breeding crops under high
fertilization inputs on the mycorrhizal responsiveness and dependency of cereals such as wheat and maize (Hetrick et al.,
1993, Lehmann et al., 2012; Ellouze et al., 2016; Zhang et al., 2019).
In this study, we used a eld trial set up by the Wheatamix consortium using 16 bread wheat varieties characterized for 27
below- and aboveground functional traits (Cantarel et al., 2021). These varieties were divided into four functional clusters
according to previously acquired traits’ database. The eld experiment was designed with 88 wheat plots harbouring a single
variety or mixtures of 2, 4 or 8 varieties differing in the number of functional groups (Dubs et al., 2018; Dubs et al., 2023). We
aimed to understand how the levels of genetic and functional diversity of bread wheat are affecting AMF communities and
two enzymatic activities involved in organic N and P mineralization in the root zone (leucine-amino-peptidase (LAP) and
phosphatases, respectively). We hypothesized that wheat diversity (variety number and functional diversity) could enhance
AMF diversity and stimulate the potential capacities of wheat mixtures for foraging (AMF colonization) and mining
(enzymatic activities) soil nutrients such as N and P.
Material And Methods
Field experiment
The experiment is related to the large Wheatamix project investigating the effect of intraspecic diversity of bread wheat on
ecosystem functioning and services (Dubs et al., 2018; Dubs et al., 2023). Our study was conducted on a eld experiment
carried out in 2016 at the INRAE Experimental Station in Versailles, France (48°48′26″N, 02°05′13″E, elevation 114 m). The full
experimental design is detailed in Dubs et al (2018). Soil texture was silty (62.96% silt, 19.36% sand, 17.67% clay) with 12.00
(standard deviation SD 1.05) mg. kg− 1 of total organic C, 0.93 (SD 0.05) mg. kg− 1 of total N, and 34.43 (SD 10.62) mg. kg− 1
of available (Olsen) P. In a previous step, 58 bread wheat varieties (
Triticum aestivum
L.) of diverse origins (elite varieties,
modern varieties bred for organic farming, MAGIC recombinant lines and few landraces) were phenotyped for 27 above- and
below-ground traits related to agronomic and ecological functions (Cantarel et al., 2021). This database was used for a
multi-trait classication of varieties into four functional groups (Dubs et al., 2018), hereafter referred to as clusters. In our

Page 4/20
eld experiment, a subset of 16 varieties was selected with four varieties of each of the four functional clusters (Table S1).
Wheat varieties (Altigo, Trémie, F426 and A22) of the cluster 1 (cl1) represent a functional group characterized by highest
susceptibility to fungal diseases, low SRL and weak ag leaf N content. The functional group cl2 was composed of wheat
varieties (Renan, Skerzzo, Midas, Alauda) less susceptible to fungal diseases (only one susceptible variety) but also with low
SRL. The cl3 functional group was composed of tall wheat varieties used in organic agriculture (landrace : “Blé Autrichien”,
varieties bred for organic agriculture: Hermès, Maxi, Ritter), with high SRL and mean susceptibility to fungal diseases (two
out of 4 being rather sensitive). Finally, the cl4 functional group contained elite varieties (Grapeli, Soissons, Arezzo, Boregar)
with high SRL and lowest susceptibility to fungal diseases (Table S1). Preferential uptake of ammonium or nitrate was also
a distinctive trait between wheat clusters, with high level of NO3– uptake capacity for cl2 and high levels of NH4+ uptake
capacity of cl3 (Cantarel et al., 2021).
Varieties were grown alone or as mixtures of two, four or eight varieties. To explore a gradient of functional diversity among
varietal diversity, mixtures were composed of wheat varieties belonging to the same or different clusters (Fig.1). In other
words, for a given varietal richness level, the number of functional groups varied from one to the highest possible number. In
total, 88 diversity modalities (16 mono varietal and 72 mixtures) were sown on a randomized design, using plots of 10.5 m x
8 m, sown at a density of 180 seeds m−². Since the objective was to quantify the effects of varietal richness and functional
diversity and not to assess signicant differences between pairs of mixtures, there was no replicate of each varietal mixture
but true replicates (i.e. different variety compositions) for each variety number x functional group number combination, as
done in current experimental designs exploring biodiversity-ecosystem functioning relationships (e.g. Weisser et al., 2017).
Wheat was sown in October (2015) after a preceding maize crop. The crop received 170 kg N ha− 1 in two doses, which is
matching with a wheat grain yield objective of 8.5 t ha− 1. Note that the climatic conditions in France in 2016 were
characterized by abnormally warm temperatures in late autumn followed by abnormally wet conditions in spring, leading to
extreme yield losses on wheat (Ben-Ari et al., 2018). Seeds were coated with a pesticide mix (CELEST, 2 cm3 kg− 1 –
Fludioxonil 25 g dm− 3 and SIGNAM 600 cm3 kg− 1 – Cypermethryne 300 g dm− 3). No additional fungicide or insecticide
treatment was applied afterwards. In March 2016, one herbicide treatment was applied at growth stage 31 (rst node
detectable; 50 g ha− 1 Harmony extra, 250 g ha− 1 Archipel, and 1 dm3 ha− 1 adjuvant Actirob 842 g dm− 3 esteried rapeseed
oil base).
Plant and soil sampling
Plants were sampled in May 2016 at the heading stage (variable according to varieties). On each plot, 50 plants were
sampled along two 3-m rows, to ensure that plot heterogeneity was accounted for and that sampling encompassed all the
wheat varieties present in the plot considered. Ten soil cores were sampled in the topsoil root zone at the same locations in
the plot as for plant sampling, and pooled into a single soil sample per plot. Roots were washed, weighted and three
segments of two centimeters were sampled from each plant. The vials containing either soil or root samples were directly
covered by liquid N2 and stored at -80°C. For yield data, the central section of each plot (1.75 × 8 m) was harvested in early
August 2016, with a cut-width of 1.75 m (MB Hege 140 combine harvester; Hege Maschinen GmbH). Grain yield are
expressed as mean grain weight measured at 15% humidity in t ha− 1.
Quantication of soil chemical properties and enzymatic activities
Soil organic C (SOC) and total N (Ntot) were determined by dry combustion (NF ISO 10694 and NF ISO 13878). Available P
content was extracted using the Olsen method and assayed colorimetrically (NF ISO 11263). Two enzymatic activities
(leucine aminopeptidases (LAP) and phosphatases) were measured after three hours incubation at soil pH 6.55 ± 0.07 (pH in
water extract) on 1 g of soil (after 12 hours thawing at 4°C) by uorometric method using methylumbelliferyl (MUB)-
substrates (details in Bell et al. 2013). Enzymatic activities in the soil are expressed in nanomoles of substrate mineralized
per g of soil per minute.
Measurements of AM fungal abundance and diversity

Page 5/20
Wheat roots were ground in a mortar with liquid N2, and a volume of 250 mm3 was used for DNA extraction. DNA quality and
concentrations were measured using Invitrogen™ Quant-iT™ PicoGreen™ dsDNA Assay Kit.
Abundance of Glomeromycotina in wheat roots was assessed by qPCR using FLR3-FLR4 primers (approx. 380 pb; Golotte et
al. 2004). Final nucleotide “T” was removed from the FLR3 original primer to enclose more AMF species and reduce positive
bias toward the Glomeraceae family (personal communication from D. van Tuinen). The qPCR reaction was carried out on 5
ng of root DNA in a nal volume of 10 mm3 comprising 5 mm3 Mix Sso advanced SYBR green Biorad, 0.5 mm3 of each
primer at 10 µM, 2 mm3 of DNA extract, and 2 mm3 of ultrapure water. The PCR cycle was as follows: 2 min at 98°C, (5 sec at
98°C, 30 sec at 60°C, 30 sec at 72°C) for 39 cycles, plus melting curve measurement. Each plate included duplicate reactions
per DNA sample and triplicate for standard set. If variation coecient exceeded 20% between duplicates, the result was
conrmed by a third measurement. Standard curves were obtained by serial dilution (10− 8–10− 3) of linearized plasmids
containing a cloned FLR3-FLR4 gene (certied as Glomeromycotina by Sanger sequencing). For conciseness, the number of
gene copies per ng of root DNA in wheat roots will hereafter be referred in the text as AMF abundance.
For metabarcoding, amplicons were constructed following a two-step PCR protocol as described in Battie-Laclau et al (2020).
Two Glomeromycota specic primer-pairs were used, FLR3/FLR4 (modied on the nal nucleotide “T” as previously;
approximatively 380 pb; Golotte et al. 2004) and NS31/AML2 (approximatively 480 pb; Simon et al., 1992, Lee et al., 2008),
targeting respectively the Large Sub-Unit (LSU) region and the Small Sub-Unit (SSU) region of the ribosomal DNA (rDNA). For
the rst round of PCR (PCR1), reactions were carried on two replicated dilutions of 15 ng.mm− 3 of DNA per sample. PCR
conditions are presented in Table S5. For each marker, the two PCR1 amplicons were pooled and puried by magnetic beads
(Clean PCR, Proteigene, France). The second PCR was performed using a Nextera® XT Index Kit (Illumina, San Diego, USA)
following the manufacturer’s instructions. After a purication with magnetic beads, these nal PCR products were
multiplexed and sequenced on a MiSeq Illumina sequencer using MiSeq Reagent Kit v3 (600-cycle; Illumina).
Bioinformatic analyses
We analyzed DNA sequence through the bioinformatics pipeline described in
https://adrientaudiere.gitlab.io/solfami_bioinfo_ssu/and/solfami_bioinfo_lsu/. In short, sequences were quality ltered using
the lterAndTrim function from the dada2 package (Callahan et al. 2016a), by rst truncating reads of a quality score inferior
to 10 and second, discarding sequences with less than 50 bp. Then we followed dada2 classic pipeline (Callahan et al.
2016b) to obtain chimera-free amplicon sequence variants (called ASV) using paired-end. Each ASV longer than 200 pb was
then taxonomically assigned to MaarjAM database (Öpik et al 2010) with the assignTaxonomy function from dada2
(Callahan et al. 2016a), which implements the RDP classier of Wang et al. (2007).
AMF diversity
Alpha-diversity was evaluated by Hill numbers (using the ‘vegan’ R package, Oksanen et al 2022) without previous
normalization following McMurdie & Holes (2014). Hill diversity indices (Hill 0 = Richness; Hill 1 = Hill-Shannon; Hill 2 = Hill-
Simpson) consider both the number and the relative abundance of species, with decreasing sensitivity to rare species and to
sample size (Roswell et al., 2021). Differences in terms of AMF community composition (beta-diversity) between functional
clusters was analyzed by PERMANOVA on Bray-Curtis distances using the ‘adonis
’
function from the ‘vegan’ R package
(Oksanen et al 2022), after normalization by standardizing abundances to the median sequencing depth.
Statistical analyses
All statistical analyses were realized using R (version 4.0.4). Linear mixed models were performed with the package ‘nlme’ to
determine whether the number of wheat varieties and the number of functional wheat clusters inuenced variables related to
wheat plant characteristics involved in nutrient acquisition, focusing on AMF abundance in roots as well as phosphatases
and LAP enzymatic activities in the root zone. The same analyses were done on the alpha-diversity of AMF in roots.

Page 6/20
The full linear mixed models also included soil proprieties as co-variables that could have an impact on the variables of
interest (C, N, Olsen P) and the spatial y-coordinates was set as random effect as rst exploration with linear models detected
some spatial heterogeneity along the y-axis of the eld experimental site (Fig.1). Co-variables of the models were
standardized (centered on the matrix and scaled on each variable’s standard deviation;
scale
function in R) to discard
spurious effects of various units of covariates included models. The nal model was selected step by step by keeping
variables that signicantly contributed to the model using the Akaike information criterion. Main factors of interest (number
of wheat genotypes and of functional clusters) were always kept in the nal model. All models were tested for normality and
AMF abundance and Hill indexes had to be log-transformed. Analyses of Deviance (Anova Type II) were performed on the
nal models, with the
Anova
function of the ‘car’ R package. The full model is detailed in Table S2. Direct correlations
between enzymatic activities and AMF abundance were analyzed by Pearson coecient and plotted with
ggscatter
(package
‘ggpub
r’
).
Comparisons of the measured variables between the four functional wheat clusters were realized on a data subset including
only functionally homogeneous mixtures (plots composed of varieties belonging to the same cluster). This dataset included
seven plots for each cluster type (four with one variety, two with two varieties, and one with four varieties).
Results
Wheat shoot mass, nitrogen content and grain yield
Neither the number of wheat genotypes nor the number of functional clusters in mixtures did explain shoot biomass and
total N uptake (p-values >0.1; Table S2) at the heading stage or the nal yield (p-value >0.1; Table S2). However, yield was
signicantly different between the four functional clusters (wheat grown as functionally homogeneous mixtures), with a
signicantly higher grain yield for Cluster 4 composed of disease-resistant elite varieties (Fig. 2. A.). Such difference between
the functional clusters in mixtures was not visible on shoot biomass at the heading stage in May(Fig 2. B.), and no
signicant difference between total N content in shoot biomass was found (Fig. 2. C).
Effect of wheat diversity on enzymatic activities in the root zone and AMF abundance in roots
There were no signicant differences in enzymatic activities in the root zone between the four functional clusters (Fig. 2 D.
and Fig. 2 E.). The number of wheat varieties and of functional clusters did not affect phosphatase activities either (Table 1).
However, both varietal and functional diversity had a signicant and positive effect on the LAP activities in the root zone
(Table 1), with signicantly higher LAP activities measured on soil from the root zone of mixtures with four or eight varieties
compared to lower numbers of varieties (two or one varieties) (Fig. 3 A.). Nevertheless, there was no signicant effect of
functional group number on LAP activities, based on pairwise comparisons (Fig. 3 B.). Total N in shoot biomass was
signicantly impacted by LAP activities (p-val= 0.0407) with a slight but positive relation (slope= 0.36; Table S1 B).
There were no signicant differences in AMF abundance in roots between the four functional clusters (Fig. 2 F.). The number
of wheat varieties however signicantly impacted AMF abundance in roots (Table 1), with higher abundance (2.3-fold
increase on average) in plots mixing eight wheat varieties compared to pure cultures (Fig. 3 E.). There was no effect of
functional cluster number on AMF abundance in roots (Fig. 3 F.).
As co-variable in linear mixed models, total N in soil had a positive correlation with LAP activities slope=0.019; Table 1) but it
was negatively correlated with phosphatases activities (slope=-0.22; Table 1)). The AMF abundance in roots was also
impacted by total N in soil, with a positive correlation (slope=0.179; Table 1), but negatively affected by available (Olsen) P in
soil (slope=-0.208; Table 1). Phosphatase activities were negatively and linearly related with AMF abundance in roots
(Pearson, R=-0,37, p-val = 0.002). On the contrary, LAP activities exhibited a positive linear relationship with the abundance of
AMF in roots (Fig. 4; Pearson,R=0.49, p-val < 0.001).
AMF community composition

Page 7/20
For the LSU and SSU regions respectively, a total of 240 and 409 Glomeromycota ASVs (Amplicon Sequence Variants) were
obtained from 8 753 361 and 18 228 610 sequences. For the LSU, mean sequencing depth per sample was 99 661, grouped
into 277 ASVs after discarding ASVs with less than ve reads. For the SSU, mean sequencing depth per sample was 209 016
after discarding one outlining sample with less than 50 000 sequences, and 392 ASVs (> 5 sequences).
In the same order of abundance, the number of ASVs from the LSU and SSU regions belonged to the orders of Glomerales
(LSU: 153; SSU: 206), Diversisporales (LSU: 3; SSU: 44), Archeosporales (LSU: 1; SSU: 41), Paraglomerales (LSU: none; SSU:
3), and non-identied (LSU: 83; SSU: 115). The ten most abundant ASVs represented 92.32 % of all LSU sequences, among
which
Funneliformis
was the dominant genus (8 out of 10 ASVs; Table 2). For the SSU region, the ten most abundant ASVs
represented 96.45 % of all sequences, with ASVs from the genera of
Funneliformis
,
Scutelospora
,
Archeospora
,
Diversispora
and two uncultured Glomeromycota. For both LSU and SSU regions, the most abundant ASV blasted on
Funneliformis
caledonium
(Table 2).
Effect of wheat diversity on AMF diversity in roots
There was no difference in AMF community composition between wheat functional clusters, with no signicant variations in
beta-diversity (Permanova,
p-value
=0.165 and 0.098, on respectively the LSU and SSU data; Table S3). However, on the SSU,
wheat functional clusters signicantly differed in alpha-diversity (Hill 1 index; Fig 2 G.) with signicantly higher AMF diversity
in wheats of cluster 4 than for that of cluster 1. No such signicant difference between wheat functional clusters was found
for the LSU data (Hill 1 index; Fig 2 H.). The number of wheat varieties did not inuence AMF alpha-diversity neither on the
LSU nor on the SSU sequencing data (Table 1). However, the number of wheat functional clusters had a signicant impact on
Hill1 index for the LSU and SSU sequencing data (Table 1, Fig. 3 H., J.) but Tukey’s post-hoc test revealed signicant
differences between mixtures treatments only for the SSU marker, with higher diversity for mixtures composed of two than
three wheat functional clusters (Fig. 3 J.). The Hill 0 index (specic richness) was never signicantly impacted by wheat
diversity, while Hill 2 followed the trend of Hill 1(Table 1).
Discussion
The number of varieties, but not the type or number of variety functional groups, increases AMF abundance
Mixtures with increasing numbers of wheat varieties signicantly increased AMF abundance in roots (Table1, Fig.3E).
Different non-exclusive mechanisms could explain such an effect. First, this could be due to a sampling effect, where
increasing the number of varieties in mixtures also increases the chance to include particularly mycotrophic varieties. An
ex
situ
experiment showed indeed a high variability in AMF colonization rates between genotypes in durum wheat (
Triticum
turgidum
L.), ranging from 7–84% (Ganugi et al., 2021). Such an increase in AMF abundance could also be due to a
complementarity effect where host diversity enabled niche differentiation for AMF colonization ; or simpler said: each AMF
strain found its “favorite” wheat variety partner. Also, if increasing the number of wheat varieties enabled higher plant
photosynthesis (through potential above- and belowground resource sharing, higher disease resistance, and competitive
replacement), carbon availability to support AMF symbiosis could have been enhanced, thus increasing AMF abundance. In
the experiment, there was no evidence however of improved biomass or grain yield when increasing the number of wheat
varieties. This does not necessarily relates with the amount of carbon allocated to AMF, though, which would have been
extremely dicult to measure in such a eld experiment. In addition, as the ten most abundant ASV represented > 90% of all
sequences, with no differences between functional wheat clusters (Permanova), these common AMF strains were certainly
present on all wheat, thus creating a dense common mycorrhizal network. Investment into common mycorrhizal networks
has been shown to depend on the identity of mixed plant species or genotypes. Engelmoer et al. (2015) found that mixing
different host species (
Daucus carota
L.,
Cichorium intybus
L. and
Medicago truncatula
Gaertn.) reduced the investment of
plants into extraradical hyphae of their common network. At the intra-specic level of host plant diversity, File et al (2012)
showed higher hyphal length in an AMF network between sibling plants compared to populations of more distantly related
plants of the same species (
Ambrosia artemisiifolia
L.). As extra-radical hyphal length has been shown to be the best

Page 8/20
indicator of AMF contribution to plant nutrition in the study of Sawers et al. (2017), it would have been interesting to evaluate
the effect of intra-specic diversity on plant investment into the common mycorrhizal network, to verify whether it followed
the observed increase of AMF abundance in roots with increasing genetic diversity in wheat variety mixtures. The role of AMF
on wheat development was not directly investigated here, but a pot experiment on maize recently showed substantial
overyielding of varietal mixtures only occurred when plants were grown in association with the AM fungus
Funneliformis
mosseae
(Wang et al., 2020).
In contrast with the effect of wheat variety diversity, there was no effect of the functional diversity of wheat variety mixtures
on AMF abundance (Table1, Fig.3F.). Since the clusters of wheat varieties differed for criteria such as root morphology and
resistance against pathogens (Table S1), this offered an interesting opportunity to challenge two common statements and
debated hypotheses, concerning trade-offs between AMF root colonization and either root morphology or disease resistance.
Actually, the four wheat clusters did not differ in root AMF abundance (Fig.2F.). We noted no difference between clusters
with high SRL (Clusters 3 and 4) and those with low SRL (Clusters 1 and 2) and thus found no visible trade-off between soil
foraging strategies by AMF root colonization and high SRL (Fig.2F.; Table S2), as previously found by Hetrick (1991). Ruiz-
Lozano et al. (1999) showed that genes involved in resistance to powdery mildew in barley (
Hordeum vulgare
L.) reduced
root colonization by the AM fungus
Funneliformis mosseae
(called by then
Glomus mosseae
). However, resistance to fungal-
borne disease of wheat did not reduce AMF abundance in roots in our experiment, when comparing wheat plots of Cluster 4,
composed of only elite resistant varieties, to those of the highly susceptible varieties of Cluster 1 (Fig.2F.). When looking at
the direct effect of mean susceptibility to yellow rust on the abundance of AMF in roots (regardless of functional cluster, on
the total trait matrix), a signicant effect was observed (p-value < 0.001; Table S2) but with a negative correlation. This result
was conrmed when looking at the effect of the observed pressure of yellow rust on wheat plots during the experiment and
the AMF abundance (p-value < 0.05; Table S2). Disease pressure was very high in France in 2016 due to abundant rainfall in
Spring, as occurred at our experimental site, and the effect of wheat variety mixtures in the present eld experiment have
been further discussed by Vidal et al. (2020). This correlation showed that disease symptoms altering plant development
also had a negative impact on AMF colonization, possibly by limiting C allocation or AMF colonization.
Activities of leucine aminopeptidases but not phosphatases positively correlated to wheat variety diversity and AMF
abundance
Neither the number of wheat varieties nor the number of wheat functional clusters did impact phosphatase activities in the
root zone (Table1; Fig.3C. and 3 D.). In contrast, these two components of wheat diversity, i.e. genetic and functional
diversity, both had a positive effect on LAP activities in the root zone (Table1; Fig.3A.). Leucine aminopeptidases (LAPs) are
metallopeptidases that cleave N-terminal residues from proteins and peptides and are expressed by soil bacteria
(Loeppmann et al., 2016). Proteins are an important source of N, which can represent 40% of the total soil N (Schulten &
Schnitzer et al., 1997). Increasing genetic and functional diversity of wheat might have triggered the demand for uptake of N
by plants, stimulating mechanisms to access organic N pools. LAP activities in the root zone were positively correlated to
total N content in wheat shoot biomass (Table S1), as well as to AMF abundance in roots (Fig.4). The relation between LAP
activities and AMF abundances might not be direct: both variables can be affected by a common external factor or variable,
such as wheat diversity. However, enhanced LAP activities concomitant to increased AMF abundance in the root can also be
the consequence of increasing root–microorganism competition for soil inorganic N (Kuzyakov & Xu 2013). Liu et al. (2021)
showed that N uptake from organic patches via AM fungal hyphae was directly affected by soil LAP activities, possibly due
to a stimulation of the microbial activities in the soil. Previous studies demonstrated that AMF can stimulate soil microbial
activities related to N cycling by inuencing the structure of bacterial communities (Nuccio et al., 2012; Jansa et al., 2019 for
a review). Although much less probable, it should be noted that little is known about LAP activities of AMF
per se
. In the
genome of
Rhizophagus irregularis
, three genes have been identied, each encoding a protein with a putative LAP function
(JGI mycocosm.jgi.doe.gov; Chen et al 2018; Table S4). One of the genes (coding for the 1524872 protein) is highly
expressed in the extraradical mycelium and the mycelium in planta (personal com. C. Roux; unpublished data). Additionally,
a peptide signal, enabling the release of the protein outside the cell, is predicted only for this protein (not present on the two

Page 9/20
other genes), but with low prediction robustness. Hence, this protein could be an interesting candidate to explain patterns of
LAP activities variation with the amount of AMF copies measured in the roots. However, its occurrence in the soil remains to
be veried. Further studies should also test to which extent protocols used for soil enzymatic activities may extract cytosolic
enzyme by breaking hyphae, as they often include steps of soil freezing and thawing, sieving and blending (as was the case
in the protocol used here).
The type or number of functional groups, but not the number of varieties, weakly alters AMF diversity
To our knowledge, this is the rst study focusing on the effect of plant intra-specic diversity and its vertical effect on AMF
diversity. We expected an increased AMF diversity with increasing variety number. However, wheat variety diversity had no
signicant effect on AMF diversity (Table1, Fig.3G. and 3 I.), while there was a signicant but weak effect of the functional
diversity of wheat variety mixtures on some of the indicators of AMF diversity, such as SSU Hill 1, SSU Hill 2 and LSU Hill
(Table1, Fig.3J.). In addition, the functional clusters of wheat varieties signicantly differed in AMF diversity (Fig.2G.).
Indeed, based on the LSU data, wheat plots of Cluster 4 (composed of resistant elite varieties), displayed signicantly higher
AMF diversity than those of Cluster 1 (composed of less resistant elite varieties and varieties from a highly recombinant
inbred MAGIC panel). Again, AMF diversity was possibly linked to host resistance, due to (i) either a functional link between
disease resistance mechanisms and AMF selectivity, or (ii) the amount of carbon allocated by the host plants, with stronger
competition between AMF when the resource is scarce (i.e. in wheat affected by fungal pathogens). The rather weak effect of
wheat functional diversity on AMF diversity observed in the present experiment can be explained by similar AMF
communities composition between functional clusters (Table S3), and between varieties (not tested here). As we expected
that intra-specic diversity of host plants might induced subtle variation in AMF community composition (compared to
communities of different plant species), we used two different AMF markers. The SSU region was used for its better
coverage of the different AMF families, completed by the LSU marker, for its better taxonomic resolution than the slowly
evolving SSU region (Krüger et al., 2012; Hart et al., 2015; Delavaux et al., 2021). On both markers, the genus
Funneliformis
was strongly dominating. On the SSU, a single
Funneliformis
ASV representing 93.3% of all sequences, subdivided into eight
dominating ASVs representing 90.7% of all sequences on the LSU.
Funneliformis
was also the dominating genus in two eld
studies in Canada on large sets of durum wheat genotypes including landraces and commercial varieties (Ellouze et al.,
2018; Stefani et al., 2020). Both studies sequenced different regions of the SSU and revealed only weak differences in alpha-
diversity and composition of AMF communities between wheat genotypes in roots, rhizosphere and bulk soil. Stefani et al.
(2020) found no difference at all, while Ellouze et al. (2018) did not nd any difference when looking at roots, but observed
that durum wheat genotype differently shaped the composition of AMF communities in their rhizosphere. Jacquiod et al.
(2021) found signicant differences in the diversity and the composition of rhizosphere microbial communities of wheat
elites and landraces, with root-associated fungi being particularly dependent of the interaction between the plant genotype
and the environment. Under low fertilizer inputs, OTU richness was higher in ancient wheat varieties than in modern varieties,
but it was the opposite upon addition of inorganic fertilizers. This phenotypic plasticity with varying environmental
conditions needs to be hold in mind: indeed, the observed responses of the wheat variety functional clusters discussed in our
study were obtained under a specic environment, which corresponded to a rather high N and P fertilizer input history. Would
an increase of AMF diversity induce a better functioning of agroecosystems is still an open question, which the present work
hardly addressed. This would require to explore a broader diversity of environmental conditions, including a range of low
fertilizer input agroecosystems. AMF fungi are recognized to have some degree of functional diversity (Van Der Heijden et al.,
2018), with a positive effect on some key functions in ecosystems. AMF diversity effect on plant species complementarity as
has been investigated in experimental settings (microcosm or macrocosm experiments; Wagg et al., 2015), but remain
challenging to demonstrate and thus hardly tangible in eld conditions.
Conclusion & Perspectives
Although wheat variety diversity did not affect aboveground biomass and yield, the within-eld number of wheat varieties
grown in mixtures stimulated AMF abundance in roots and enhanced leucine amino-peptidase (but not phosphatase)

Page 10/20
activities in the root zone. This possibly impacted the plant nutrient acquisition capacity through improved access to N from
organic sources. The effect of functional cluster diversity on these processes was less prominent, and underlying
mechanisms (complementarity, sampling effect...) need to be further understood. Concerning AMF diversity, mixing varieties
of two different wheat functional clusters enhanced AMF diversity, but the effect was subtle and did not increase with
increasing wheat variety diversity. Our results showed the presence of a limited number of dominant AMF strains, suggesting
a dense common mycorrhizal network between plants in wheat variety mixtures. The ecological (and economic) costs of the
excess use of inorganic N and P fertilizers call for an urgent reduction of those inputs. Our results showed that crop intra-
specic diversity is an interesting lever to induce changes in processes involved in nutrient acquisition strategies, although
this should be explored under a diversity of environmental conditions. In this respect, our work is calling for additional
studies under more limiting conditions (in terms of nutrient availability), where positive interactions could be even more
important, according to the stress gradient hypothesis.
Declarations
Acknowledgments
We thank Florence Dubs for her help on collecting the dataset, and Christophe Roux for comments and discussions. Data
used in this work were partly produced through the GenSeq technical facilities of the « Institut des Sciences de l’Evolution de
Montpellier », thanks to the support of the program ‘Investments for the future’ (ANR-10- LABX-04-01) granted to the LabEx
CeMEB (Montpellier).
Funding
: This work was supported by funding of the ANR WHEATAMIX project (grant ANR-13-AGRO-0008, French National
Research Agency) and the SolFaMi project (grant of the INRAE Metaprogramme EcoServ 2019).

Competing interested:
The authors have no relevant nancial or non-nancial interests to disclose.
Authors contribution:
Jérôme Enjalbert and Xavier Le Roux contributed to the eld study conception and design. Field
sampling was performed by Philippe Hinsinger, Didier Arnal, Damien Dezette, Jérôme Enjalbert and Xavier Le Roux.
Laboratory analyses were made by Josiane Abadie, Damien Dezette and Elisa Taschen, and bioinformatics by Adrien
Taudière. Data collection and analysis were performed by Elisa Taschen, who also wrote the rst draft of the manuscript,
with helpful discussion with Claude Plassard and Esther Guillot, and all authors commented on previous versions of the
manuscript. All authors read and approved the nal manuscript. Wheatamix consortium regroups many collaborators
participating in the eld experiment and the elaboration of the trait data set of wheat varieties.”
Data Availability
The datasets generated during and/ analyzed during the current study are available in the Data INRAE repository,
[PERSISTENT LINK TO DATASETS : Submission in progress].
References
1. Barot S, Allard V, Cantarel A, et al (2017) Designing mixtures of varieties for multifunctional agriculture with the help of
ecology. A review. Agronomy for sustainable development 37:1–20
2. Battie-Laclau P, Taschen E, Plassard C, et al (2020) Role of trees and herbaceous vegetation beneath trees in maintaining
arbuscular mycorrhizal communities in temperate alley cropping systems. Plant and Soil 453:153–171
3. Bell CW, Fricks BE, Rocca JD, et al (2013) High-throughput uorometric measurement of potential soil extracellular
enzyme activities. JoVE (Journal of Visualized Experiments) e50961
4. Ben-Ari T, Boé J, Ciais P, et al (2018) Causes and implications of the unforeseen 2016 extreme yield loss in the
breadbasket of France. Nat Commun 9:1627

Page 11/20
5. Bergmann J, Weigelt A, Van der Plas F, et al (2020) The fungal collaboration gradient dominates the root economics
space in plants. Science Advances 6:eaba3756
. Borg J, Kiær LP, Lecarpentier C, et al (2018) Unfolding the potential of wheat cultivar mixtures: A meta-analysis
perspective and identication of knowledge gaps. Field Crops Research 221:298–313
7. Callahan BJ, McMurdie PJ, Rosen MJ, et al (2016a) DADA2: High-resolution sample inference from Illumina amplicon
data. Nature methods 13:581–583
. Callahan BJ, Sankaran K, Fukuyama JA, et al (2016b) Bioconductor workow for microbiome data analysis: from raw
reads to community analyses. F1000Research 5:
9. Cantarel AA, Allard V, Andrieu B, et al (2021) Plant functional trait variability and trait syndromes among wheat varieties:
the footprint of articial selection. Journal of Experimental Botany 72:1166–1180
10. Chen EC, Morin E, Beaudet D, et al (2018) High intraspecic genome diversity in the model arbuscular mycorrhizal
symbiont Rhizophagus irregularis. New Phytologist 220:1161–1171
11. Delavaux CS, Sturmer SL, Wagner MR, et al (2021) Utility of large subunit for environmental sequencing of arbuscular
mycorrhizal fungi: a new reference database and pipeline. New Phytologist 229:3048–3052.
https://doi.org/10.1111/nph.17080
12. Dubs F, Le Roux X, Allard V, et al (2018) An experimental design to test the effect of wheat variety mixtures on
biodiversity and ecosystem services. hal-01843564
13. Dubs F, Enjalbert J, Barot S, et al. (2023) Unfolding the link between multiple ecosystem services and bundles of
functional traits to design multifunctional crop variety mixtures. (under revision)
14. Elliott AJ, Daniell TJ, Cameron DD, Field KJ (2021) A commercial arbuscular mycorrhizal inoculum increases root
colonization across wheat cultivars but does not increase assimilation of mycorrhiza-acquired nutrients. Plants, people,
planet 3:588–599
15. Ellouze W, Hamel C, DePauw RM, et al (2016) Potential to breed for mycorrhizal association in durum wheat. Canadian
journal of microbiology 62:263–271
1. Ellouze W, Hamel C, Singh AK, et al (2018) Abundance of the arbuscular mycorrhizal fungal taxa associated with the
roots and rhizosphere soil of different durum wheat cultivars in the Canadian prairies. Canadian journal of microbiology
64:527–536
17. El Moujahid L, Le Roux X, Michalet S, et al (2017) Effect of plant diversity on the diversity of soil organic compounds.
PLOS One 12: e0170494
1. Engelmoer DJ, Kiers ET (2015) Host diversity affects the abundance of the extraradical arbuscular mycorrhizal network.
New Phytologist 205:1485–1491
19. Ercoli L, Schüßler A, Arduini I, Pellegrino E (2017) Strong increase of durum wheat iron and zinc content by eld-
inoculation with arbuscular mycorrhizal fungi at different soil nitrogen availabilities. Plant and soil 419:153–167
20. Erel R, Bérard A, Capowiez L, et al (2017) Soil type determines how root and rhizosphere traits relate to phosphorus
acquisition in eld-grown maize genotypes. Plant Soil 412:115–132
21. Ezawa T, Saito K (2018) How do arbuscular mycorrhizal fungi handle phosphate? New insight into ne-tuning of
phosphate metabolism. New phytologist 220:1116–1121
22. File AL, Klironomos J, Maherali H, Dudley SA (2012) Plant kin recognition enhances abundance of symbiotic microbial
partner https://doi.org/10.1371/journal.pone.0045648
23. Gaba S, Lescourret F, Boudsocq S, et al (2015) Multiple cropping systems as drivers for providing multiple ecosystem
services: from concepts to design. Agron Sustain Dev 35:607–623
24. Ganugi P, Masoni A, Sbrana C, et al (2021) Genetic variability assessment of 127 Triticum turgidum L. accessions for
mycorrhizal susceptibility-related traits detection. Scientic reports 11:1–11

Page 12/20
25. Garcia de Leon D, Vahter T, Zobel M, et al (2020) Different wheat cultivars exhibit variable responses to inoculation with
arbuscular mycorrhizal fungi from organic and conventional farms. PLoS One 15:e0233878
2. Gollotte A, Van Tuinen D, Atkinson D (2004) Diversity of arbuscular mycorrhizal fungi colonising roots of the grass
species Agrostis capillaris and Lolium perenne in a eld experiment. Mycorrhiza 14:111–117
27. Hart MM, Aleklett K, Chagnon P-L, et al (2015) Navigating the labyrinth: a guide to sequence-based, community ecology
of arbuscular mycorrhizal fungi. New phytologist 207:235–247
2. Hetrick BAD (1991) Mycorrhizas and root architecture. Experientia 47:355–362
29. Hetrick BAD, Wilson GWT, Cox TS (1993) Mycorrhizal dependence of modern wheat cultivars and ancestors: a synthesis.
Canadian journal of Botany 71:512–518
30. Hinsinger P, Betencourt E, Bernard L, et al (2011) P for two, sharing a scarce resource: soil phosphorus acquisition in the
rhizosphere of intercropped species. Plant physiology 156:1078–1086
31. Hong P, Schmid B, De Laender F, et al (2022) Biodiversity promotes ecosystem functioning despite environmental
change. Ecology letters 25:555–569
32. Ingraa R, Amato G, Sosa-Hernández MA, et al (2020) Nitrogen Type and Availability Drive Mycorrhizal Effects on
Wheat Performance, Nitrogen Uptake and Recovery, and Production Sustainability. Front Plant Sci 11:760.
https://doi.org/10.3389/fpls.2020.00760
33. Jacott CN, Murray JD, Ridout CJ (2017) Trade-offs in arbuscular mycorrhizal symbiosis: disease resistance, growth
responses and perspectives for crop breeding. Agronomy 7:75
34. Jacquiod S, Raynaud T, Pimet E, et al (2021) Changes in wheat rhizosphere microbiota in response to chemical inputs,
plant genotype and phenotypic plasticity. bioRxiv
35. Jansa J, Forczek ST, Rozmoš M, et al (2019) Arbuscular mycorrhiza and soil organic nitrogen: network of players and
interactions. Chemical and Biological Technologies in Agriculture 6:1–10
3. Krüger M, Krüger C, Walker C, et al (2012) Phylogenetic reference data for systematics and phylotaxonomy of arbuscular
mycorrhizal fungi from phylum to species level. New phytologist 193:970–984
37. Kuzyakov Y, Xu X (2013) Competition between roots and microorganisms for nitrogen: mechanisms and ecological
relevance. New Phytologist 198:656–669. https://doi.org/10.1111/nph.12235
3. Le Roux X, Schmid B, Poly F, et al (2013) Soil environmental conditions and buildup of microbial communities mediate
the effect of grassland plant diversity on nitrifying and denitrifying enzyme activities. PLOS One
10.1371/journal.pone.0061069
39. Lee J, Lee S, Young JPW (2008) Improved PCR primers for the detection and identication of arbuscular mycorrhizal
fungi. FEMS microbiology ecology 65:339–349
40. Lehmann A, Barto EK, Powell JR, Rillig MC (2012) Mycorrhizal responsiveness trends in annual crop plants and their wild
relatives—a meta-analysis on studies from 1981 to 2010. Plant and Soil 355:231–250
41. Litrico I, Violle C (2015) Diversity in Plant Breeding: A New Conceptual Framework. Trends in Plant Science 20:604–613.
https://doi.org/10.1016/j.tplants.2015.07.007
42. Liu H, Wu Y, Xu H, et al (2021) Mechanistic understanding of interspecic interaction between a C4 grass and a C3
legume via arbuscular mycorrhizal fungi, as inuenced by soil phosphorus availability using a 13C and 15N dual-
labelled organic patch. The Plant Journal 108:183–196
43. Loeppmann S, Semenov M, Blagodatskaya E, Kuzyakov Y (2016) Substrate quality affects microbial- and enzyme
activities in rooted soil. Journal of Plant Nutrition and Soil Science 179:39–47. https://doi.org/10.1002/jpln.201400518
44. Loreau M (1998) Separating sampling and other effects in biodiversity experiments. Oikos 82:600–602
45. McMurdie PJ, Holmes S (2014) Waste Not, Want Not: Why Rarefying Microbiome Data Is Inadmissible. PLOS
Computational Biology 10:e1003531

Page 13/20
4. Montazeaud G, Violle C, Roumet P, et al (2020) Multifaceted functional diversity for multifaceted crop yield: towards
ecological assembly rules for varietal mixtures. Journal of Applied Ecology 57:2285–2295
47. Neuenkamp L, Moora M, Öpik M, et al (2018) The role of plant mycorrhizal type and status in modulating the
relationship between plant and arbuscular mycorrhizal fungal communities. New Phytologist 220:1236–1247
4. Nuccio EE, Hodge A, Pett-Ridge J, et al (2012) An arbuscular mycorrhizal fungus modies the soil microbial community
and nitrogen cycling during litter decomposition. LLNL-JRNL-554312
49. Oksanen J, Simpson G, Blanchet F, Kindt R, Legendre P, Minchin P, O'Hara R, Solymos P, Stevens M, Szoecs E, Wagner H,
Barbour M, Bedward M, Bolker B, Borcard D, Carvalho G, Chirico M, De Caceres M, Durand S, Evangelista H, FitzJohn R,
Friendly M, Furneaux B, Hannigan G, Hill M, Lahti L, McGlinn D, Ouellette M, Ribeiro Cunha E, Smith T, Stier A, Ter Braak C,
Weedon J (2022). _vegan: Community Ecology Package_. R package version 2.6-4, <https://CRAN.R-
project.org/package=vegan>
50. Öpik M, Vanatoa A, Vanatoa E, et al (2010) The online database MaarjAM reveals global and ecosystemic distribution
patterns in arbuscular mycorrhizal fungi (Glomeromycota). New Phytologist 188:223–241.
https://doi.org/10.1111/j.1469-8137.2010.03334.x
51. Reiss ER, Drinkwater LE (2018) Cultivar mixtures: a meta-analysis of the effect of intraspecic diversity on crop yield.
Ecological Applications 28:62–77
52. Roswell M, Dushoff J, Winfree R (2021) A conceptual guide to measuring species diversity. Oikos 130:321–338
53. Schulten H-R, Schnitzer M (1997) The chemistry of soil organic nitrogen: a review. Biology and fertility of soils 26:1–15
54. Simon L, Lalonde M, Bruns T (1992) Specic amplication of 18S fungal ribosomal genes from vesicular-arbuscular
endomycorrhizal fungi colonizing roots. Applied and environmental microbiology 58:291–295
55. Stefani F, Dupont S, Laterrière M, et al (2020) Similar arbuscular mycorrhizal fungal communities in 31 durum wheat
cultivars (Triticum turgidum L. var. durum) under eld conditions in Eastern Canada. Frontiers in plant science 11:1206
5. Thirkell TJ, Cameron DD, Hodge A (2016) Resolving the ‘nitrogen paradox’of arbuscular mycorrhizas: fertilization with
organic matter brings considerable benets for plant nutrition and growth. Plant, Cell & Environment 39:1683–1690
57. Thonar C, Schnepf A, Frossard E, et al (2011) Traits related to differences in function among three arbuscular
mycorrhizal fungi. Plant Soil 339:231–245. https://doi.org/10.1007/s11104-010-0571-3
5. Tilman D, Knops J, Wedin D, et al (1997) The inuence of functional diversity and composition on ecosystem processes.
Science 277:1300–1302
59. Van Der Heijden MG, Klironomos JN, Ursic M, et al (1998) Mycorrhizal fungal diversity determines plant biodiversity,
ecosystem variability and productivity. Nature 396:69–72
0. Vidal T, Saint-Jean S, Lusley P, et al (2020) Cultivar mixture effects on disease and yield remain despite diversity in
wheat height and earliness. Plant Pathology 69:1148–1160
1. Wagg C, Barendregt C, Jansa J, van der Heijden MGA (2015) Complementarity in both plant and mycorrhizal fungal
communities are not necessarily increased by diversity in the other. Journal of Ecology 103:1233–1244
2. Wang L, Zhang L, George TS, Feng G (2022) A core microbiome in the hyphosphere of arbuscular mycorrhizal fungi has
functional signicance in organic phosphorus mineralization. New Phytologist 238: 859-873
3. Wang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naive Bayesian classier for rapid assignment of rRNA sequences into
the new bacterial taxonomy. Applied and environmental microbiology 73:5261–5267
4. Wang X-X, Hoand E, Feng G, Kuyper TW (2020) Arbuscular mycorrhizal symbiosis increases phosphorus uptake and
productivity of mixtures of maize varieties compared to monocultures. Journal of Applied Ecology 57:2203–2211
5. Wang X-X, Hoand E, Feng G, Kuyper TW (2017) Phosphate uptake from phytate due to hyphae-mediated phytase
activity by arbuscular mycorrhizal maize. Frontiers in plant science 8:684
. Weisser WW, Roscher C, Meyer ST, et al (2017) Biodiversity effects on ecosystem functioning in a 15-year grassland
experiment: patterns, mechanisms, and open questions. Basic and Applied Ecology 23: 1-73

Page 14/20
7. Zhang S, Lehmann A, Zheng W, et al (2019) Arbuscular mycorrhizal fungi increase grain yields: A meta-analysis. New
Phytologist 222:543–555
Tables
Table 1.ANOVA analyses on linear models searching for the effect of wheat diversity, in terms of numbers (Nb) of varieties or
of functional clusters, on two soil enzyme activities (leucine amino-peptidases and phosphatases), AMF abundance in wheat
roots and AMF diversity (Hill indices) evaluated using two markers targeting the Large Sub-Unit (LSU) or Small Sub-Unit
(SSU) region of the ribosomal DNA.

Page 15/20
Explanatory
variables Chisquare Df Pr(>Chisq) P-
value Intercept slope
Leucine amino peptidase
activities Nb varieties 11.6436 3 0.008709 **  
Nb clusters 9.1372 3 0.027521 *
Total N 9.3264 1 0.002259 ** 0.219 0.019
Phosphatase activities Nb varieties 0.3782 3 0.9447   
Nb clusters 4.4307 3 0.2186
Total N 27.1270 1 1.905e-07 *** 2.96 -0.22
AMF abundance   ¤ Nb varieties 11.7872 3 0.008149 **  
Nb clusters 6.1081 3 0.106467
Pi Olsen 8.8856 1 0.002874 ** 7.092 -0.208
Total N 8.9464 1 0.002780 ** 7.092 0.179
SSU Hill 0 ¤ Nb varieties 3.9492 3 0.2670   
Nb clusters 3.4090 3 0.3328
Total N 16.5858 1 4.65e-05 *** 42.31
SSU Hill 1 ¤ Nb varieties 5.1316 3 0.1624113   
Nb clusters 9.0520 3 0.0286080 *
Total N 11.3465 1 0.0007559 *** 0.431 0.054
SSU Hill 2 ¤ Nb varieties 4.5809 3 0.205185   
Nb clusters 9.2412 3 0.026250 *
Total N 7.5348 1 0.006052 ** 0.113 0.018
LSU Hill 0 ¤ Nb varieties 3.2251 3 0.3582   
Nb clusters 2.2177 3 0.5285   
LSU Hill 1 ¤ Nb varieties 2.2634 3 0.5195775   
Nb clusters 9.2822 3 0.0257653 *
Total C 11.7779 1 0.0005994 *** 1.339 0.113
LSU Hill 2 ¤ Nb varieties 1.0462 3 0.7900645
Nb clusters 7.3533 3 0.0614488 .
Total C 11.0527 1 0.0008856 *** 1.339 0.113
¤ log transformed

Table 2.Blast results (NCBI) of the ten most abundant Amplicon Sequence Variants (ASV), representing 92.2 % of the
sequences of the Large Sub-Unit (LSU) dataset and 96.5 % of the Small Sub-unit (SSU) dataset. When possible, the closest
blast was selected from reliable taxonomic identied species only (isolates and strains, referenced in collections).

Page 16/20
rDNA
portion ASV % of
sequences Genus Blast species E
value Per.
Ident Accession
LSU ASV_1 67.8
Funneliformis Funneliformis
caledonium
3E-
156 97.58% FN547496.1
ASV_18 11.6
Funneliformis Funneliformis
caledonium
3E-
156 97.58% JQ048873.1
ASV_40 3.21
Funneliformis Funneliformis
caledonium
3E-
156 97.58% JQ048874.1
ASV_58 2.93
Funneliformis Funneliformis
mosseae
1E-
169 100.00% AY541909.1
ASV_49 2.04
Funneliformis Funneliformis
geosporum
2E-
158 97.89% EU931263.1
ASV_102 1.27
Funneliformis Funneliformis
geosporum
3E-
160 98.18% EU931263.1
ASV_77 1.1
Funneliformis Funneliformis
caledonium
1E-
159 98.18% JQ048874.1
ASV_91 0.9
Rhizophagus Rhizophagus
irregularis
3E-
165 99.39% HF968916.1
ASV_151 0.77
Funneliformis Funneliformis
caledonium
1E-
154 97.28% JQ048874.1
ASV_85 0.7
Septoglomus Septoglomus
constrictum
2E-
163 98.79% AF304971.1
SSU ASV_1 93.6
Funneliformis Funneliformis
caledonium
3E-
126 100.00% KU136397.1
ASV_86 0.7
Scutellospora Scutellospora
calospora
1E-
124 99.60% KU136427.1
ASV_105 0.5
Archaeospora uncultured
Archaeospora
2E-
123 99.60% MH629114.1
ASV_80 0.4
Archaeospora Archaeospora
trappei
3E-
126 100.00% Y17634.3
ASV_155 0.3
Archaeospora Archaeospora
trappei
1E-
124 99.60% Y17634.3
ASV_95 0.2
Diversispora Diversispora sp.
6E-
123 99.20% MF621782.1
ASV_78 0.2
Uncult.Glomeromycotina Uncult.
Glomeromycotina
2E-
123 99.60% JN794959.1
ASV_138 0.2
Uncult.
Glomeromycotina Uncult.
Glomeromycotina
3E-
125 100.00% LT833532.1
ASV_206 0.1
Funneliformis uncultured
Funneliformis
6E-
118 98.01% MH629586.1
ASV_122 0.1
Diversispora Diversispora
aurantia
9E-
126 100.00% EF581880.1
Figures

Page 17/20
Figure 1
Aerial picture and design of the eld experiment. A. Spatial distribution of the wheat plots in the eld experiment (with x and
y-coordinates), each plot being buffered by rows of triticale. B. Scheme displaying how wheat diversity varied in terms of
variety number and number of functional clusters of these varieties. Plots composed of varieties belonging to a single
functional cluster are called “functionally homogeneous mixtures”.

Page 18/20
Figure 2
Boxplots presenting the values of: A. Grain yield (t ha-1), B. Shoot dry biomass (g for 50 plants), C. Total N in dry shoot
biomass (mg. on 50 plants), D. Leucine amino-peptidase activity (LAP; nmol. of substrate g-1.min-1), E. Phosphatases activity
(µmol. of substrate g-1.min-1), F. AMF abundance in wheat roots (gene copy number ng-1 root DNA), G. AMF diversity in roots
(Hill 1 index) on the LSU marker, H. AMF diversity in roots (Hill 1 index) on the SSU marker, for each of the four functional
clusters. For each variable, signicant differences between clusters are indicated by different letters (p <0.05), and bold
points represent the marginal means (with condence interval at 0.95) from the linear mixed model.

Page 19/20
Figure 3
Boxplot showing variation according to the number of wheat varieties (varietal diversity, left panels) and the number of
wheat functional clusters (functional diversity, right panels) of A.B. soil leucine amino-peptidase (LAP) activities (nmol. of
substrate mineralized per gram of soil per minute) in the root zone, C.D. Phosphatase activities (nmol. of substrate g-1.min-1)
in the root zone, E.F. AMF abundance in wheat roots (AMF gene copies ng-1. root DNA), and alpha diversity of AMF index Hill
1 calculated from G.H. the LSU sequencing data and I.J. the SSU sequencing data. For each panel, signicant differences
between variety numbers or cluster numbers are indicated by different letters (p <0.05; or Ns for non-signicant according to
Tukey tests), and bold points represent the marginal means from the used model.

Page 20/20
Figure 4
Correlation between the activity of leucine amino-peptidases (LAP) and arbuscular mycorrhizal fungi (AMF) abundance in
roots, across all the 88 wheat plots. The Pearson coecient (R) and p-value are indicated.
Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.
supplementarytable.docx