ORIGINAL PAPER
Arbuscular mycorrhizal fungi reduce the differences
in competitiveness between dominant and subordinate
plant species
Pierre Mariotte & Claire Meugnier & David Johnson &
Aurélie Thébault & Thomas Spiegelberger &
Alexandre Buttler
Received: 18 June 2012 /Accepted: 2 October 2012 /Published online: 14 October 2012
#
Springer-Verlag Berlin Heidelberg 2012
Abstract In grassland communities, plants can be classified
as dominants or subord ina tes acc ordi ng to their rela tive
abundances, but the factors controlling such distributions
remain unclear. Here, we test whether the presence of the
arbuscular mycorrhizal (AM) fungus Glomus intraradices
affects the competitiveness of two dominant (Taraxacum
officinale and Agrostis capillaris) and two subordinate
species (Prunella vulgaris and Achillea millefolium). Plants
were grown in pots in the presence or absence of the fungus,
in monoculture and in mixtures of both species groups with
two and four species. In the absence of G. intraradices,
dominants were clearly more competitive than subordinates.
In inoculated pots, the fungus acted towards the parasitic
end of the mutualism–parasitism continuum and had an
overall negative effect on the growth of the plant species.
However, the negative effects of the AM fungus were more
pronounced on dominant species reducing the differences in
competitiveness between dominant and subordinate species.
The effects of G. intraradices varied with species composi-
tion highlighting the importance of plant community to
mediate the effects of AM fungi. Dominant species were
negatively affected from the AM fungus in mixtures, while
subordinates grew identically with and without the fungus.
Therefore, our findings predict that the plant dominance
hierarchy may flatten out when dominant species are more
reduced than subordinate species in an unfavourable AM
fungal relationship (parasitism).
Keywords Competitive effects
.
Glomus intraradices
.
Grasslands
.
Mutualism–parasitism
.
Plant diversity
.
Plant hierarchy
.
Subordinate species
Introduction
Semi-natural grasslands are widespread components of north-
temperate landscapes and have important roles in providing
grazing for livestock and acting as reservoirs of both carbon
(Follett and Reed 2010) and biodiversity (Cremene et al.
2005;Bauretal.2006). In particular, grasslands developing
on calcareous substrates tend to contain a large diversity of
P. Mariotte (*)
:
C. Meugnier
:
A. Thébault
:
T. Spiegelberger
:
A. Buttler
School of Architecture, Civil and Environmental Engineering
(ENAC), Laboratory of Ecological Systems (ECOS),
Ecole Polytechnique Fédérale de Lausanne (EPFL),
Station 2,
1015 Lausanne, Switzerland
e-mail: pierre.mariotte@epfl.ch
P. Mariotte
:
A. Thébault
:
A. Buttler
Swiss Federal Institute for Forest,
Snow and Landscape Research (WSL),
Site Lausanne, Station 2,
1015 Lausanne, Switzerland
D. Johnson
Institute of Biological and Environmental Sciences,
University of Aberdeen,
Cruickshank Building, St Machar Drive,
Aberdeen AB24 3UU, UK
T. Spiegelberger
Irstea, Research Unit Mountain Ecosystems (UR EMGR),
2 rue de la Papeterie, Saint-Martin-d’Hères, BP 76,
38402 Grenoble, France
A. Buttler
Laboratoire de Chrono-Environnement,
UMR CNRS 6249, UFR des Sciences et Techniques,
Université de Franche-Comté,
16 route de Gray,
25030 Besançon, France
Mycorrhiza (2013) 23:267–277
DOI 10.1007/s00572-012-0465-8
plants including several rare, threatened and iconic species.
For example, in the Swiss Jura wood pastures, plant commu-
nities result from a traditional management with regular graz-
ing and trampling disturbance and have a long succession
history with well-established vegetation communities where
up to 40 species may inhabit 1 m
2
(Gigon and Leutert 1996;
Buttler et al. 2009). Given their crucial and contrasting eco-
system services, it is important that species-rich grasslands are
conserved. This aim requires understanding of the processes
that govern the composition and stability of species-rich plant
communities.
Observations of plant species abundances in semi-natural
grasslands reveal distinct frequency distributions, with some
species found frequently and in high abundance (dominants),
and some also found frequently but in low abundance
(subordinates; Grime et al. 1987; Olff and Bakker 1998).
Some species may also be considered transients because they
rarely persist (Whittaker 1965). Dominant species are gener-
ally few in number, tall and account for a large proportion of
the total community biomass. In contrast, subordinate species
consistently co-occur with particular dominants, are often
small in stature and contribute marginally to the total biomass
of the community, although they are the most diverse compo-
nent of communities. The role of dominant species in ecosys-
tem functioning has received considerable research attention
and, according to the “mass ratio ” theory (Grime 1998),
ecosystem properties are determined by dominant species
independent of changes in species richness that involve var-
iations in the number of subordinate species. However, more
recent studies show that less abundant species may have a
larger influence on ecosystem properties and functioning than
their relative abundance suggests (Lyons et al. 2005;Boeken
and Shachak 2006). Thus, it becomes important to consider
the factors that determine the abundance not only of domi-
nants but also of subordinates.
In semi-natural, extensively managed grasslands, the
roots of almost all plants are heavily colonised by arbuscular
mycorrhizal (AM) fungi (Read et al. 1976). AM fungi often
form large mycelial networks throughout soil and greatly
facilitate acquisition and uptake of scarce or immobile min-
eral nutrients, particularly phosphorus (Johnson et al. 2001).
In forming mycelial networks, individual mycor rhizal fungi
can be supported by several host plants, and so the hyphae
can facilitate seedling establishment (van der Heijden et al.
2004) and affect plant competition. Whilst the AM symbi-
osis is typically considered a mutualistic one, increasing
evidence points to a more diverse range of interactions. In
a meta-analysis, around 45 % of studies found positive effects
of AM fungi on plant growth, 30 % showed no effect and
25 % showed negative effects (van der Heijden and Horton
2009). This analysis supports the idea that AM fungi act along
a continuum between mutualism and parasitism (Johnson et
al. 1997; Klirono mos 2003) and species competitiveness
could be increased as well as decreased in the presence of
AM fungi. Here, the net outcome of the symbiosis is likely
dependent on the environment or may develop in different
directions throughout the lifetime of a plant.
The contrasting functional characteristics of individual
AM fungi therefore suggest that they may have important
roles in shaping plant community composition and stability
in grasslands. The effect of AM fungi on plant communities
is also related to nutrient availability in soils, and it has been
shown that low phosphorus status stimulates colonisation of
roots by AM fungi, which feeds back to affect plant diver-
sity and productivity (Collins and Foster 2009). AM fungi
are completely dependent on host plants for carbon, and
their biomass and activity means they are substantial sinks
for p lant assimilate (Johnson et al. 2002), which could
therefore completely change the dominance hierarchy
(Gross et al. 2010) following its degree of profit from
particular plant species. Indeed, previ ous studies have dem-
onstrated that subordinate and dominant plant species can
show distinct responses to AM fungi (van der Heijden et al.
1998a; Yao et al. 2007; Karanika et al. 2008) which may
influence pl ant domi nance (Gr ime et al. 1987
;vander
Heijden et al. 1998b; Hartnett and Wilson 1999). Consider-
ing that AM fungi can act along a continuum betwee n
mutualism and parasitism (Johnson et al. 1997), there are
two mechanisms which may explain the effects of AM fungi
on plant dominance hierarchies: (1) dominant and subordi-
nate species have different AM fungal dep endency and
species with higher dependency are favoured in competition
in mutualistic interactions or (2) dominant and subordinate
species show different responses to less favourable relation-
ships with AM fungi, and one of these groups may be more
affected by parasitism. For the first mechanism, these
responses have been developed into a model whereby the
relative response of dominants and s ubordinates to AM
fungi determines plant community composition and domi-
nance hierarchies (Urcelay and Diaz 20 03 ). When mycor-
rhizal dependence of subordinates is strong, dominance
rankings flatten out, so that overall plant species diversity
increases. In contrast, when mycorrhizal dependence of
dominants is strong and subordinates is weak, dominance
rankings steepen in response to AM fungi, so that overall
plant species diversity decreases. In the second mechanism,
when dominant species are more affected than subordinate
species in unfavourable relationships with AM fungi, dom-
inance rankings may flatten out, and inversely, when subor-
dinate species show higher biomass reduction than
dominant species, dominance ranking steepens in the pres-
ence of AM fungi. To date, there have been few studies that
explicitly test these hypotheses, with most of the evidence
derived indirectly from experiments or observations related
to the relative influence of mycorrhizal symbiosis on plant
communities (Klironomos et al. 2011).
268 Mycorrhiza (2013) 23:267–277
Plant communities of wooded pastures in the Swiss Jura
Mountains result from traditional cattle and forest manage-
ment, and vegetation communities are well-established on
unimproved soils with intermediate fertility. The se grassland
communities serve as an ideal model to explore the effect of
AM fungi on plant interactions since they are very diverse
and show a typical lognormal rank–abundance curve (Grime
1998) with a few dominant species accounting for a high
proportion of the total community biomass. In this study, we
chose four plant species which are native to species-rich
calcareous grasslands. According to their relative abundance
in the field, two of these species can be considered to be
dominants and two as subordinates. Previous experiments in
the field (unpublished) showed that these subordinate spe-
cies were significantly more colonised by AM fungi (about
20 % more) compared to dominants during seedling estab-
lishment. AM fungi could therefore be more beneficial to
less competitive subordinates and explain their persistence
in plant communities. In this paper, we report investigations
of the competitive effect of two dominant and two subordi-
nate species in different speci es composition when grown
under control conditions and in the presence or absence of
the ubiquitous AM fungus Glomus intr aradices.We
hypothesise that inoculation with G. intraradices will pro-
mote growth of subordinate plant species and reduce the
differences in competitiveness with dominant species which
will not or less draw benefit from the fungus.
Materials and methods
Plant species selection
Dominant and subordinate species were selected from meas-
urements of community composition in semi-natural calcare-
ous grasslands in the Swiss Jura Mountains (Les Amburnex,
western Switzerland, 6°13′50″ E, 46°32′50″ N). The field
research site is an extensively grazed pasture with a vertic
cambisol which contains intermediate value of carbonates
(15 % CaCO
3
) and relatively low level of available P
(Table 1). Absolute plant cover was determined within 36
plots (1.2×1.2 m) using a Braun-Blanquet index. A species
was classified as dominant if its frequency was greater than
75 % (100 % frequency means that the species is present in
all plots) and its cumulative relative cover greater than 25 %.
A species was classified as subordinate if its frequency was
greater than 75 % and its cumulative relative cover between
2 and 25 % (adapted from Grime 1998). Based on their field
abundance and results of their competitive abilities (effect
and response) when growing together in a previous glass-
house experiment (Mariotte et al. 2012), we assigned
Prunella vulgaris and Achillea millefolium as subordinate
species and Agrostis capillaris and Tara xacum officinal e as
dominant species.
Experimental design
The pot experiment was set up as a randomised split-block
design (Fig. 1) with two factors: (1) AM fungal inoculation
treatment varying at the block level and (2) plant mixture
type varying at the plot level. The inoculation treatment
contained two levels: w ith G. intraradices inoculum (M
treatment) and with non-mycorrhizal control inoculum
(NM treatment). We chose the ubiquitous G. intraradices
because it has a broad host range and is the most frequently
detected phylotype in all grassland systems (Öpik et al.
2006) and the most abundant in Swiss meadows (Sýkorová
et al. 2007). In this pot experiment, we used soil inoculum of
G. intr aradices Schenck & Smith (isolate BEG 21), native
from a calcareous grassland, which was provided by Prof.
M.G.A. van der Heijden, Agroscope ART Zurich, Switzer-
land. G. intraradices was previously pro pagated from
spores as a pure culture of Plantago lanceolata growing in
autoclaved soil for 8 months in 2009 to provide recent high-
quality inoculum (further details about t his inoculum in
Veiga et al. 2012). The plant mixture type contained three
levels: monocultures (n0 4), two-species mixtures including
each combination of one dominant and one subordinate (n0
4) and four-species mixture including two different domi-
nants and two different subordinates (n 0 1). All pots
contained four individuals to maintain stable plant densities
for each mixture type. The interaction of mixture type and
AM inoculation treatment resulted in 18 combinations
which were replicated six times, yielding a total of 108 pots
arranged in six blocks.
Seeds germination, soil mixture and growing conditions
Seeds of th e four species were g athere d from the Swiss
plateau (provided b y FENACO SA, Yverd on-les-bains,
Switzerland), stored at 4 °C prior to the start of the exper-
iment, surface sterilised in 5 % bleach for 10 min and then
Table 1 Soil characteristics of the field soil (Les Amburnex) and the
pot mixture used in the glasshouse experiment
Field site Pot mixture
Available P (mgkg
−1
)
acetate NH
þ
4
þ EDTA 1 : 10
4.47±0.15 4.13±0.12
Total P (%) 0.18±0.03 0.21±0.02
Available Ca (gkg
−1
)
acetate NH
þ
4
þ EDTA 1 : 10
11.27±0.29 13.91±0.76
Total organic C (%) 5.82±0.22 4.09±0.19
Total N (%) 0.67±0.06 0.57±0.03
pH (H
2
O) 6.81±0.06 7.76±0.04
Mycorrhiza (2013) 23:267–277 269
rinsed five times with distilled water. Around 700 seeds of
each species were germinated on autoclaved water agar gel
(1.6 %) in a phytotron over 6 days in the dark at 22 °C. One
week later, seedlings were transplanted wi thin an agar gel
cube (1 cm
3
) and assembled following the different species
mixtures into 1.5 L sterile plastic pots containing 800 ml of
a gamma sterilised sand (Berns et al. 2008) mixture (90 %
sand and 10 % compost) which recreated the low P avail-
ability measured in the field sit e (Table 1). Half of the pots
received 40 ml of G. intraradices inoculum added on the top
soil (M treatment), while the control pots received 40 ml of
autoclaved inoculum (NM treatment). To compensate for
the microbial comm unity which may coexist with AM fungi
in mycorrhizal pots, it is common practice to add an inoc-
ulum washing (filtered over 20 μm) to the non-mycorrh izal
pots (e.g. Gavito et al. 2003; Van der Heijden et al. 2003;
Chen et al. 2007). However, this does not necessarily in-
clude all microbes, and the microbial community added
alone may have different effects on plant growth when
added alone in the control pots (positive or negative due to
changes in soil interactions) as to when coexisting with the
AM fungus in mycorrhizal pots. For this reason, and as no
effects of pathogens were detected in previous experiments
using the same high-quality inoculum of G. intraradices
BEG21 (Van der Heijden et al. 1998a, 2006; Wagg et al.
2011; Veiga et al. 2012), a microbial wash was not added to
control pots. Neverthel ess, the inoculum was checked for
the absence of other micro organisms using two different
approaches. Firstly, an inoculum washi ng (filt ered ove r
20 μm) was put into culture (six replicates) in nutrient
agar-agar (Sigma-Aldrich, St. Louis, MO, USA), and no
significant development of bacteria or fungi was detected
after 5 days. Secondly, P. lanceolata roots in the inoculum
were stained with acid fuchsin (Gerdemann 1955)and
scored f or root colo nisation on 50 intersections per root
sample (10 replicates). Ninety percent of P. lanceolata root
length was colonised by G. intraradices (see also Veiga et
al. 2012 ), and fungal structures other than those belonging
to AM fungi were not observed.
Pots were maintained in a glasshouse from April to July
2010 where day temperature varied between 25 and 38 °C,
night temperature was 16 °C and the day length was 15 h.
Because the glasshouse permitted 90 % penetration of photo-
synthetically active radiation, additional light was not provid-
ed. During the first month, each pot was watered every day
with the same volume of water (200 ml) and then adjusted to
maintain 10 % water content. Plants received no fertilisation.
After 105 days, shoots were harvested for each species by
cutting at ground level, and roots were collected by washing
and sieving each individual. Root subsamples of selected
individuals were used for determination of AM fungal col-
onisation. Shoot and root biomass were measured after
drying in an oven at 60 °C for 72 h. The average of the
two (two-species mixture) or four (monoculture) individuals
per species in each pot was calculated to obtain the overall
root and shoot biomass per species per pot. As we did not
Fig. 1 Schematic diagram of
the experimental design. Circles
represent pots and contain plant
combinations of dominant
species: T. officinale (D1) and
A. capillaris (D2), and
subordinate species: P. vulgaris
(S1) and A. millefolium (S2).
Each pot contained four
individuals of the same species
in monoculture (white), two
individuals per species (one
dominant and one subordinate)
in two-species mixture (grey)
and one individual per species
(two dominants and two subor-
dinates) in mixture of four spe-
cies (dark grey). Combinations
are reproduced either with (M )
or without (NM) AM fungal in-
oculation which varied at the
plot level in six blocks
270 Mycorrhiza (2013) 23:267–277
observe any differences in root/shoot ratios following inoc-
ulation with G. intraradices, or by changing the mixtures of
species, we focused on total plant biomass as the response
variable in our analyses.
Estimation of AM fungal colonisation
The primary objective was to determine whether the inoculum
had successfully colonised roots rather than to test for effects
of mixture type and plant identity. Roots from each plant
species in each community composition (four replicates) were
stained with acid fuchsin as described by Gerdemann (1955).
A minimum of 50 intersections per root sample were viewed
microscopically and scored for the presence of vesicles, hy-
phae only and arbuscules. The absence of colonisation was
also confirmed in uninoculated plants.
Relative effect of G. intraradices on plant growth
For each plant species grown in monoculture, as two-species
mixtures or as four-species mixtures, the relative effect of G.
intraradices was measured as:
AM fungal effect
ij
¼ ln Y
ij
M=Y
ij
NM
where Y
ij
M represents total plant biom ass (root and shoot) of
speci es i growing with species j in presence of the AM
fungal inoculum and Y
ij
NM the biomass of the same species
i growing with species j without the AM fungal inoculum.
Positive values for AM fungal effect indicate a better growth
in presence of G. intraradices (mutualism) and, inversely,
negative values indicate a negative growth response to in-
oculation by G. intraradices (parasitism).
Competitiveness of dominant and subordinate plant species
The relative yield per plant (RYP) of each species was
calculated from total biomass (root and shoot) in the mycor-
rhizal (M) and non-m ycorrhizal (NM) treatments of two-
and four-species mixtures, as:
RYP
ij
¼ Y
ij
=Y
ii
where RYP
i
is the relative yield per plant of species i (i.e.
target) grown with species j (i.e. neighbour), Y
ij
is the yield
of plant i when grown with species j and Y
ii
is the yield of
plant i when grown in monoculture (Engel and Weltzin
2008). When RYP
ij
>RYP
ji
, j is less competitive than spe-
cies i and conversely, when RYP
ij
<RYP
ji
, j is a better
competitor than i .
Competitive effect (CE) of dominant species is defined as
the mean RYP of T. officinale and A. capillaris grown with
each subordinate species, and competitive effect of subordi-
nate species is defined as the mean RYP of P. vulgaris and A.
millefolium grown with each dominant species (Goldberg and
Landa 1991). Species with a greater CE are better competitors,
and inversely, species with a lower CE are less competitive. To
test whether competitiveness changed when plants were
colonised by G. intraradices, the CE of dominant and subor-
dinate species was calculated for both mycorrhizal and non-
mycorrhizal plants in two- and four-species mixtures.
Statistical analysis
Statistical analyses were undertaken in R version 2.11.1 (R
Development Core Team 2010). Total plant biomass (after
square root transforming), relative AM fungal effect and
competitive effect were analysed using linear mixed-
effects models with plot nested into block as random factors
followed by Tukey’s post hoc. For total plant biomass,
datasets from each mixture type were analysed separately,
and the analysis of relativ e AM fungal effect was a lso
separated for each species.
Results
Colonisation of roots by G. intraradices
The root systems of the plant species inoculated with G.
intraradices were heavily colonised, but AM fungal coloni-
sation did not significantly differ between species. Coloni-
sation of roots by vesicles (mean of 46 % for T. officinale,
56 % for A. capillaris, 80 % for P. vulgaris and 63 % for A.
millefolium) and hyphae only (30 % for T. officinale,33%
for A. capillaris,11%forP. vulgaris and 24 % for A.
millefolium) was high, but there was little development of
arbuscules (less than 2 % for each species). Plan ts not
inoculated with G. intraradices remained uncolonised.
Plant productivity in experimental systems
Both su bordinate and dominant plant species either pro-
duced less or equal biomass in the presence of G. intra-
radices compared to un-inoculated controls (Fig. 2). The
effects of the AM fungus were very consistent among spe-
cies when they were grown in monoculture (Fig. 2a), reduc-
ing biomass by between 42 and 44 % compared to the non-
mycorrhizal controls. This resulted in an overall significant
effect (F
1,5
0 95.69, P<0.001) of the presence of G. intra-
radices and an overall significant effect of species (F
3,30
0
12.32, P<0.001) but no interaction (F
3,30
0 0.29, P0 0.82).
The patterns in plant biomass were broadly simil ar when
plants were grow n in two-species (Fig. 2b) or four-species
(Fig. 2c) mixtures. In the two-species mixture, there was an
overall significant interaction between addition of G. intra-
radices×plant species (F
3,78
0 10. 14, P <0.001). Both
Mycorrhiza (2013) 23:267–277 271
