A novel proxy to examine interspecific phosphorus facilitation between plant species

Abstract

Resource complementarity can contribute to enhanced ecosystem functioning in diverse plant communities, but the role of facilitation in the enhanced complementarity is poorly understood.
Here, we use leaf manganese concentration ([Mn]) as a proxy for rhizosheath carboxylate concentration to explore novel mechanisms of complementarity mediated by phosphorus (P) facilitation.
In pot experiments, we showed that mixtures involving Carex korshinskyi, an efficient P‐mobilizing species, exhibited greater biomass and relative complementarity effect than combinations without C. korshinskyi on P‐deficient soils. Compared with monocultures, leaf [Mn] and [P] of species that are inefficient at P mobilization increased by 27% and 21% when grown with C. korshinskyi (i.e. interspecific P facilitation via carboxylates) rather than next to another inefficient P‐mobilizing species. This experimental result was supported by a meta‐analysis including a range of efficient P‐mobilizing species. Phosphorus facilitation enhanced the relative complementarity effect in low‐P environments, related to a greater change in several facilitated species of their root morphological traits relative to those in monoculture.
Using leaf [Mn] as a proxy, we highlight a vital mechanism of interspecific P facilitation via belowground processes and provide evidence for the pivotal role of P facilitation mediated by the plasticity of root traits in biodiversity research.

Full text

A novel proxy to examine interspecific phosphorus facilitation
between plant species
Rui-Peng Yu
1
,YeSu
1
, Hans Lambers
2,3
, Jasper van Ruijven
4
, Ran An
1
, Hao Yang
1
,
Xiao-Tong Yin
1
, Yi Xing
1
, Wei-Ping Zhang
1
and Long Li
1
1
Key Laboratory of Plant-Soil Interactions, Ministry of Education, Beijing Key Laboratory of Biodiversity and Organic Farming, College of Resources and Environmental Sciences, China
Agricultural University, Beijing 100193, China;
2
School of Biological Sciences and Institute of Agriculture, University of Western Australia, 35 Stirling Highway, Perth, WA 6009, Australia;
3
National Academy of Agriculture Green Development, China Agricultural University, Beijing 100193, China;
4
Plant Ecology and Nature Conservation Group, Wageningen University &
Research, PO Box 47, Wageningen 6700 AA, the Netherlands
Author for correspondence:
Long Li
Email: lilong@cau.edu.cn
Received: 11 February 2023
Accepted: 24 May 2023
New Phytologist (2023)
doi: 10.1111/nph.19082
Key words: biodiversity–ecosystem
functioning, carboxylate release, Carex
korshinskyi, phosphorus-acquisition root
traits, relative changes of root traits, relative
complementarity effect.
Summary
Resource complementarity can contribute to enhanced ecosystem functioning in diverse
plant communities, but the role of facilitation in the enhanced complementarity is poorly
understood.
Here, we use leaf manganese concentration ([Mn]) as a proxy for rhizosheath carboxylate
concentration to explore novel mechanisms of complementarity mediated by phosphorus (P)
facilitation.
In pot experiments, we showed that mixtures involving Carex korshinskyi, an efficient P-
mobilizing species, exhibited greater biomass and relative complementarity effect than combi-
nations without C. korshinskyi on P-deficient soils. Compared with monocultures, leaf [Mn]
and [P] of species that are inefficient at P mobilization increased by 27% and 21% when
grown with C. korshinskyi (i.e. interspecific P facilitation via carboxylates) rather than next to
another inefficient P-mobilizing species. This experimental result was supported by a meta-
analysis including a range of efficient P-mobilizing species. Phosphorus facilitation enhanced
the relative complementarity effect in low-P environments, related to a greater change in sev-
eral facilitated species of their root morphological traits relative to those in monoculture.
Using leaf [Mn] as a proxy, we highlight a vital mechanism of interspecific P facilitation via
belowground processes and provide evidence for the pivotal role of P facilitation mediated by
the plasticity of root traits in biodiversity research.
Introduction
Plant diversity is usually positively correlated with ecosystem
functioning, for example, enhanced productivity and nutrient
uptake (Tilman et al., 2014; Barry et al., 2019a; Brooker
et al., 2021). The positive biodiversity–ecosystem functioning
(BEF) relationships in diverse communities may occur because
species with large biomass in monoculture can also win in com-
petition in mixtures, coined a positive selection effect (Loreau &
Hector, 2001). Increased ecosystem functioning in communities
may also result from a positive complementarity effect, which
occurs when species generally perform better in mixtures, regard-
less of their monoculture biomass. This effect may arise from
processes including niche differentiation (e.g. temporal and spa-
tial resource partitioning), direct abiotic facilitation (e.g. nutrient
enrichment), and indirect biotic facilitation (e.g. disease suppres-
sion, Loreau & Hector, 2001; Wright et al., 2017; Barry
et al., 2019a). A positive complementarity effect, rather than a
positive selection effect, contributes more to greater ecosystem
productivity in diverse plant communities (van Ruijven &
Berendse, 2005; Cardinale et al., 2007), while spatial resource
partitioning alone does not always provide a full explanation for
a positive complementarity effect in grasslands (Jesch et al., 2018;
Barry et al., 2019b). Therefore, facilitation is receiving increasing
attention as a mechanism enhancing ecosystem functioning. This
occurs when one or more facilitated species in diverse plant com-
munities increase their productivity in the presence of a facilitat-
ing species (Wright et al., 2017;Yuet al., 2021). Even if the
enhanced performance of facilitated species is usually regarded as
interspecific facilitation, there are few proxies to study mechan-
isms of facilitation in species-diverse communities.
It is widely accepted that legumes may facilitate neighbors via
belowground facilitation of nitrogen (N) acquisition driven by
increased soil N availability via dinitrogen (N
2
) fixation in sym-
biosis with soil microorganisms (Wright et al., 2017; Barry
et al., 2019a). However, beyond N facilitation, belowground
phosphorus (P) facilitation has received little attention in BEF
studies (Yu et al., 2021). Carboxylates and protons in root
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023) 1
www.newphytologist.com
Research

exudates mobilize sorbed P, and phosphatases hydrolyze soil
organic P and convert it to plant-available inorganic P (Lambers
et al., 2006). Our previous findings showed that grassland species
vary in their capacity to mobilize sorbed or organic soil P. For
example, some sedges (e.g. Carex korshinskyi and Carex durius-
cula) and forbs (e.g. Potentilla tanacetifolia and Artemisia frigida)
exhibit relatively greater release of carboxylates or phosphatases
than grass species do (e.g. Stipa grandis and Cleistogenes squarrosa)
(Yu et al., 2020a,b). Species that exude phosphatases, carboxy-
lates, or protons into rhizosheath soil, mobilizing and converting
organic or sorbed soil P to plant-available forms for plant uptake,
are defined as efficient P-mobilizing species. Conversely, species
with a limited capacity to mobilize sorbed or organic P are con-
sidered species that are inefficient at P mobilization (Horst
et al., 2001;Liet al., 2014).
When growing species with divergent P-mobilizing capacities
together, interspecific P facilitation by efficient P-mobilizing spe-
cies may enhance leaf P concentration ([P]), productivity, and/or
shoot P content of neighboring species that are inefficient at P
mobilization (Karanika et al., 2007;Liet al., 2007,2014). Plant
species can integrate information relating to neighbor identity
and nutrient availability in mixtures and coordinate root growth
based on nutrient status and signals (Cahill et al., 2010). Species
that are inefficient at P mobilization usually exhibit greater speci-
fic root length, root branching intensity, or root hair length than
efficient P-mobilizing species (Lambers et al., 2006; Wen
et al., 2019). These root morphological traits may make them
more efficient at acquiring P when soil P availability is higher
(Zhang et al., 2016). Our previous study showed that compared
with monocultures, species that are inefficient at P mobilization
but exhibit a greater increase of root morphological traits (e.g.
total root length and proportion of fine roots) in response to a
facilitator may have an enhanced opportunity to be facilitated
compared with other species that exhibit a smaller response of
root traits to a facilitator (Yu et al., 2020a). However, little is
known about the pathways involved in P facilitation based on
measurements of root morphological traits. In addition, because
roots intermingle in mixtures, it is difficult to precisely test how
efficient P-mobilizing species facilitate their neighbors by measur-
ing rhizosphere processes in situ.
Carboxylates not only mobilize sorbed soil inorganic P (Pi) to
unsorbed Pi in the soil solution, but also some metal cations, for
example, manganese (Mn). Then, soil Mn
2+
and Pi can be taken
up by roots from the rhizosphere; Mn availability is also increased
by root-released carboxylates, which chelate Mn and reduce
Mn
4+
to Mn
2+
(Lambers et al., 2015). Recent studies show that
leaf Mn concentration ([Mn]) is correlated with carboxylate con-
centration in the rhizosheath at various soil P levels; leaf P con-
tent is also positively correlated with rhizosheath carboxylates and
root morphological traits of chickpea (Huang et al., 2017; Pang
et al., 2018; Wen et al., 2021). When soil P availability is very
low, facilitated plants may not show increased leaf [P], and hence
having a proxy is very useful, especially in field observations
where increased productivity may be hard to measure (Shen
et al., 2023). Therefore, leaf [Mn] can be used as an easily-
measurable proxy that reflects the mobilization of sorbed soil P
for a given species (Pang et al., 2018; Wen et al., 2021) or a range
of species (Lambers et al., 2015,2021,2022; Lambers, 2022).
Although previous studies have shown that leaf [Mn] of species
that are inefficient at P mobilization may increase in biculture
(Gardner & Boundy, 1983; Muler et al., 2014), to our knowl-
edge, no study has unequivocally demonstrated whether leaf
[Mn] can be used as a proxy to test for interspecific P facilitation
by carboxylates. The roles of interspecific P facilitation and rela-
tive changes of functional traits between mixtures and monocul-
tures in a complementarity effect have received little attention.
Based on previous advances and knowledge gaps, we present
three hypotheses. (1) Species combinations that include efficient
P-mobilizing species will exhibit a greater productivity and com-
plementarity effect, especially on P-deficient soils. (2) Leaf [Mn]
of species that are inefficient at P mobilization is greater when
grown with efficient P-mobilizing species rather than next to
another inefficient P-mobilizing species, implying the existence
of interspecific P facilitation via carboxylates. Such P facilitation
by efficient P-mobilizing species occurs in species-diverse plant
communities in both natural ecosystems and agroecosystems. (3)
An enhanced complementarity effect via interspecific P facilita-
tion is partly mediated by greater relative changes of root traits
between mixtures and monocultures of facilitated species.
To test these hypotheses, we applied a three-step approach:
first, we grew five dominant and subdominant species from a
typical steppe that differ in their capacity to mobilize P at two soil
P levels in monoculture. We classified the species into efficient P-
mobilizing species and inefficient ones based on P-mobilization
traits and tested the relationship between leaf [Mn] and rhi-
zosheath carboxylate concentration among these species. Then,
we grew 10 two-species mixtures, using the additive partitioning
method to calculate the relative selection effect and relative com-
plementarity effect to allow comparison among combinations
with varied biomass (Loreau & Hector, 2001; Craven
et al., 2016). We compared the differences between combinations
that included efficient P-mobilizing species and combinations
without efficient P-mobilizing species. Second, we used leaf [Mn]
and [P] to examine whether leaf [Mn] can be used as a proxy for
interspecific P facilitation by carboxylates. Then, we tested
whether the strength of P facilitation increases with enhanced
dominance of the facilitator in our glasshouse study. We also
examined whether the enhanced leaf [Mn] in bicultures including
efficient P-mobilizing species was a general pattern by using a
meta-analysis. Finally, we analyzed the relative change of root
traits in mixtures than the corresponding monoculture of each
facilitated species. We further tested the role of relative changes
of root traits in interspecific P facilitation (proxied by leaf [Mn])
and relative complementarity effect.
Materials and Methods
Glasshouse study
Experimental setup Five dominant and subdominant species of
a typical steppe were selected, including Stipa grandis P. Smirn.
(perennial bunchgrass, dominant species), Leymus chinensis
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
2
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

(Trin.) Tzvel. (perennial rhizome grass, dominant species), Cleis-
togenes squarrosa (Trin.) Keng (perennial bunchgrass, subdomi-
nant species), Carex korshinskyi Kom. (sedge, subdominant
species), and Artemisia frigida Willd. (forb, subdominant species)
(Yu et al., 2020b). Seeds of each species were collected from mul-
tiple individuals in a typical steppe in Inner Mongolia (43.55°N,
116.70°E) and thoroughly mixed to avoid any genotype domi-
nating in any pots (Yu et al., 2020a). Seeds were surface-sterilized
with hydrogen peroxide (15% v/v) for 15 min and then washed
three times in deionized water.
Soils from the top 20-cm layer were collected in a typical
steppe in Inner Mongolia. The soil was air-dried, passed through
a 2-mm sieve and evenly mixed. The soil properties were as fol-
lows: bulk density, 1.52 g cm
3
; soil texture: sand 84%, silt 10%,
clay 6%; organic matter, 22.5 g kg
1
; total N, 1.16 g kg
1
; total
P, 0.28 g kg
1
; Olsen P, 2.85 mg kg
1
; pH (soil : water ratio =
1 : 2.5), 7.44. We used a P addition with 0 mg kg
1
as low-P
treatment (LP), and 60 mg kg
1
as high-P treatment (HP), sup-
plied as KH
2
PO
4
. The bulk soil Olsen P concentration after P
fertilization was 22.4 mg kg
1
. As the soils were collected in a
typical steppe, which is the natural condition for plant growth,
no additional basal nutrients were supplied. Pots with a base dia-
meter of 12 cm, a top diameter of 18 cm, and a height of 14 cm
were filled with 1.5 kg of air-dried soil. Pots were arranged as ran-
domized complete block design with four replicates for each spe-
cies in monoculture and for each combination in the mixture in
each of the P treatments (LP and HP).
Plant growth conditions The experiment was carried out in a
glasshouse with natural light at China Agricultural University,
Beijing. The temperature was 22–30°C during the day. Seeds of
experimental species with uniform seed mass were sown in pots.
For information of the experiment and the tested species, see
Supporting Information Table S1. For information of the species
combinations, see Table S2. We standardized the density to six
uniform individuals in monoculture and three individuals of each
species in a mixture 30 d after the emergence of all species. Two
species in this study had rhizomes, that is, L. chinensis and C. kor-
shinskyi. We controlled their density on the 30
th
day. No addi-
tional density control was further conducted, although these
species might produce more individuals by rhizome during the
experiment. During the glasshouse experiment, any other species
emerging in pots with a given combination were removed. Each
pot had several small holes at the bottom, through which water
could be taken up. Before planting, all pots were irrigated
through the holes at the bottom, reaching pot capacity. After 15
d of growth, the pots were maintained at 75% of pot capacity
(15% water content) to support the optimal growth measured by
weight, until the final harvest. Plants were harvested 100 d after
sowing.
Harvest and measurements We gently lifted the entire root sys-
tems from the pots and separated these from the soil. Soil clumps
and excess soil were removed, and the soil that tightly adhered to
roots was considered rhizosheath soil (Pang et al., 2017). After
that, shoots were harvested at the base, and we separated the root
samples of each species in monoculture into two even parts, that
is, subsample A and subsample B, for further physiological and
morphological trait measurements.
We carefully lifted the entire root system from its pot and
sorted it into species in the mixtures; the roots attached to the
stem base were the roots of the target species. Then, shoots of
each species in the mixture were harvested at the base. Roots of
each species in the mixture were separated into two even parts as
subsample A and subsample B, that is, subsample A comprised half
of the root samples of species 1 (e.g. S. grandis) and species 2 (e.g.
L. chinensis) in the S. grandis/L. chinensis combination; the root
samples of these two species were stored separately for further
root morphology measurements. Most roots were sorted into
species; the remaining mixed roots (subsample C) were carefully
collected without separating them into species. Since roots were
twined in mixture, we did not measure physiological traits in
mixture, but root morphological traits of each species were
measured.
P-mobilization traits in monoculture. A subsample (subsample
A) in monoculture was transferred into a beaker with 100 ml of
0.2 mM CaCl
2
(Pearse et al., 2007). Roots were gently dunked
for 60 s to remove the rhizosheath soil as much as possible and
minimize root damage. A subsample of 10 ml supernatant was
transferred into a centrifuge tube with two drops of microbial
inhibitor Micropur (Sicheres Trinkwasser, Munich, Germany) at
0.01 g l
1
and three drops of concentrated phosphoric acid. The
samples were stored at 20°C until analysis of carboxylates by
HPLC–MS/MS after passing through a 0.22 μm filter. The mea-
surement of carboxylates followed the method of Fiori
et al.(2018). The remaining soil in the beaker was air-dried and
weighed to calculate carboxylate concentrations based on the
weight of rhizosheath soil.
We used a brush to remove the rhizosheath soil from another
subsample (subsample B) of the root systems in monoculture to
measure rhizosheath acid phosphatase activity (Apase) and pH.
Apase was measured according to Tabatabai & Bremner (1969):
to 1.00 g fresh soil, 0.80 ml acetate buffer at pH 5.2 and 0.20 ml
p-nitrophenyl phosphate (p-NPP) substrate was added to be
incubated at 30°C for 1 h; 1.00 ml of 0.5 M NaOH was added
to terminate the reaction. Absorption was measured spectropho-
tometrically at 405 nm (Uvmini-1240; Shimadzu Corp., Kyoto,
Japan). Rhizosheath pH was measured with a soil : water ratio =
1 : 2.5 with a pH meter (UB-7; Denver Instrument Co., Arvada,
CO, USA).
Root morphological traits in monoculture and mixture. The
subsample A in monoculture and mixture was washed to remove
the remaining soil and intact root segments were selected and
stored in 50% (v/v) ethanol at 4°C for root hair measurement
(Haling et al., 2016). Root hairs were photographed using a
SZX16 Wide Zoom Versatile Stereo Microscope (Olympus
Corp., Tokyo, Japan). The length of 10 root hairs (2–8 cm from
the root tip) for each species per replicate was measured using
IMAGEJ (NIH Image, Bethesda, MD, USA), and the number of
root hairs was counted as root hair density (Haling et al., 2016).
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023)
www.newphytologist.com
New
Phytologist Research 3
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

The subsample B in monoculture and mixture was washed sev-
eral times to remove remaining soil and scanned at 600 dpi (Per-
fection V750 Pro; Epson, Suwa, Japan). Root length and average
root diameter were analyzed using the WINRHIZO scanner-based
system (WINRHIZO system; Regent Instruments Inc., Quebec,
QC, Canada). Fourier-transform infrared (FTIR) spectroscopy
was used to determine the proportion of roots of a given species
in mixed root samples (subsample C) (Streit et al., 2019), as
detailed in Methods S1; Tables S3 and S4. Roots of subsamples A,
B, and Cwere dried at 70°C for 48 h, and then root biomass of
each species and total root biomass were calculated based on sub-
samples A and Bin monocultures, based on all subsamples in mix-
tures.
Harvested shoot samples were oven-dried at 70°C for 48 h,
and weighed and ground for nutrient analyses. Leaf material was
digested and inductively coupled plasma optical emission spec-
troscopy (ICP-OES, OPTIMA 3300 DV; Perkin-Elmer, Wal-
tham, MA, USA) was used for leaf [Mn] and [P] measurement.
Calculation As the above- and belowground biomass differed
among the 10 species combinations, for each species combina-
tion, we standardized the complementarity effect and selection
effect by the mean biomass of two species in monoculture of the
corresponding treatment (i.e. relative complementarity and selec-
tion effect; Craven et al., 2016). Then, the strengths of mixtures
across species combinations can be compared. The relative com-
plementarity and selection effects were calculated following the
additive partitioning approach (Loreau & Hector, 2001; Craven
et al., 2016):
ΔRB ¼RBORBE¼Bmixture=Bmonoculture 0:5
In this equation, RB indicates relative biomass; the observed
RB (RB
O
) is the biomass of a given species in the mixture divided
by that in monoculture. The expected RB (RB
E
) is the propor-
tion of a given species planted in the mixture, which was 0.5 in
this study. The relative complementarity effect (rCE) was calcu-
lated as:
rCE ¼nmean ΔRBðÞmean MðÞ=mean MðÞ
¼nmean ΔRBðÞ
In this calculation, nis the number of species in the mixture
which was always two in this study; Mis the monoculture bio-
mass of the given species. The relative selection effect (rSE) is cal-
culated as:
rSE ¼ncovariance ΔRB, MðÞ=mean MðÞ
A positive relative complementarity effect reflects that, on aver-
age, species perform better in the mixture than in monoculture.
A positive relative selection effect indicates that the enhanced bio-
mass is controlled by the species with greater performance in
monoculture, which also dominates the biomass in the mixture
(Loreau & Hector, 2001).
The calculation of root morphological traits in monoculture
and mixture was based on subsample B. Specific root length was
calculated as total root length divided by root biomass (dry
weight). The proportion of fine roots was calculated as the root
length of fine roots with a diameter <0.2 mm divided by total
root length (Bergmann et al., 2017). Root branching intensity
was calculated as the number of root tips divided by the root
length of the subsample (Kramer-Walter et al., 2016).
The relative change of root traits in the mixture from mono-
culture was calculated as:
Relative change of a given trait ¼traitimix traitimono
ðÞ=traitimono
where trait
imix
and trait
imono
were the targeted trait of a given
species in mixture and monoculture, respectively. The domi-
nance of efficient P-mobilizing species in the mixture was calcu-
lated as the biomass of the given species divided by the total
biomass of the mixture.
Statistical analyses
Monoculture All statistical analyses were conducted in R
v.4.1.3 (R Core Team, 2022). We conducted ANOVAs with
block as a factor. First, we used one-way ANOVAs to examine
the effect of P addition on the biomass, leaf [P] and P content of
each species in monoculture. One-way ANOVAs were used
because species varied in their biomass; we mainly focused on the
effect of P addition on the performance of a given species in this
study. Then, two-way ANOVAs were performed to test the
effects of P addition and species identity on P-mobilization traits
(i.e. rhizosheath Apase, pH and carboxylates) and leaf [Mn].
Tukey’s post-hoc HSD test was performed at the 5% probability
level in ANOVAs. To test the differences in P mobilization capa-
city among species, we carried out a principal component analysis
using P-mobilization traits in monoculture, using the VEGAN
package (Oksanen et al., 2022). Correlations were then con-
ducted to examine the relationship between leaf [Mn] and rhi-
zosheath carboxylate concentration, pH, and Apase.
In addition, we conducted a cluster analysis using P-
mobilization traits to group species into efficient P-mobilizing
species and species that are inefficient at P mobilization in mono-
culture in P-deficient soils and P-sufficient soils, following
Ward’s method (Borcard et al., 2018). We partitioned species
into two groups according to the average silhouette width (Bor-
card et al., 2018). The package ‘CLUSTER’ was used for cluster ana-
lysis (Maechler et al., 2022).
Mixture Linear mixed-effect models were performed to test the
effect of P addition and group (i.e. mixtures that included effi-
cient P-mobilizing species and mixtures without efficient P-
mobilizing species) on biomass, P content, relative diversity effect
(i.e. relative selection effect and complementarity effect), and
relative biomass. Phosphorus addition and group were treated as
fixed effects, block and combination identity were treated as ran-
dom effects.
Two-way ANOVAs were used to examine the effects of P addi-
tion and neighbor identity (four different species that grew next
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
4
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

to the focal species) on relative biomass of each species. The effect
of group on leaf [Mn], leaf [P] and relative changes of trait from
monoculture on low-P soils, and the differences in relative com-
plementarity effect among 10 species combinations on P-
deficient soils were tested, using one-way ANOVAs.
Two-sided t-tests were conducted to examine whether the
relative selection and complementarity effect significantly dif-
fered from zero, that is, expected performance (Loreau & Hec-
tor, 2001). If the relative selection effect was significantly
greater than zero, this indicates that the contribution to
enhanced performance of mixtures was due to increased biomass
in mixtures by the species that also had the highest biomass in
monoculture. If the relative complementarity effect was positive,
this reflects interspecific facilitation and niche differentiation.
We also used two-sided t-tests to test whether the observed rela-
tive biomass of a given species significantly differed from 0.5,
that is, expected relative biomass. When the observed relative
biomass was >0.5, the given species was overperforming in
mixtures.
Correlations between relative changes of leaf [Mn] and leaf
[P], between relative changes of leaf [Mn] of neighbor species
that are inefficient at P mobilization and the dominance of effi-
cient P-mobilizing species, and between relative complementarity
effect and relative changes of trait of species that are inefficient at
P mobilization were performed. To test the differences in relative
changes of root trait and leaf [Mn] of species that are inefficient
at P mobilization in response to neighbor identity, we also carried
out a principal component analysis using five root morphological
traits and leaf [Mn].
Meta-analysis Literature including the difference in leaf [Mn]
in monoculture and mixture was collected through Web of
Science and China National Knowledge Infrastructure. We
further extracted leaf [P] data in the selected literature. A
weighted response ratio approach was used to conduct the meta-
analysis (Hedges et al., 1999; Luo et al., 2006). For full details of
data compilation and statistical analyses, see Methods S2. Litera-
ture included in this meta-analysis is given in Dataset S1.
Results
Species vary in root traits to mobilize P in monoculture
Phosphorus addition increased the aboveground biomass of all
tested species, except for C. korshinskyi, but only increased the
belowground biomass of S. grandis (Table 1). Phosphorus addi-
tion increased aboveground [P] and P content of most species
(Table S5). Species varied in root traits that relate to mobilization
of sorbed P, while P addition did not affect root traits. Carex kor-
shinskyi exhibited greater rhizosheath carboxylate concentration
and lower rhizosheath pH than other species did (Figs 1a,S1a).
Species varied significantly in their capacity to mobilize sorbed
soil P, and C. korshinskyi exhibited significantly greater
P-mobilization traits than other species did (PERMANOVA test
P<0.01, Fig. 1c; Table S6). We further conducted a cluster ana-
lysis using these P-related traits to classify the five species into
efficient P-mobilizing species and species that are inefficient at P
mobilization. The results showed that C. korshinskyi was an effi-
cient P-mobilizing species on both P-deficient and P-sufficient
soils (Fig. 1d). Others were considered species that are inefficient
at P mobilization.
Artemisia frigida and C. korshinskyi also exhibited a relatively
greater leaf [Mn] than other species did (Fig. 1b), and leaf [Mn]
was positively correlated with main and total carboxylate concen-
tration, negatively with rhizosheath pH, and did not correlate
with Apase in both P-deficient and P-enriched soils (Figs 1e,f,
S2,S3).
Combinations that included C. korshinskyi exhibited a
greater relative complementarity effect
To explore whether mixtures differed in productivity and diver-
sity effects, we partitioned 10 species combinations into combi-
nations that involved an efficient P-mobilizing species (i.e. C.
korshinskyi in this study) and combinations that did not include
an efficient P-mobilizing species based on the cluster analysis.
Combinations with C. korshinskyi exhibited a greater above- and
belowground biomass but not aboveground P content than
monocultures and mixtures without an efficient P-mobilizing
species in low-P soils (Tables 1,S5).
Relative selection effect was not affected by P addition and
group (i.e. mixtures that included C. korshinskyi and mixtures
without C. korshinskyi) (Fig. 2a,d). Aboveground relative comple-
mentarity effects were significantly positive and greater in combi-
nations including an efficient P-mobilizing species than those in
Table 1 Above- and belowground biomass in monoculture and mixture at
two phosphorus (P) levels.
Biomass (g) Species/group
P level
LP HP
Aboveground biomass in
monoculture
Sg (n=4) 0.84b 1.47a
Lc (n=4) 3.43b 4.26a
Ck (n=4) 1.87a 2.02a
Cs (n=4) 5.21b 7.07a
Af (n=4) 3.15b 5.57a
Belowground biomass in
monoculture
Sg (n=4) 0.25b 0.43a
Lc (n=4) 1.06a 1.66a
Ck (n=4) 0.41a 0.58a
Cs (n=4) 0.61a 0.71a
Af (n=4) 0.45a 0.65a
Aboveground biomass in
monoculture and
mixture
Monoculture (n=20) 2.90bB 4.08aA
With (n=16) 4.26aA 4.09aA
Without (n=24) 3.86bAB 5.10aA
Belowground biomass in
monoculture and
mixture
Monoculture (n=20) 0.56bB 0.81aA
With (n=16) 1.18aA 1.16aA
Without (n=24) 0.80bAB 1.19aA
nis the number of data in the analysis. Lowercase letters indicate
differences between P levels, uppercase letters represent differences
among monoculture, mixtures that include (with) and do not include an
efficient P-mobilizing species (without) at a given P level. The same letter
means no significant difference (Tukey HSD). Af, Artemisia frigida;
Ck, Carex korshinskyi; Cs, Cleistogenes squarrosa; HP, P-sufficient soil;
Lc, Leymus chinensis; LP, P-deficient soil; Sg, Stipa grandis.
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023)
www.newphytologist.com
New
Phytologist Research 5
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

the absence of C. korshinskyi in P-deficient soils, but not in high-
P soils (Fig. 2b). Furthermore, species that are inefficient at P
mobilization usually exhibited greater relative aboveground bio-
mass when grown with C. korshinskyi than when grown next to
another inefficient P-mobilizing species, particularly on low-P
soils (Fig. 2c; Table 2). Our evidence also shows that combina-
tions that included C. korshinskyi exhibited a greater below-
ground complementarity effect and relative biomass of species
that are inefficient at P mobilization than those that did not
include C. korshinskyi, independent of P-addition treatment
(Fig. 2e,f).
Leaf [Mn] can be used as a proxy for interspecific P
facilitation via carboxylates
We aimed to examine whether leaf [Mn] can be used to detect
interspecific facilitation of P acquisition. We compared leaf [Mn]
and [P] in monocultures and mixtures with or without C. kor-
shinskyi. We show that species that are inefficient at P mobiliza-
tion that grew with C. korshinskyi exhibited a significantly greater
leaf [Mn] and [P] than those in monoculture or in mixtures with-
out an efficient P-mobilizing neighbor in P-deficient soils, but
not in high-P soils (Fig. 3a,b). The changes in leaf [Mn] were
Fig. 1 Variations in phosphorus (P)-
mobilization traits among species in
monoculture. Effects of P addition on (a)
rhizosheath carboxylate concentration and
(b) leaf manganese concentration ([Mn])
among species (S) in monoculture. The
yellow square indicates the mean value, the
central line represents the median, the
bottom and top of the box indicate the 25
th
and 75
th
percentiles, respectively. Whiskers
are the smallest and largest value within the
1.5 times interquartile range below and
above the 25
th
and 75
th
percentiles. n=4.
Lowercase letters indicate differences
between treatments if the interaction effect
was significant. Uppercase letters mean
differences among species. The same letter
means there was no significant difference
(Tukey HSD). (c) Principal component
analysis (PCA) for root traits associated with
P mobilization in monoculture of five species
in monoculture. PC1 represents the first axis,
PC2 represents the second axis, and the
percentage number represents the
proportion of variation the axis explained.
A PERMANOVA test showed that species
exhibited significant differences based on P-
mobilization traits (P<0.01). (d) Cluster
analysis across species monocultures based
on root physiological traits, that is,
rhizosheath acid phosphatase activity
(Apase), pH, and total carboxylate
concentration using Ward’s method. Species
shown in blue and red boxes represent
inefficient and efficient P-mobilizing species,
respectively. (e, f) Correlations between total
carboxylate concentration and pH in
rhizosheath soil and leaf [Mn] in species
monoculture. Light grey bands represent
95% confidence intervals. Af, Artemisia
frigida; Ck, Carex korshinskyi; Cs,
Cleistogenes squarrosa; HP, P-sufficient soil;
Lc, Leymus chinensis; LP, P-deficient soil;
Sg, Stipa grandis.*,P<0.05; **,P<0.01;
***,P<0.001; ns, not significant.
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
6
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

positively correlated with that of leaf [P] (Fig. 3c), indicating soil
P mobilized by C. korshinskyi facilitated P uptake of neighboring
inefficient species, implying the existence of interspecific P facili-
tation via carboxylate release. We further showed that changes in
leaf [Mn] of inefficient species were positively correlated with the
above- and belowground dominance of C. korshinskyi in mixtures
(Fig. 3e,f).
To examine whether the observed results in our glasshouse
experiment were similar to those reported in the literature, we
performed a meta-analysis testing the effect of species mixtures
on leaf [Mn] and [P]. Species included in this meta-analysis
were grouped into efficient P-mobilizing species and
inefficient ones based on references (Table S7). No publica-
tion bias of leaf [Mn] and [P] was found in this meta-analysis
(Fig. S4).
Our analysis showed that compared with monoculture,
grown with efficient P-mobilizing species, leaf [Mn] and [P] of
inefficient species were 9% and 28% greater without P addition,
and 16% and 7% under P addition, respectively (Fig. 3d).
However, compared with monocultures, the leaf [Mn] and [P]
did not differ when two species that are inefficient at P mobili-
zation were grown together, supporting the results of the pre-
sent study.
Interspecific P facilitation enhanced complementarity and
required greater changes in traits of facilitated species in
mixtures than monocultures
To explore the role of root traits of facilitated species in interspe-
cific P facilitation via carboxylates (proxied by leaf [Mn]), we
further calculated the change in traits of species that are ineffi-
cient at P mobilization relative to that in corresponding mono-
cultures. The results showed that species that are inefficient at P
mobilization exhibited greater relative changes of leaf [Mn], pro-
portion of fine roots, root branching intensity, and root hair
length when growing with C. korshinskyi than when grown with
another inefficient P-mobilizing species on P-deficient soils
(Fig. 4a). When grown with C. korshinskyi,S. grandis exhibited
greater relative change of root morphological traits than other
species did (Fig. 4b). Artemisia frigida exhibited a greater relative
change of leaf [Mn] when grown with C. korshinskyi, while the
relative change of root morphological traits was more negative in
mixtures than in monocultures (Fig. 4b).
The changes in leaf [Mn] of S. grandis,L. chinensis and A. fri-
gida were positively associated with the relative complementarity
effect (Fig. 4c). By contrast, the relative change of leaf [P] did not
show a significant correlation with the relative complementarity
Fig. 2 Aboveground and belowground
relative diversity effects and relative biomass
in mixtures. Effects of phosphorus addition
(P) and group (G) on relative selection effect,
relative complementarity effect, and relative
biomass of (a–c) aboveground biomass and
(d–f) belowground biomass. The group
includes mixture with Carex korshinskyi
(With, n=16) and without C. korshinskyi
(Without, n=24). The yellow square
indicates the mean value, the central line
represents the median, the bottom and top
of the box indicate the 25
th
and 75
th
percentiles, respectively. Whiskers are the
smallest and largest value within the 1.5
times interquartile range below and above
the 25
th
and 75
th
percentiles. Lowercase
letters indicate differences across treatments
if the interaction effect was significant. The
same letter means there was no significant
difference (Tukey HSD). No lowercase letters
were used if the interaction effect was not
significant. HP, P-sufficient soil; LP, P-
deficient soil. # in bars indicate a significant
difference between relative selection effect
and complementarity effect from zero, and
between relative biomass from 0.5. #, P<
0.05; ##, P<0.01; ###, P<0.001. Asterisks
indicate a significant difference in P addition,
group, and their interaction effects. *,P<
0.05; **,P<0.01; ***,P<0.001; ns, not
significant.
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023)
www.newphytologist.com
New
Phytologist Research 7
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

effect, except for S. grandis of aboveground complementarity
(Fig. S5). Relative changes of leaf [Mn] of S. grandis also exhib-
ited a positive correlation with changes in proportion of fine
roots and root branching intensity when grown with different
species (Fig. 4c). However, the relative change of root traits from
the monoculture of other species did not exhibit a positive corre-
lation with that of leaf [Mn], except for proportion of fine roots
and root hair length of A. frigida (Figs 4c,S6).
In addition, C. korshinskyi and S. grandis exhibited greater rela-
tive biomass in a S. grandis/C. korshinskyi combination (Table 2),
therefore exhibiting a greater above- and belowground relative
complementarity effect than other combinations on P-deficient
soils (Fig. 4d).
Discussion
Plant growth was limited by P in this study (Table 1). Carex kor-
shinskyi was grouped as an efficient P-mobilizing species, with
greater carboxylate and proton release, had similar biomass on
low-P and high-P soils, due to its greater capacity to mobilize
sorbed P than other species (Fig. 1c,d). Most biodiversity
research focuses on the effects of species richness on aboveground
productivity, while belowground overperformance and comple-
mentarity effects have received far less attention, probably due to
method limitations. By calculating the root biomass of each spe-
cies via FTIR spectroscopy, we quantified belowground comple-
mentarity effect in mixtures. We found that combinations
involving C. korshinskyi usually exhibited greater above- and
belowground biomass and relative complementarity effects than
those in the absence of C. korshinskyi on P-deficient soils (Table 1;
Fig. 2), which supports the first hypothesis. Recent studies also
show that plant mixtures enhance productivity and complemen-
tarity effects via increased soil P availability in diverse ecosystems,
in which interspecific P facilitation is a key mechanism to drive
overperformance in species mixtures (Yu et al., 2020a; Chen
et al., 2022). We found that aboveground relative biomass of
S. grandis,L. chinensis, and C. squarrosa was near or >1 when
neighbored with C. korshinskyi on P-deficient soils (Table 2),
which probably reflects interspecific facilitation (Wagg
et al., 2019). We further explore the potential proxies to study
mechanisms of facilitation in species-diverse communities.
We found that leaf [Mn] was positively correlated with total
and main carboxylate concentrations in the rhizosheath on P-
deficient and P-sufficient soils (Figs 1e,S2b). Oxalate, fumarate,
and malonate accounted for 96% of total carboxylates (Fig. S2a),
where oxalate and malonate have a similar efficiency in mobiliz-
ing sorbed P to citrate (Str¨
om et al., 2005; Playsted et al., 2006;
Pang et al., 2018). Then, we showed that greater leaf [Mn] and
[P] of species that are inefficient at P mobilization were observed
only in the presence of C. korshinskyi, an efficient P-mobilizing
species, on P-deficient soils (Fig. 3a,b). The mechanism under-
pinning this species-specific interaction presumably is the chela-
tion of Mn by carboxylates and the decrease in rhizosheath pH
by efficient P-mobilizing species, as Mn reduction in the soil is
enhanced at low pH (Gardner et al., 1982). The decrease in rhi-
zosheath pH by efficient P-mobilizing species also enhances soil
P availability through the dissolution of P-containing minerals,
facilitating P uptake of neighboring species that are inefficient at
Table 2 Above- and belowground relative biomass of each species when grown with different neighbors in phosphorus (P)-deficient soils (LP) and P-
sufficient soils (HP).
Species Neighbor
Aboveground Belowground
LP HP Variable Pvalue LP HP Variable Pvalue
Sg Lc 0.61 0.13 P <0.001 1.27 0.53 P <0.05
Ck 0.95 0.41 N <0.001 0.77 0.45 N <0.01
Cs 0.21 0.06 P ×N 0.14 0.33 0.15 P ×N 0.24
Af 0.59 0.45 0.78 0.77
Lc Sg 0.77 0.87 P 0.20 0.53 0.74 P 0.37
Ck 1.39 1.02 N <0.001 1.88 1.05 N <0.05
Cs 0.74 0.63 P ×N 0.27 0.89 0.71 P ×N 0.20
Af 0.63 0.58 1.00 1.14
Ck Sg 1.12 1.10 P 0.56 2.56 1.99 P 0.38
Lc 0.30 0.32 N <0.001 0.92 0.85 N <0.01
Cs 0.36 0.22 P ×N 0.94 1.00 0.40 P ×N 0.79
Af 0.75 0.66 1.21 1.35
Cs Sg 0.85 1.02 P 0.54 0.72 0.88 P 0.72
Lc 0.73 0.59 N <0.01 0.80 0.80 N 0.71
Ck 1.04 0.77 P ×N<0.05 0.80 0.78 P ×N 0.65
Af 0.82 0.93 0.91 0.85
Af Sg 0.26 0.35 P <0.05 0.27 0.51 P 0.53
Lc 0.37 0.21 N <0.001 0.72 0.45 N <0.01
Ck 0.63 0.35 P ×N<0.05 0.69 0.57 P ×N 0.21
Cs 0.15 0.06 0.27 0.21
n=4. Neighbor identity effect refers to the effects of four different species grown next to the focal species on above- and belowground relative biomass of
each species. Af, Artemisia frigida; Ck, Carex korshinskyi; Cs, Cleistogenes squarrosa; Lc, Leymus chinensis; N, neighbor identity effect; P, P addition
effect; Sg, Stipa grandis.
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
8
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

P mobilization (Li et al., 2007,2014). The changes in leaf [Mn]
were positively correlated with leaf [P], implying the neighboring
inefficient species took up the Mn
2+
and soluble P mobilized by
the facilitator (Lambers et al., 2015). In support of this, our
meta-analysis also shows that greater leaf [Mn] and [P] of species
only occurred when grown with an efficient P-mobilizing species,
rather than with another inefficient species in both agroecosys-
tems and natural systems, especially without P addition (Fig. 3d).
Leaf [P] of species increased under P addition, indicating the
amelioration of P status, while leaf [Mn] did not change in HP
treatment (Fig. 3a,b). Conversely, leaf [Mn] of facilitated species
increased when grown next to efficient P-mobilizing species in
both the glasshouse experiment and a meta-analysis. The changes
in leaf [P] are affected by multiple factors, for example, plant P-
acquisition strategy, neighbor identity, and soil P availability
which do not specifically reflect interspecific P facilitation (Lam-
bers, 2022). Increased leaf [Mn] of facilitated species is likely
related to the enhanced soil P availability as a result of rhizo-
sphere processes of the facilitator (Lambers et al., 2015; Lambers,
2022). Therefore, we highlight that leaf [Mn], rather than leaf
[P], is a good proxy to reflect interspecific P facilitation.
We only found a single efficient P-mobilizing species, that is,
C. korshinskyi, in the present study. However, mycorrhizal species
Melissilus ruthenicus,P. tanacetifolia and Filifolium sibiricum in
our previous research (Yu et al., 2020a), and non-mycorrhizal
species such as Lupinus albus (Dissanayaka et al., 2015) and
Fig. 3 Leaf manganese concentration ([Mn]),
leaf phosphorus concentration ([P]), and
their relative change between mixtures and
monocultures in the glasshouse experiment
and meta-analysis. Effects of group (G) on
(a) leaf [Mn] and (b) leaf [P] of species that
are inefficient at P mobilization. The group
includes monoculture (n=16), neighbored
with Carex korshinskyi (With, n=16) and
grown without C. korshinskyi (Without, n=
48). The yellow square indicates the mean
value, the central line represents the median,
the bottom and top of the box indicate the
25
th
and 75
th
percentiles, respectively.
Whiskers are the smallest and largest value
within the 1.5 times interquartile range
below and above the 25
th
and 75
th
percentiles. Lowercase letters indicate
differences among the group on P-deficient
soils (LP) and P-sufficient soils (HP). The
same letter means there was no significant
difference (Tukey HSD). (c) Correlations
between the relative change of leaf [P] and
that of leaf [Mn] of species that are
inefficient at P mobilization in combination
with and without C. korshinskyi at two soil
P-levels. (d) The weighted response ratio
(RR
++
) of leaf [Mn] and leaf [P] of inefficient
grown with different neighbors (N) in the
meta-analysis. The list of inefficient and
efficient P-mobilizing species is shown in
Supporting Information Table S7. Bars
represent 95% confidence intervals of RR
++
.
The numbers outside and inside the brackets
indicate the sample size of leaf [Mn] and leaf
[P] data, respectively. (e, f) Correlations
between relative change of leaf [Mn] of
neighboring species from corresponding
monocultures and the dominance of C.
korshinskyi on low-P soils. Panels (a–c, e, f)
showed the results at present glasshouse
experiment, panel (d) showed the results
from the meta-analysis. Light grey bands
represent 95% confidence intervals. *,P<
0.05; **,P<0.01; ***,P<0.001; ns, not
significant.
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023)
www.newphytologist.com
New
Phytologist Research 9
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Banksia attenuata (Muler et al., 2014) in other studies are also
considered efficient P-mobilizing species. Therefore, we surmise
that species that are inefficient at P mobilization might exhibit
greater productivity when grown with an efficient P-mobilizing
species by P facilitation on low-P soils (Li et al., 2014). This
implies that the effects of interspecific P facilitation associated
with a positive BEF relationship by facilitators are common,
rather than species specific. Therefore, the results from our glass-
house experiment and the meta-analysis provide further support
that leaf [Mn] can be used as a proxy of interspecific P facilitation
via carboxylate release in various ecosystems, supporting the sec-
ond hypothesis.
Fig. 4 The role of relative change of leaf
manganese concentration ([Mn]) and root
morphological traits between mixtures and
monocultures of species that are inefficient at
phosphorus (P) mobilization in
complementarity. (a) Relative change of
functional traits of species that are inefficient
at P mobilization in response to the group
includes a mixture with Carex korshinskyi
(With, red polygon, n=16) and without C.
korshinskyi (Without, blue polygon, n=48)
on low-P soils. Asterisks indicate a significant
difference between groups. (b) Principal
component analysis (PCA) for relative
change of traits of species grown with or
without C. korshinskyi on P-deficient soils.
PC1 represents the first axis, PC2 represents
the second axis, and the percentage number
represents the proportion of variation the
axis explained. (c) Correlations between
relative complementarity effect and the
relative change of leaf manganese
concentration ([Mn]) of each species that are
inefficient at P mobilization, and between the
relative change of proportion of fine roots
(PFR) and root branching intensity (RBI) and
that of leaf [Mn] of each species that is
inefficient at P mobilization grown with one
of four species (n=16) on low-P soils. Light
grey bands represent 95% confidence
intervals. (d) Relative complementarity effect
of above- and belowground biomass across
combinations on P-deficient soils (n=4). Bars
are means SE. # indicates a significant
difference between the relative
complementarity effect from zero.
#, P<0.05; ##, P<0.01; ###, P<0.001.
Lowercase letters indicate differences among
combinations. The same letter means that
there was no significant difference (Tukey
HSD). Af, Artemisia frigida; Ck, Carex
korshinskyi; Cs, Cleistogenes squarrosa; Lc,
Leymus chinensis; RHD, root hair density;
RHL, root hair length; Sg, Stipa grandis; SRL,
specific root length. *,P<0.05; **,P<0.01;
***,P<0.001.
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
10
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Grown with C. korshinskyi, the neighboring species usually
exhibited greater relative changes of a range of traits (i.e. leaf
[Mn], proportion of fine roots, root branching intensity, and root
hair length) than when grown with another species that was inef-
ficient at P mobilization on low-P soils (Fig. 4a). We further
showed that the roots of grass species were thinner than those of
A. frigida when grown with C. korshinskyi (Fig. 4b); S. grandis
exhibited greater changes of morphological traits than other spe-
cies did, and interspecific P facilitation (proxied by leaf [Mn])
enhanced relative complementarity effect via changes of root
morphological traits (Fig. 4b,c). In the presence of C. korshinskyi,
the S. grandis/C. korshinskyi combination also exhibited a greater
relative complementarity effect than other species combinations
(Fig. 4d). This is probably because the root growth of facilitated
species can change in response to increased P availability in the
rhizosheath soil of the facilitator (Zhang et al., 2016,2020). Our
previous results showed that grass species exhibit greater inherent
root morphological traits and relative changes of those traits than
forbs do; this may allow grasses to obtain greater benefits from
efficient P-mobilizing neighbors than forbs do (Yu et al., 2020a).
This is probably also related to contrasting P-acquisition strate-
gies and trade-offs among root traits between grasses and forbs
(Wen et al., 2019). The morphological traits examined in this
study are important for P acquisition (Lambers et al., 2006),
while our evidence shows that specific root length, root hair
length and density did not exhibit a positive correlation with leaf
[Mn] of the tested species, indicating that the changes in the
appropriate trait (e.g. the proportion of fine roots in the present
study) may be required in interspecific P facilitation. In addition,
C. korshinskyi exhibited greater relative biomass when grown with
S. grandis which may be related to the lower competition by
S. grandis; and the strength of interspecific P facilitation may
increase with enhanced dominance of C. korshinskyi (Table 2;
Fig. 3e,f). Therefore, S. grandis may obtain more benefit via both
increased dominance of the facilitator to mobilize more sorbed P
into the soil solution, and S. grandis exhibits greater relative
changes of a range of traits in response to the facilitator, support-
ing our third hypothesis.
By contrast, although the relative change of leaf [Mn] of A. fri-
gida was positively correlated with the relative complementarity
effect, the inherently coarse roots and negative response of root
morphological traits to C. korshinskyi may offset the benefits via
carboxylates, resulting in a lower relative complementarity effect
than that in other mixtures that include C. korshinskyi (Fig. 4b,
d). The L. chinensis/C. korshinskyi combination exhibited a rela-
tively greater complementarity than that of other mixtures, and
enhanced interspecific P facilitation may increase belowground
complementarity in the mixture (Fig. 4c,d). However, we did not
observe a positive correlation between root morphological traits
of L. chinensis and interspecific P facilitation (Figs 4c,S6). Com-
plementarity effects may be driven by multiple mechanisms, and
as such, variations in another mechanism may obscure clear cor-
relations between complementarity effects and one particular
mechanism (e.g. P facilitation) in the L. chinensis/C. korshinskyi
combination. Leymus chinensis has greater stoichiometric stability
than other species in a typical steppe, that is, a greater ability to
maintain plant nutrient status despite variations in soil nutrient
availability (Yu et al., 2010). We surmise that the inherently thin
roots and the rhizome of L. chinensis facilitate the growth and
limited response to soil nutrient availability via changes in root
traits. Although greater relative biomass of some species that are
inefficient at P mobilization is observed when neighbored with
C. korshinskyi on high-P soils (Table 2), the results likely reflect
stronger competition of focal species than C. korshinskyi, rather
than interspecific P facilitation (Figs 2b,3a,b).
To our knowledge, this is the first time to comprehensively use
leaf [Mn] as a proxy to examine interspecific P facilitation via car-
boxylates in biodiversity research. Importantly, using leaf [Mn]
as proxy, we highlight that direct interspecific facilitation of P
acquisition via belowground processes is a vital mechanism that
can enhance complementarity and productivity. We can examine
P facilitation by measuring a straightforward leaf trait rather than
collecting complicated root traits in species-diverse plant commu-
nities. What is less clear is how much of the biodiversity effects in
bicultures containing C. korshinskyi is due to facilitation as
opposed to niche differentiation. This question may be addressed
directly by including single plant controls, as, for example, in
Sch¨
ob et al.(2018). Future research would also consider more
species combinations with divergent P strategies to explore the
general role of interspecific P facilitation in BEF relationships.
Leaf [Mn] exhibits a positive correlation with rhizosheath car-
boxylates, whereas soil [Mn] is intercepted by arbuscular mycor-
rhizal fungi (AMF), and plant Mn uptake is affected by the soil
microbiome (Kothari et al., 1991; Lehmann & Rillig, 2015), soil
Mn availability and plant strategy (Tian et al., 2016; Lambers
et al., 2021). The roots of species that were included in this study
are colonized by AMF except for C. korshinskyi; the colonization
rates range from 16% to 39% in tested species (Tian
et al., 2009). Therefore, further studies should explore the role of
the AMF, soil microbiome and Mn-acquisition strategies in
interspecific P facilitation in species-diverse communities.
In conclusion, we show that leaf [Mn] is a proxy for interspeci-
fic P facilitation via rhizosheath carboxylates in grassland
communities, as well as in shrub and crop species through a
meta-analysis. There is a potential application using leaf [Mn] as
a trait to examine P facilitation in other ecosystems such as forests
and intercrops. Furthermore, combinations that include an effi-
cient P-mobilizing species tend to exhibit a greater relative above-
and belowground complementarity effect on P-impoverished
soils which is related to direct P facilitation via carboxylate
release. Neighboring species that exhibit a greater response of
specific traits obtain greater benefits from the facilitator, in which
S. grandis may be better matched with C. korshinskyi. This is the
first report demonstrating how interspecific P facilitation via
carboxylate release increases above- and belowground comple-
mentarity, enhancing our understanding of BEF relationship and
species coexistence.
Acknowledgements
This study was financially supported by the National Natural
Science Foundation of China (Project no. 32101297). R-PY was
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023)
www.newphytologist.com
New
Phytologist Research 11
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

also financially supported by the fellowship of China Postdoc-
toral Science Foundation (Project no. 2021M690164). We thank
Dr Jun Wang for help with the measurement of carboxylates.
Competing interests
None declared.
Author contributions
LL and R-PY designed the experiment. R-PY, YS, RA, HY, X-
TY, and YX collected data and performed analyses. HL provided
the idea to focus on leaf [Mn] as a proxy to examine interspecific
phosphorus facilitation via carboxylate release in mixtures. R-PY
drafted the paper, and HL, JR, W-PZ, and LL contributed sub-
stantially to revisions.
ORCID
Ran An https://orcid.org/0000-0002-4714-8454
Hans Lambers https://orcid.org/0000-0002-4118-2272
Long Li https://orcid.org/0000-0003-0523-3308
Jasper van Ruijven https://orcid.org/0000-0003-0003-2363
Ye Su https://orcid.org/0000-0001-8682-3363
Yi Xing https://orcid.org/0000-0001-6558-5466
Hao Yang https://orcid.org/0000-0003-2138-7307
Xiao-Tong Yin https://orcid.org/0000-0003-2108-4621
Rui-Peng Yu https://orcid.org/0000-0001-9385-9444
Wei-Ping Zhang https://orcid.org/0000-0002-7769-1255
Data availability
Data are available in Zenodo at doi: 10.5281/zenodo.7919182.
References
Barry KE, Mommer L, van Ruijven J, Wirth C, Wright AJ, Bai Y, Connolly J,
De Deyn GB, de Kroon H, Isbell F et al. 2019a. The future of
complementarity: disentangling causes from consequences. Trends in Ecology &
Evolution 34: 167–180.
Barry KE, van Ruijven J, Mommer L, Bai Y, Beierkuhnlein C, Buchmann N, de
Kroon H, Ebeling A, Eisenhauer N, Guimaraes-Steinicke C et al. 2019b.
Limited evidence for spatial resource partitioning across temperate grassland
biodiversity experiments. Ecology 101: e02905.
Bergmann J, Ryo M, Prati D, Hempel S, Rillig MC. 2017. Root traits are more
than analogues of leaf traits: the case for diaspore mass. New Phytologist 216:
1130–1139.
Borcard D, Gillet F, Legendre P. 2018. Numerical ecology with R. New York, NY,
USA: Springer.
Brooker RW, George TS, Homulle Z, Karley AJ, Newton AC, Pakeman RJ,
Sch¨
ob C, Wright A. 2021. Facilitation and biodiversity–ecosystem function
relationships in crop production systems and their role in sustainable farming.
Journal of Ecology 109: 2054–2067.
Cahill JF Jr, McNickle GG, Haag JJ, Lamb EG, Nyanumba SM, St Clair CC.
2010. Plants integrate information about nutrients and neighbors. Science 328:
1657.
Cardinale BJ, Wright JP, Cadotte MW, Carroll IT, Hector A, Srivastava DS,
Loreau M, Weis JJ. 2007. Impacts of plant diversity on biomass production
increase through time because of species complementarity. Proceedings of the
National Academy of Sciences, USA 104: 18123–18128.
Chen X, Chen HYH, Chang SX. 2022. Meta-analysis shows that plant mixtures
increase soil phosphorus availability and plant productivity in diverse
ecosystems. Nature Ecology & Evolution 6: 1112–1121.
Craven D, Isbell F, Manning P, Connolly J, Bruelheide H, Ebeling A, Roscher
C, van Ruijven J, Weigelt A, Wilsey B et al. 2016. Plant diversity effects on
grassland productivity are robust to both nutrient enrichment and drought.
Philosophical Transactions of the Royal Society of London. Series B: Biological
Sciences 371: 20150277.
Dissanayaka DMS, Maruyama H, Masuda G, Wasaki J. 2015. Interspecific
facilitation of P acquisition in intercropping of maize with white lupin in two
contrasting soils as influenced by different rates and forms of P supply. Plant
and Soil 390: 223–236.
Fiori J, Amadesi E, Fanelli F, Tropeano CV, Rugolo M, Gotti R. 2018. Cellular
and mitochondrial determination of low molecular mass organic acids by LC-
MS/MS. Journal of Pharmaceutical and Biomedical Analysis 150:33–38.
Gardner W, Boundy K. 1983. The acquisition of phosphorus by Lupinus albus
L., IV. The effect of interplanting wheat and white lupin on the growth and
mineral composition of the two species. Plant and Soil 70: 391–402.
Gardner W, Parbery D, Barber D. 1982. The acquisition of phosphorus by
Lupinus albus L. I. Some characteristics of the soil/root interface. Plant and Soil
68:19–32.
Haling RE, Yang Z, Shadwell N, Culvenor RA, Stefanski A, Ryan MH, Sandral
GA, Kidd DR, Lambers H, Simpson RJ. 2016. Root morphological traits that
determine phosphorus-acquisition efficiency and critical external phosphorus
requirement in pasture species. Functional Plant Biology 43: 815–826.
Hedges LV, Gurevitch J, Curtis PS. 1999. The meta-analysis of response ratios
in experimental ecology. Ecology 80: 1150–1156.
Horst WJ, Kamh M, Jibrin JM, Chude VO. 2001. Agronomic measures for
increasing P availability to crops. Plant and Soil 237: 211–223.
Huang G, Hayes PE, Ryan MH, Pang J, Lambers H. 2017. Peppermint trees
shift their phosphorus-acquisition strategy along a strong gradient of plant-
available phosphorus by increasing their transpiration. Oecologia 185: 387–
400.
Jesch A, Barry KE, Ravenek JM, Bachmann D, Strecker T, Weigelt A, Buchmann
N, De Kroon H, Gessler A, Mommer L. 2018. Below-ground resource
partitioning alone cannot explain the biodiversity-ecosystem function relationship:
a field test using multiple tracers. Journal of Ecology 106:2002–2018.
Karanika ED, Alifragis DA, Mamolos AP, Veresoglou DS. 2007. Differentiation
between responses of primary productivity and phosphorus exploitation to
species richness. Plant and Soil 297:69–81.
Kothari SK, Marschner H, Romheld V. 1991. Effect of a vesicular-arbuscular
mycorrhizal fungus and rhizosphere micro-organisms on manganese reduction
in the rhizosphere and manganese concentrations in maize (Zea mays L.). New
Phytologist 117: 649–655.
Kramer-Walter KR, Bellingham PJ, Millar TR, Smissen RD, Richardson SJ,
Laughlin DC, Mommer L. 2016. Root traits are multidimensional: specific
root length is independent from root tissue density and the plant economic
spectrum. Journal of Ecology 104: 1299–1310.
Lambers H. 2022. Phosphorus acquisition and utilization in plants. Annual
Review of Plant Biology 73:17–42.
Lambers H, de Britto Costa P, Cawthray GR, Denton MD, Finnegan PM,
Hayes PE, Oliveira RS, Power SC, Ranathunge K, Shen Q et al. 2022.
Strategies to acquire and use phosphorus in phosphorus-impoverished and fire-
prone environments. Plant and Soil 476: 133–160.
Lambers H, Hayes PE, Lalibert´
e E, Oliveira RS, Turner BL. 2015. Leaf
manganese accumulation and phosphorus-acquisition efficiency. Trends in
Plant Science 20:83–90.
Lambers H, Shane MW, Cramer MD, Pearse SJ, Veneklaas EJ. 2006. Root
structure and functioning for efficient acquisition of phosphorus: matching
morphological and physiological traits. Annals of Botany 98: 693–713.
Lambers H, Wright IJ, Guilherme Pereira C, Bellingham PJ, Bentley LP,
Boonman A, Cernusak LA, Foulds W, Gleason SM, Gray EF et al. 2021. Leaf
manganese concentrations as a tool to assess belowground plant functioning in
phosphorus-impoverished environments. Plant and Soil 461:43–61.
Lehmann A, Rillig MC. 2015. Arbuscular mycorrhizal contribution to copper,
manganese and iron nutrient concentrations in crops –a meta-analysis. Soil
Biology and Biochemistry 81: 147–158.
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
12
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Li L, Li S, Sun J, Zhou L, Bao X, Zhang H, Zhang F. 2007. Diversity enhances
agricultural productivity via rhizosphere phosphorus facilitation on
phosphorus-deficient soils. Proceedings of the National Academy of Sciences, USA
104: 11192–11196.
Li L, Tilman D, Lambers H, Zhang F. 2014. Plant diversity and overyielding:
insights from belowground facilitation of intercropping in agriculture. New
Phytologist 203:63–69.
Loreau M, Hector A. 2001. Partitioning selection and complementarity in
biodiversity experiments. Nature 412:72–76.
Luo Y, Hui D, Zhang D. 2006. Elevated CO
2
stimulates net accumulations
of carbon and nitrogen in land ecosystems: a meta-analysis. Ecology 87:
53–63.
Maechler M, Rousseeuw P, Struyf A, Hubert M, Hornik K. 2022. CLUSTER:
cluster analysis basics and extensions. R package v.2.1.4. [WWW document]
URL https://cran.r-project.org/web/packages/cluster/index.html [accessed 15
June 2023].
Muler AL, Oliveira RS, Lambers H, Veneklaas EJ. 2014. Does cluster-root
activity benefit nutrient uptake and growth of co-existing species? Oecologia
174:23–31.
Oksanen J, Simpson GL, Blanchet FG, Kindt R, Legendre P, Minchin PR,
O’Hara RB, Solymos P, Stevens MHH, Szoecs E et al. 2022. VEGAN:
community ecology package. R package v.2.6-4. [WWW document] URL
https://github.com/vegandevs/vegan [accessed 15 June 2023].
Pang J, Bansal R, Zhao H, Bohuon E, Lambers H, Ryan MH, Ranathunge
K, Siddique KH. 2018. The carboxylate-releasing phosphorus-mobilizing
strategy can be proxied by foliar manganese concentration in a large set of
chickpea germplasm under low phosphorus supply. New Phytologist 219:
518–529.
Pang J, Ryan MH, Siddique KHM, Simpson RJ. 2017. Unwrapping the
rhizosheath. Plant and Soil 418: 129–139.
Pearse SJ, Veneklaas EJ, Cawthray G, Bolland MD, Lambers H. 2007.
Carboxylate composition of root exudates does not relate consistently to a crop
species’ ability to use phosphorus from aluminium, iron or calcium phosphate
sources. New Phytologist 173: 181–190.
Playsted CW, Johnston ME, Ramage CM, Edwards DG, Cawthray GR,
Lambers H. 2006. Functional significance of dauciform roots: exudation of
carboxylates and acid phosphatase under phosphorus deficiency in Caustis
blakei (Cyperaceae). New Phytologist 170: 491–500.
R Core Team. 2022. R: a language and environment for statistical computing.
Vienna, Austria: R Foundation for Statistical Computing.
van Ruijven J, Berendse F. 2005. Diversity-productivity relationships: initial
effects, long-term patterns, and underlying mechanisms. Proceedings of the
National Academy of Sciences, USA 102: 695–700.
Sch¨
ob C, Brooker RW, Zuppinger-Dingley D. 2018. Evolution of facilitation
requires diverse communities. Nature Ecology & Evolution 2: 1381–1385.
Shen Q, Ranathunge K, Zhong H, Finnegan PM, Lambers H. 2023. Facilitation
of phosphorus acquisition by Banksia attenuata allows Adenanthos cygnorum
(Proteaceae) to extend its range into severely phosphorus-impoverished
habitats. Plant and Soil. doi: 10.1007/s11104-023-05935-6.
Streit J, Meinen C, Nelson WCD, Siebrecht-Sch¨
oll DJ, Rauber R. 2019. Above-
and belowground biomass in a mixed cropping system with eight novel winter
faba bean genotypes and winter wheat using FTIR spectroscopy for root species
discrimination. Plant and Soil 436: 141–158.
Str¨
om L, Owen AG, Godbold DL, Jones DL. 2005. Organic acid behaviour in a
calcareous soil implications for rhizosphere nutrient cycling. Soil Biology and
Biochemistry 37: 2046–2054.
Tabatabai M, Bremner J. 1969. Use of p-nitrophenyl phosphate for assay of soil
phosphatase activity. Soil Biology and Biochemistry 1: 301–307.
Tian H, Gai J, Zhang J, Christie P, Li X. 2009. Arbuscular mycorrhizal fungi
associated with wild forage plants in typical steppe of eastern Inner Mongolia.
European Journal of Soil Biology 45: 321–327.
Tian Q, Liu N, Bai W, Li L, Chen J, Reich PB, Yu Q, Guo D, Smith MD,
Knapp AK et al. 2016. A novel soil manganese mechanism drives plant species
loss with increased nitrogen deposition in a temperate steppe. Ecology 97:65–
74.
Tilman D, Isbell F, Cowles JM. 2014. Biodiversity and ecosystem functioning.
Annual Review of Ecology, Evolution, and Systematics 45: 471–493.
Wagg C, Barry KE, O’Brien MJ, McKenzie-Gopsill A, Roscher C, Eisenhauer
N, Schmid B, Schmid B. 2019. Not even wrong: comment by Wagg et al.
Ecology 100: e02805.
Wen Z, Li H, Shen Q, Tang X, Xiong C, Li H, Pang J, Ryan MH, Lambers H,
Shen J. 2019. Tradeoffs among root morphology, exudation and mycorrhizal
symbioses for phosphorus-acquisition strategies of 16 crop species. New
Phytologist 223: 882–895.
Wen Z, Pang J, Ryan MH, Shen J, Siddique KHM, Lambers H. 2021. In
addition to foliar manganese concentration, both iron and zinc provide proxies
for rhizosheath carboxylates in chickpea under low phosphorus supply. Plant
and Soil 465:31–46.
Wright AJ, Wardle DA, Callaway R, Gaxiola A. 2017. Theoverlookedroleof
facilitation in biodiversity experiments. Trends in Ecology & Evolution 32: 383–390.
Yu Q, Chen Q, Elser JJ, He N, Wu H, Zhang G, Wu J, Bai Y, Han X. 2010.
Linking stoichiometric homoeostasis with ecosystem structure, functioning and
stability. Ecology Letters 13: 1390–1399.
Yu R, Lambers H, Callaway RM, Wright AJ, Li L. 2021. Belowground
facilitation and trait matching: two or three to tango? Trends in Plant Science
26: 1227–1235.
Yu R, Li X, Xiao Z, Lambers H, Li L. 2020a. Phosphorus facilitation and
covariation of root traits in steppe species. New Phytologist 226: 1285–1298.
Yu R, Zhang W, Yu Y, Yu S, Lambers H, Li L. 2020b. Linking shifts in species
composition induced by grazing with root traits for phosphorus acquisition in a
typical steppe in Inner Mongolia. Science of the Total Environment 712: 136495.
Zhang D, Lyu Y, Li H, Tang X, Hu R, Rengel Z, Zhang F, Whalley WR,
Davies WJ, Cahill JF Jr et al. 2020. Neighbouring plants modify maize root
foraging for phosphorus: coupling nutrients and neighbours for improved
nutrient-use efficiency. New Phytologist 226: 244–253.
Zhang D, Zhang C, Tang X, Li H, Zhang F, Rengel Z, Whalley WR, Davies
WJ, Shen J. 2016. Increased soil phosphorus availability induced by faba bean
root exudation stimulates root growth and phosphorus uptake in neighbouring
maize. New Phytologist 209: 823–831.
Supporting Information
Additional Supporting Information may be found online in the
Supporting Information section at the end of the article.
Dataset S1 List of selected literature included in the meta-analy-
sis.
Fig. S1 Effects of phosphorus addition on rhizosheath pH and
acid phosphatase activity among species in monoculture.
Fig. S2 The average proportion of five exuded carboxylates and
their relationships with leaf manganese concentration in mono-
culture.
Fig. S3 Correlation between rhizosheath acid phosphatase activ-
ity and leaf manganese concentration in monoculture.
Fig. S4 Funnel plot to evaluate the possibility of publication bias
in the meta-analysis.
Fig. S5 Correlations between relative complementarity effect and
the relative change of leaf phosphorus concentration.
Fig. S6 Correlations between relative change of specific root
length, root hair length, and root hair density of species that was
inefficient at P mobilization and that of leaf manganese concen-
tration.
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
New Phytologist (2023)
www.newphytologist.com
New
Phytologist Research 13
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Methods S1 Fourier-transform infrared spectroscopy analysis.
Methods S2 A detailed description of the data compilation and
statistical analysis of the meta-analysis.
Table S1 Information of the experiment and the tested species.
Table S2 Information of the species combinations.
Table S3 The calibration curve range in two species of 34 cali-
bration samples with known species composition used.
Table S4 Statistical parameters of the Fourier transform infrared
models in terms of calibration.
Table S5 Aboveground phosphorus concentration and content
in monoculture and mixture.
Table S6 PERMANOVA tests based on root phosphorus-mobi-
lization traits between each species pair in monoculture.
Table S7 Classification of efficient phosphorus (P)-mobilizing
species and species that are inefficient at P mobilization based on
the literature as cited.
Please note: Wiley is not responsible for the content or function-
ality of any Supporting Information supplied by the authors. Any
queries (other than missing material) should be directed to the
New Phytologist Central Office.
New Phytologist (2023)
www.newphytologist.com
Ó2023 The Authors
New Phytologist Ó2023 New Phytologist Foundation
Research
New
Phytologist
14
14698137, 0, Downloaded from https://nph.onlinelibrary.wiley.com/doi/10.1111/nph.19082 by University Of Western Australi, Wiley Online Library on [27/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License