ArticlePDF Available

Impacts of species richness on productivity in a large-scale subtropical forest experiment

Authors:
  • Institute of Evolutionary Biology and Environmental Studies

Abstract

Biodiversity experiments have shown that species loss reduces ecosystem functioning in grassland. To test whether this result can be extrapolated to forests, the main contributors to terrestrial primary productivity, requires large-scale experiments.We manipulated tree species richness by planting more than 150,000 trees in plots with 1 to 16 species. Simulating multiple extinction scenarios, we found that richness strongly increased stand-level productivity. After 8 years, 16-species mixtures had accumulated over twice the amount of carbon found in average monocultures and similar amounts as those of two commercial monocultures. Species richness effects were strongly associated with functional and phylogenetic diversity. A shrub addition treatment reduced tree productivity, but this reduction was smaller at high shrub species richness. Our results encourage multispecies afforestation strategies to restore biodiversity and mitigate climate change.
FOREST ECOLOGY
Impacts of species richness on
productivity in a large-scale
subtropical forest experiment
Yuanyuan Huang
1
*, Yuxin Chen
1
*, Nadia Castro-Izaguirre
1
, Martin Baruffol
1,2
,
Matteo Brezzi
1
, Anne Lang
3
, Ying Li
3
, Werner Härdtle
3
, Goddert von Oheimb
4,5
,
Xuefei Yang
6,7
, Xiaojuan Liu
1,8
, Kequan Pei
8
, Sabine Both
6
, Bo Yang
9
,
David Eichenberg
6,10
, Thorsten Assmann
3
, Jürgen Bauhus
11
, Thorsten Behrens
12
,
François Buscot
4,13
, Xiao-Yong Chen
14
, Douglas Chesters
15
, Bing-Yang Ding
16
,
Walter Durka
4,17
, Alexandra Erfmeier
4,18
, Jingyun Fang
19
, Markus Fischer
20
,
Liang-Dong Guo
21
, Dali Guo
22
, Jessica L. M. Gutknecht
23
, Jin-Sheng He
19
,
Chun-Ling He
15
,AndyHector
24
,LydiaHönig
6
,Ren-YongHu
25
, Alexandra-Maria Klein
26
,
Peter Kühn
12
,YuLiang
8
, Shan Li
8
,StefanMichalski
17
, Michael Scherer-Lorenzen
27
,
Karsten Schmidt
12
,ThomasScholten
12
, Andreas Schuldt
3,4
,XuezhengShi
28
,
Man-Zhi Tan
28
,ZhiyaoTang
19
,StefanTrogisch
4,6,27
,ZhengwenWang
29
,ErikWelk
4,6
,
Christian Wirth
4,10
,TesfayeWubet
4,13
,WenhuaXiang
30
, Mingjian Yu
31
,Xiao-DongYu
15
,
Jiayong Zhang
32
, Shouren Zhang
8
, Naili Zhang
8
,Hong-ZhangZhou
15
,Chao-DongZhu
15
,
Li Zhu
8
, Helge Bruelheide
4,6
,KepingMa
8
,PascalA.Niklaus
1
,BernhardSchmid
1
Biodiversity experiments have shown that species loss reduces ecosystem functioning in
grassland. To test whether this result can be extrapolated to forests, the main contributors to
terrestrial primary productivity, requires large-scale experiments. We manipulated tree species
richness by planting more than 150,000 trees in plots with 1 to 16 species. Simulating
multiple extinction scenarios, we found that richness strongly increased stand-level
productivity. After 8 years, 16-species mixtures had accumulated over twice the amount of
carbon found in average monocultures and similar amounts as those of two commercial
monocultures. Species richness effects were strongly associated with functional and
phylogenetic diversity. A shrub addition treatment reduced tree productivity, but this
reduction was smaller at high shrub species richness. Our results encourage multispecies
afforestation strategies to restore biodiversity and mitigate climate change.
Forest ecosystems harbor around two-thirds
of all terrestrial plant species. Observational
studies suggest that species-rich forests ex-
ceed the productivity of less diverse forests
(13), but covarying factors [such as spatial
heterogeneity in abiotic environment (1), species
composition (2), and successional stages (2)] make
assigning causation difficult. Systematic experimen-
tal manipulations of plant species composition in
grassland (46) have shown that plant diversity
promotes community productivity through niche
partitioning among species, specifically with re-
spect to abiotic resources (7)orinteractionswith
enemies (8), or through increasing the contribu-
tion of highly productive species in more diverse
communities (9). These two types of biological
mechanisms are thought to be captured by the
complementarity and selection effects calculated
by the additive partitioning of net biodiversity
effects (10). Complementarity effects are large and
positive when most speciesinamixturecontrib-
ute more than expected on the basis of their
monoculture values to community values, and
negative when most species in a mixture con-
tribute less than expected, whereas selection
effects are large when a single or few species
show a disproportionate contribution to com-
munity values (10). It has been postulated that
biodiversity effects may be weak or absent in
forests, especially in those of high species rich-
ness, because the coexistence of so many species
may require similar niches and competitive abil-
ities (1,1113).
Several forest biodiversity experiments have
recently been initiated (14,15), mostly in the tem-
perate zone or in small plots with limited species
richness gradients (1622). Here, we report results
of the BEF-Chinaexperiment (BEF, biodiversity
ecosystem functioning) that was established in
a highly diverse subtropical forest in southeast
China (23). The experiment is characterized by a
large species richness gradient, multiple simu-
lated extinction scenarios, high replication, and
extended duration (2009 to present). We studied
how changing tree species richness affected stand-
level development of tree basal area, aboveground
volume, and abovegroundcarbon(C)from2013
to 2017 (24). The experiment was implemented
at two sites (site A and site B) of ~20 ha each,
with communities assembled from six partially
overlapping species pools (three per site). A com-
plete pool represented a 16-species community,
which was repeatedly divided to yield reduced
richness levels of eight, four, two, and one species;
in addition, 24-species communities were created
by combining species of all three pools present at
each site (fig. S1) (24). Of the 42 tree species used
in the experiment (table S1), 40 occurred with the
same frequency at each richness level. The re-
maining two species were typical plantation
species in the area and were established in
reference monocultures. A special feature of the
design is that within each pool, communities
form nested series that simulate different trajec-
tories of trait-based species extinctions (fig. S2
and table S2). We analyzed trajectories related to
RESEARCH
Huang et al., Science 362,8083 (2018) 5 October 2018 1of4
1
Department of Evolutionary Biology and Environmental Studies, University of Zürich, Winterthurerstrasse 190, 8057 Zürich, Switzerland.
2
Instituto de Investigación de Recursos Biológicos Alexander von
Humboldt, calle 28A # 5-09, Bogotá DC, Colombia.
3
Institute of Ecology, Leuphana University of Lüneburg, Universitätsallee 1, 21335 Lüneburg, Germany.
4
German Centre for Integrative Biodiversity
Research (iDiv) HalleJenaLeipzig, Deutscher Platz 5e, 04103 Leipzig, Germany.
5
Technische Universität Dresden, Institute of General Ecology and Environmental Protection, Post Office Box 1117, 01735
Tharandt, Germany.
6
Martin Luther University Halle-Wittenberg, Am Kirchtor 1, 06108 Halle (Saale), Germany.
7
Kunming Institute of Botany, Chinese Academy of Sciences, 134, Lanhei Road, Kunming,
650204, China.
8
State Key Laboratory of Vegetation and Environmental Change, Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China.
9
Key Laboratory of Plant Resources and
Biodiversity of Jiangxi Province, Jingdezhen University, 838 Cidu Avenue, Jingdezhen, Jiangxi 333000, China.
10
Institut für Spezielle Botanik und Funktionelle Biodiversität, University of Leipzig, 04103
Leipzig, Germany.
11
Chair of Silviculture, Faculty of Environment and N atural Resources, University of Freiburg, Tennenbacherstraße 4, 79106 Freiburg, Germany.
12
Department of Geosciences, Soil Science
and Geomorphology, University of Tübingen, Rümelinstraße 19-23, 72074 Tübingen, Germany.
13
Helmholtz Centre for Environmental ResearchUFZ, Department of Soil Ecology, Theodor-Lieser-Straße 4,
06120 Halle (Saale), Germany.
14
School of Ecological and Environmental Sciences, ECNU-UH Joint Translational Science and Technology Research Institute, East China Normal University, Shanghai 200241,
China.
15
Institute of Zoology, Chinese Academy of Sciences, 1 Beichen West Road, Chaoyang District, Beijing 100101, China.
16
School of Life and Environment Sciences, Wenzhou University, Wenzhou,
China.
17
Helmholtz Centre for Environmental ResearchUFZ, Department of Community Ecology, Theodor-Lieser-Straße 4, 06120 Halle (Saale), Germany.
18
Institute for Ecosystem Research, Kiel University,
Olshausenstraße 75, 24118 Kiel, Germany.
19
Department of Ecology, Peking University, 5 Yiheyuan Road, Beijing 100871, China.
20
Institute of Plant Sciences, University of Bern, Altenbergrain 21, 3013 Bern,
Switzerland.
21
State Key Laboratory of Mycology, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China.
22
Institute of Geographic Sciences and Natural Resources Research,
Chinese Academy of Sciences, Beijing, China.
23
Department of Soil, Water, and Climate University of Minnesota, Twin Cities, MN, USA.
24
Department of Plant Sciences, University of Oxford, South Parks
Road, OX1 3RB, UK.
25
Department of Biology, College of Life and Environmental Sciences, Wenzhou University, Wenzhou 325035 China.
26
Nature Conservation and Landscape Ecology, Faculty of Earth and
Environmental Sciences, Univ ersity of Freiburg, Germany.
27
Geobotany, Faculty of Biology, University of Freiburg, Freiburg, Germany.
28
Institute of Soil Science, the Chinese Academy of Sciences, Nanjing
210008 China.
29
Institute of Applied Ecology, Chinese Academy of Sciences, 72 Wenhua Road, Shenyang 110016, P.R. China.
30
Faculty of Life Science and Technology, Central South University of Forestry
and Technology, Changsha 410004, Hunan Province, China.
31
College of Life Sciences, Zhejiang University, Hangzhou, Zhejiang 310058, China.
32
Institute of Ecology, College of Chemistry and Life Science,
Zhejiang Normal University, Yingbin Road No. 688, Jinhua City, Zhejiang Province, China 321004.
*These authors contributed equally to this work. Deceased.
Corresponding author. Email: helge.bruelheide@botanik.uni-halle.de (H.B.); kpma@ibcas.ac.cn (K.M.); pascal.niklaus@ieu.uzh.ch (P.A.N.); bernhard.schmid@ieu.uzh.ch (B.S.)
on October 4, 2018 http://science.sciencemag.org/Downloaded from
means and diversities of the three functional
traits leaf duration (LD), specific leaf area (SLA),
and wood density (WD). These traits are often
used to characterize plant-growth strategies (25)
and are potentially related to extinction proba-
bilities under environmental change (26). In 2009
(site A) and 2010 (site B), communities of 400
trees were planted on square plots 0.067 ha in
size, which equals the Chinese area unit of 1 mu.
Communities of pools A2, A3, B2, and B3 (fig. S1)
were established in single 1-mu plots. Each com-
munity of pools A1 and B1 was replicated in five
1-mu plots, four of which formed a larger square
plot of 4 mu; these four plots received an
understory shrub species richness treatment
factorially crossed with the tree species richness
gradient: Plots had zero, two, four, or eight shrub
species randomly selected from a pool of 18
species, with shrubs planted at the same density
as the trees.
We found significant positive effects of the
logarithm of tree species richness on stand basal
area and stand volume as well as on the annual
increments of these two variables (Table 1, Fig. 1,
and figs. S3 and S4). These effects grew steadi-
ly through time (changes in stand volume per
doubling of species, with standard errors, were
0.74 ± 0.58, 1.47 ± 0.85, 2.98 ± 1.29, 4.91 ± 1.83, and
6.99 ± 2.24 m
3
ha
1
from 2013 to 2017). Mean
volume increments were larger in wetter years
(F
1,99.1
= 7.58, P= 0.007), but richness effects on
volume increments were not affected by annual
precipitation (F
1,91.7
=2.25,P=0.137).After8years
of growth (site A), the average 16-species mix-
ture stored 31.5 Mg C ha
1
above ground [95%
Bayesian credible interval (CI), 25.5 to 37.6] (24),
whichismorethandoubletheamountfoundin
monocultures (11.9 Mg C ha
1
; CI, 10.6 to 13.5) (fig.
S5) and similar to the C storage of monocultures of
the commercial plantation species Cunninghamia
lanceolata (26.3 Mg C ha
1
; CI, 19.0 to 33.2) and
Pinus massoniana (28.5 Mg C ha
1
;CI,20.8to
36.1) (fig. S5). These strong positive effects of tree
species richness must have been driven by faster
growth of live trees in more diverse stands because
tree survival rate did not increase with species
richness (fig. S6). This is in contrast to findings
in a large grassland biodiversity experiment in
which positive diversity effects on productivity
were mediated by a greater number rather than
larger size of individuals in high-diversity plots (27).
The net biodiversity effect (10)onstandvol-
ume increased through time for mixtures of all
species-richness levels (year as linear term with
F
1,38.6
=29.15,P< 0.001) (Fig. 2) and was driven
by increases in complementarity effects (year as
linear term with F
1,52.4
= 9.23, P= 0.004) (Fig. 2).
Selection effects were on average negative (F
1,37.8
=
8.75, P= 0.005) because some species with rel-
atively high monoculture stand volume had lower
performanceinmixtures,andsomewithrela-
tively low monoculture stand volume had higher
performance. This was corroborated by negative
species-level selection effects (fig. S7).
We tested whether the observed species-
richness effects could be explained by functional
or phylogenetic diversity.Forthis,wecalculated
functional diversity (FD) and functional disper-
sion (FDis) (24) on the basis of the seven plant
Huang et al., Science 362,8083 (2018) 5 October 2018 2of4
Table 1. Mixed-effects models for effects of site, tree species richness (logSR), year, and interactions on stand-level tree basal area and volume
and their increments. Fixed effects were fitted sequentially (type-I sum of squares) as indicated in the table [random terms are provided in (24)].
n, numbers of plots; df, numerator degrees of freedom; ddf, denominator degrees of freedom; and logSR, log
2
(tree species richness). F and Pindicate F ratios
and the Pvalue of the significance test, respectively.
Basal area (n= 387) Volume (n= 387) Basal area increment (n= 387) Volume increment (n= 387)
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
Source of variation df ddf F Pdf ddf F Pdf ddf F Pdf ddf F P
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
Site 1 120.0 14.35 <0.001 1 100.0 20.79 <0.001 1 121.5 8.12 0.005 1 101.3 20.79 <0.001
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
LogSR 1 111.9 7.45 0.007 1 88.9 6.62 0.012 1 113.8 15.58 <0.001 1 91.2 12.41 <0.001
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
Year 4 489.4 309.0 <0.001 4 402.3 197.10 <0.001 3 287.5 9.90 <0.001 3 281.8 35.05 <0.001
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
Site × year 4 488.3 7.75 <0.001 4 410.4 20.92 <0.001 3 301.0 9.43 <0.001 3 309.0 19.62 <0.001
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
LogSR × year 4 456.2 15.21 <0.001 4 368.9 11.98 <0.001 3 265.6 3.82 0.010 3 259.0 6.18 <0.001
.................................... .................................................................. .................................................................... .................................................................... .................................................................. ............................
Site A Site B
1248162412481624
0
100
200
300
Stand volume
(m3 ha1)
A
0
20
40
60
80
12481624
2017
2016
2015
2014
2013
B
Site A Site B
12481624 12481624
−40
0
40
80
Tree species richness
Stand volume increment
(m3 ha1 yr1)
C
0
5
10
15
20
25
30
12481624
Tree species richness
2016−2017
2015−2016
2014−2015
2013−2014
D
Fig. 1. Stand-level tree volume and its increment as a function of tree species richness from
2013 to 2017. (Aand B) Stand-level tree volume. (Cand D) Stand-level tree volume increment. In
(A) and (C), raw data points and regression lines are shown for each year. (B) and (D) show
predicted means and standard errors based on mixed-effects models (Table 1). The extremes of the
point cloud taper off toward higher diversity levels because of decreasing sample size; quantile
regressions show qualitatively the same positive relationships for the largest 10% of values at each
diversity level (fig. S4). Standard deviations of species compositions (square root of corresponding
between-composition variance components), shown as black error bars above the raw data, indicate
that there is no variance-reduction effect of increasing species richness.
RESEARCH |REPORT
on October 4, 2018 http://science.sciencemag.org/Downloaded from
functional traits LD (deciduous or evergreen), SLA,
WD, leaf dry matter content, leaf nitrogen, leaf
phosphorus, and leaf thickness or the first three
of these (LD, SLA, and WD), which contributed
most to explanatory power. We also calculated
phylogenetic diversity (PD) and mean phyloge-
netic distance (MPD) for each community (24).
All measures of functional and phylogenetic di-
versity had similar explanatory power as that of
species richness for stand-level productivity mea-
sures; differences between species-richness levels
in stand volume could also be explained by asso-
ciated differences in functional or phylogenetic
diversity (fitted before species richness in model 1
in tables S3 and S4, respectively). However, none
of the functional or phylogenetic diversity mea-
sures could explain additional variation among
communities of the same richness level (when
fitted after species richness in model 2 in tables S3
and S4, respectively). This finding is consistent
with similar reports from large-scale grassland
biodiversity experiments (28). It is conceivable
that for each particular species mixture with
high stand-level productivity, a particular combi-
nation of functional traits causes the observed
biodiversity effect; this cannot be captured by
using the same functional diversity measure for
all species mixtures.
Earlier studies have suggested that positive
biodiversity effects in forests might originate from
denser crown packing and enhanced light inter-
ception in mixed-species canopies (21,29,30). We
measured the vertical crown extent of all trees in
2016 and 2017 and tested whether plots with less
crown overlap produced greater stand-level vol-
ume (24), which was not the case (F
1,446.8
= 1.73,
P= 0.189). A reason for the absence of such a
correlation might be that depending on the par-
ticular species combination, crown dissimilarity
can result from light competition (18) or from
complementary light use among species.
Despite the absence of general effects of func-
tional diversity beyond species richness, we found
some specific effects along the multiple extinction
scenarios inherent in our experimental design (fig.
S2A) (24). Changes in FD with each halving of
species richness were negatively correlated with
stand-volume changes at high but positively cor-
related at low species richness (fig. S8A), sug-
gesting that FD captured beneficial differences
between species at low but not at high diversity.
We then focused on mixtures of two species be-
cause for these, the highest number of distinct
species compositions were available. We found
that a positive correlation of net biodiversity and
complementarity effects with functional-trait dis-
tances developed over the 5 years of measure-
ments (Fig. 3 and table S5). This was also the case
for the diversity of the trait LD, indicating that
mixtures of a shade-tolerant evergreen and a
shade-intolerant deciduous broad-leaved species
captured more light than did species pairs with
uniform leaf duration.
Extinction sequences that differed in trajecto-
ries of community-weighted means for LD, SLA,
or WD (fig. S2, B to D) did not show any signifi-
cant variation in species-richness effects on stand-
levelproductivity(fig.S8,BtoD).Thissuggests
that effects of trait-based extinctions, at least the
ones tested and often considered most important
(25,26), may not differ much from effects of ran-
dom extinction. Different results might have been
obtained with other trait-based extinction scenar-
ios, either ones that we did not analyze (for exam-
ple, based on root traits) or ones that we did not
simulate.
Plots additionally planted with shrubs (24)
had reduced stand-level tree volume (F
1,234.5
=
7.30, P= 0.007), which is consistent with other
findings that shrub removal in forests can increase
tree growth (31). However, the effect of shrub
competition decreased with increasing shrub
species richness (log shrub richness F
1,191.9
=
6.57, P= 0.011), even though stand-level basal
area of shrubs did not decrease (fig. S9). The re-
duced competition between shrubs and trees at
higher shrub diversity suggests that complemen-
tarity effects extend to tree interactions with shrubs.
Our results provide strong evidence for a posi-
tive effect of tree species richness on tree produc-
tivity at stand level in establishing subtropical
forest ecosystems and support the idea that co-
occurring species in highly diverse subtropical
forest can differ in niches and competitive abil-
ities. At the end of the observation period, mixed
stands with 16 species had accumulated about
1.7 times the amount of C found in the average
monoculture (fig. S5). This effect is, on a relative
scale, similar to the 1.8-fold average increase in
aboveground stand biomass from monocultures
to 16-species mixtures in a multisite grassland
biodiversity experiment (4). Given that plant bio-
mass is higher in forests, and that the largest
Huang et al., Science 362,8083 (2018) 5 October 2018 3of4
Fig. 2. Changes over
time in the net bio-
diversity effect (NE) and
its additive compo-
nents, complementarity
effect (CE) and selec-
tion effect (SE), on
stand-level tree volume
in mixed-species plots.
N= 65 to 77, 50 to 52, 28,
and 14 plots for two-,
four-, eight-, and
16-species mixtures,
respectively. The figure
shows means ± SEs. The
yaxis is square root
scaled to reflect the
quadratic nature of bio-
diversity effects (10).
2 species 4 species 8 species 16 species
2013
2014
2015
2016
2017
2013
2014
2015
2016
2017
2013
2014
2015
2016
2017
2013
2014
2015
2016
2017
−16
−4
0
4
16
36
64
Diversity effect ( m3 ha1)
CE
NE
SE
Tree species richness of mixtures
−225
−100
−25
0
25
100
225
01234
Net effect (m3 ha−1)
A(341; 702)
−225
−100
−25
0
25
100
225
01234
Trait distance
Complementarity effect (m3 ha−1)
B
(−704; −340)
−225
−100
−25
0
25
100
225
01234
Selection effect (m3 ha−1)
2017
2016
2015
2014
2013
C
Fig. 3. Relationship between functional trait distance and biodiversity effects on stand
volume in two-species mixtures across years. (Ato C) Each point represents a plot in a year
(n= 65 to 77 plots). Regression lines are based on mixed-effects models (24). Euclidean trait distances
were calculated with the three z-transformed traits LD, SLA, and WD. The yaxes are square
rootscaled to reflect the quadratic nature of biodiversity effects (10). Two extreme yvalues are
moved to the plot margin and given as numbers.
RESEARCH |REPORT
on October 4, 2018 http://science.sciencemag.org/Downloaded from
fraction of tree C is bound in relatively persistent
woody biomass, these effects translate into large
diversity-mediated rates of C accumulation. Specif-
ically, after 8 years of growth at site A, we found
an extra 19.5 (95% CI, 14.1 to 25.1) Mg C ha
1
accu-
mulated in 16-species mixtures relative to the aver-
age monoculture. The biodiversity-productivity
effects that we found did not differ between 1-mu
and 4-mu plots (F
1,118.5
=0.07,P>0.5forinter-
action log tree species richness × plot size). How-
ever, biodiversity effects might be even larger at
spatial scales beyond the ones that we tested
experimentally because environmental heteroge-
neity might promote spatial insurance effects
(32). Our first-order extrapolation to the global
scale indicated that a 10% decrease in tree spe-
cies richness would lead to a 2.7% decrease in for-
est productivity on average (24), which is within
the range of productivity decreases (2.1 to 3.1%)
reported for the same tree species loss scenario in
a recent observational study that used plot data
covering a large part of the global forests (3). In
that study, it was estimated that such a loss would
correspond to around $20 billion per year of com-
mercial wood production.
Substantial forest areas are managed worldwide,
with large afforestation programs underway
(33,34); in China, the total forested area in-
creased by 1.5 × 10
6
ha year
1
from 2010 to 2015,
mainly because of new monoculture plantation
of species with high short-term productivity (35).
Our experimental findings suggest that a similar
or potentially even higher productivity can be
achieved with mixed plantations of native spe-
cies. Such strategies would yield cobenefits (2)in
terms of active biodiversity management and like-
ly higher levels of stability of productivity and
ecosystem services under adverse conditions such
as pathogen infestation or future climate change,
including extreme events.
REFERENCES AND NOTES
1. A. S. Mori, J. Ecol. 106, 113125 (2018).
2. L. Gamfeldt et al., Nat. Commun. 4, 1340 (2013).
3. J. Liang et al., Science 354, eaaf8957 (2016).
4. A. Hector et al., Science 286, 11231127 (1999).
5. P. B. Reich et al., Science 336, 589592 (2012).
6. D. Tilman, D. Wedin, J. Knops, Nature 379, 718720 (1996).
7. D. Tilman, C. L. Lehman, K. T. Thomson, Proc. Natl. Acad. Sci.
U.S.A. 94, 18571861 (1997).
8. S. A. Schnitzer et al., Ecology 92, 296303 (2011).
9. M. A. Huston, Oecologia 110, 449460 (1997).
10. M. Loreau, A. Hector, Nature 412,7276 (2001).
11. H. Bruelheide et al., Ecol. Monogr. 81,2541 (2011).
12. S. P. Hubbell, Ecology 87, 13871398 (2006).
13. X. Wang et al., Ecology 97, 347360 (2016).
14. M. Scherer-Lorenzen, The functional role of biodiversity in the
context of global changein Forests and Global Change,
D. A. Coomes, D. F. R. P. Burslem, W. D. Simonson, Eds.
(Cambridge Univ. Press, 2014), pp. 195238.
15. K. Verheyen et al., Ambio 45,2941 (2016).
16. J. J. Grossman, J. Cavender-Bares, S. E. Hobbie, P. B. Reich,
R. A. Montgomery, Ecology 98, 26012614 (2017).
17. C. Potvin, N. J. Gotelli, Ecol. Lett. 11, 217223 (2008).
18. J. Sapijanskas, A. Paquette, C. Potvin, N. Kunert, M. Loreau,
Ecology 95, 24792492 (2014).
19. C. M. Tobner et al., Ecol. Lett. 19, 638647 (2016).
20. T. Van de Peer, K. Verheyen, Q. Ponette, N. N. Setiawan,
B. Muys, J. Ecol. 106, 10961105 (2017).
21. L. J. Williams, A. Paquette, J. Cavender-Bares, C. Messier,
P. B. Reich, Nat. Ecol. Evol. 1, 63 (2017).
22. D. A. Clarke, P. H. York, M. A. Rasheed, T. D. Northfield, Trends
Ecol. Evol. 32, 320323 (2017).
23. H. Bruelheide et al., Methods Ecol. Evol. 5,7489 (2014).
24. Materials and methods are available as supplementary
materials.
25. P. B. Adler, A. Fajardo, A. R. Kleinhesselink, N. J. B. Kraft, Ecol.
Lett. 16, 12941306 (2013).
26. S. Greenwood et al., Ecol. Lett. 20, 539553 (2017).
27. E. Marquard et al., J. Ecol. 97, 696704 (2009).
28. D. F. B. Flynn, N. Mirotchnick, M. Jain, M. I. Palmer, S. Naeem,
Ecology 92, 15731581 (2011).
29. T. Jucker, O. Bouriaud, D. A. Coomes, Funct. Ecol. 29,
10781086 (2015).
30. P. A. Niklaus, M. Baruffol, J.-S. He, K. Ma, B. Schmid, Ecology
98,11041116 (2017).
31. R. G. Wagner, K. M. Little, B. Richardson, K. McNabb, Forestry
79,5779 (2006).
32. F. Isbell et al., Nature 546,6572 (2017).
33. Food and Agriculture Organization of the United Nations (FAO),
Global Forest Resources Assessment (FAO, 2015).
34. R. J. Keenan et al., For. Ecol. Manage. 352,920 (2015).
35. F. Hua et al., Nat. Commun. 7, 12717 (2016).
ACKNO WLEDGM ENTS
We thank farmers for help in the field. Funding: This study was
supported by the German Research Foundation (DFG FOR 891),
the Strategic Priority Research Program of the Chinese Academy
of Sciences (nos. XDB31000000 and XDA19050000), the National
Natural Science Foundation of China (NSFC nos. 31270496 and
31300353), the Swiss National Science Foundation (SNSF nos.
130720 and 147092), and the European Union (EC 7th Framework
Program no. 608422). Author contributions: Y.H. and Y.C. are
co-first authors. H.B., W.H., J.-S. H., A.H., K.M., T.S., and B.S.
conceived the project; M.Ba., M.Br., N.C., D.E., J.F., Y.H., Y.L.,
S.L., X.L., S.M., T.S., X.Y., and B.Y. collected the data; Y.H., Y.C.,
P.A.N., and B.S. analyzed and interpreted the data and wrote the
manuscript. All authors discussed the results and contributed
to the final manuscript. Competing interests: The authors declare
no competing interests. Data and materials availability: The
data supporting the findings of this study are deposited in Dryad
with the accession code doi: 10.5061/dryad.t86145r.
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/362/6410/80/suppl/DC1
Materials and Methods
Supplementary Text
Figs. S1 to S9
Tables S1 to S5
References (3657)
2 May 2018; accepted 24 August 2018
10.1126/science.aat6405
Huang et al., Science 362,8083 (2018) 5 October 2018 4of4
RESEARCH |REPORT
on October 4, 2018 http://science.sciencemag.org/Downloaded from
Impacts of species richness on productivity in a large-scale subtropical forest experiment
Keping Ma, Pascal A. Niklaus and Bernhard Schmid
Xiao-Dong Yu, Jiayong Zhang, Shouren Zhang, Naili Zhang, Hong-Zhang Zhou, Chao-Dong Zhu, Li Zhu, Helge Bruelheide,
Zhiyao Tang, Stefan Trogisch, Zhengwen Wang, Erik Welk, Christian Wirth, Tesfaye Wubet, Wenhua Xiang, Mingjian Yu,
Michalski, Michael Scherer-Lorenzen, Karsten Schmidt, Thomas Scholten, Andreas Schuldt, Xuezheng Shi, Man-Zhi Tan,
Chun-Ling He, Andy Hector, Lydia Hönig, Ren-Yong Hu, Alexandra-Maria Klein, Peter Kühn, Yu Liang, Shan Li, Stefan
Alexandra Erfmeier, Jingyun Fang, Markus Fischer, Liang-Dong Guo, Dali Guo, Jessica L. M. Gutknecht, Jin-Sheng He,
Jürgen Bauhus, Thorsten Behrens, François Buscot, Xiao-Yong Chen, Douglas Chesters, Bing-Yang Ding, Walter Durka,
Goddert von Oheimb, Xuefei Yang, Xiaojuan Liu, Kequan Pei, Sabine Both, Bo Yang, David Eichenberg, Thorsten Assmann,
Yuanyuan Huang, Yuxin Chen, Nadia Castro-Izaguirre, Martin Baruffol, Matteo Brezzi, Anne Lang, Ying Li, Werner Härdtle,
DOI: 10.1126/science.aat6405
(6410), 80-83.362Science
, this issue p. 80Science
forests could benefit both restoration of biodiversity and mitigation of climate change. mixedeffects of tree diversity on forest productivity and carbon accumulation. Thus, changing from monocultures to more
sizes with a wide range of species richness levels. After 8 years of the experiment, the findings suggest strong positive
experiment in a subtropical forest in China. The study combines many replicates, realistic tree densities, and large plot
report the first results from a large biodiversityet al.ecosystem functioning. Is the same true for forests? Huang
Experimental studies in grasslands have shown that the loss of species has negative consequences for
Tree diversity improves forest productivity
ARTICLE TOOLS http://science.sciencemag.org/content/362/6410/80
MATERIALS
SUPPLEMENTARY http://science.sciencemag.org/content/suppl/2018/10/03/362.6410.80.DC1
REFERENCES http://science.sciencemag.org/content/362/6410/80#BIBL
This article cites 47 articles, 3 of which you can access for free
PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions
Terms of ServiceUse of this article is subject to the
is a registered trademark of AAAS.Science
licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. The title
Science, 1200 New York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive
(print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience
on October 4, 2018 http://science.sciencemag.org/Downloaded from
www.sciencemag.org/content/362/6410/80/suppl/DC1
Supplementary Material for
Impacts of species richness on productivity in a large-scale subtropical
forest experiment
Yuanyuan Huang, Yuxin Chen, Nadia Castro-Izaguirre, Martin Baruffol, Matteo Brezzi,
Anne Lang, Ying Li, Werner Härdtle, Goddert von Oheimb, Xuefei Yang, Xiaojuan Liu,
Kequan Pei, Sabine Both, Bo Yang, David Eichenberg, Thorsten Assmann,
Jürgen Bauhus, Thorsten Behrens, François Buscot, Xiao-Yong Chen, Douglas Chesters,
Bing-Yang Ding, Walter Durka, Alexandra Erfmeier, Jingyun Fang, Markus Fischer,
Liang-Dong Guo, Dali Guo, Jessica L. M. Gutknecht, Jin-Sheng He, Chun-Ling He,
Andy Hector, Lydia Hönig, Ren-Yong Hu, Alexandra-Maria Klein, Peter Kühn,
Yu Liang, Shan Li, Stefan Michalski, Michael Scherer-Lorenzen, Karsten Schmidt,
Thomas Scholten, Andreas Schuldt, Xuezheng Shi, Man-Zhi Tan, Zhiyao Tang,
Stefan Trogisch, Zhengwen Wang, Erik Welk, Christian Wirth, Tesfaye Wubet,
Wenhua Xiang, Mingjian Yu, Xiao-Dong Yu, Jiayong Zhang, Shouren Zhang,
Naili Zhang, Hong-Zhang Zhou, Chao-Dong Zhu, Li Zhu, Helge Bruelheide*,
Keping Ma*, Pascal A. Niklaus*, Bernhard Schmid*
‡Corresponding author. Email: helge.bruelheide@botanik.uni-halle.de (H.B.); kpma@ibcas.ac.cn (K.M.);
pascal.niklaus@ieu.uzh.ch (P.A.N.); bernhard.schmid@ieu.uzh.ch (B.S.)
Published 5 October 2018, Science 362, 80 (2017)
DOI: 10.1126/science.aat6405
This PDF file includes:
Materials and Methods
Supplementary Text
Figs. S1 to S9
Tables S1 to S5
References
2
Materials and Methods
Study site and experimental design
The BEF-China experimental platform is located in Jiangxi Province, China (29°08¢–29°11¢N,
117°90¢–117°93¢E). Climate at the site is subtropical, with mean annual temperature and
precipitation of 16.7 °C and 1800 mm, respectively (averaged from 1971–2000) (36). During the
study period from 2013–2017 mean annual temperature was 18.1, 18.0, 17.5, 18.0 and 17.9 °C,
whereas annual precipitation was more variable with 1354, 2110, 2632, 1944 and 2338 mm,
respectively (37). The region is covered by subtropical broad-leaved forest and plantations of two
commercially important coniferous species, Pinus massoniana and Cunninghamia lanceolata (see
below). We established a large-scale tree biodiversity experiment in 2009–2010 at two sites (A and
B) of approximately 20 ha each. These sites were previously used as Cunninghamia lanceolata
plantations, which still surround the experimental sites. We planted a total of 226,400 individual
trees on 566 plots (23). For the present study we used 396 plots, planted with a total of 158,400
individuals, in which species-loss scenarios were simulated that included each of 40 species at each
level of species richness from 24 to sixteen, eight, four, two and one species and the two conifers as
reference monocultures. Species names and abbreviations together with major characteristics and
initial size at planting are provided in Table S1.
Three pools of 16 species were created at each site: A1–A3 and B1–B3 (Tables S1 and S2).
The species in each 16-species pool were put in random sequence and then repeatedly divided in
halves until monocultures were obtained. For each site, this procedure resulted in 69 unique species
compositions plus one 24-species mixture combining all species of the three pools of that site plus
the reference monocultures of the two conifers. Each plot was 25.8 × 25.8 m in size (Chinese area
unit of 1 mu) and planted with 400 tree individuals arranged in a rectangular 20 × 20 grid with 1.29
m spacing between rows and columns, corresponding to an area of 1.675 m2 per individual at
planting. Planting density for species such as Cunninghamia lanceolata in commercial plantations
around the field site is 1 tree per 4 m2 (2 m distance between trees). Accounting for mortality,
which on average was 6 out of 16 central trees (Fig. S6), the average area per tree had increased to
about 3 m2 by 2017. Our objective when establishing a relatively high planting density was to allow
for early interactions among trees, as would be expected in natural stands at the experimental site
(11). From measurements at site A over the four-year period 2013–2016 we found that mean crown
projection area per tree increased from 1.1–2.7 m2 in the average monoculture and from 2.2–4.1 m2
in the average 16-species mixture, leading to LAI values that increased from 1.1-1.8 in the average
monoculture and from 1.5-3.7 in the average 16-species mixture. To minimize edge effects, plots
were established adjacent to each other, with trees thus forming a continuous cover across the entire
site. Site A was planted in 2009, site B in 2010.
Plots were randomly distributed in rectangular grids across the two sites (Fig. S1). Each
species composition of pools A1 and B1 was replicated across five plots, of which four were
arranged into a larger quadrat of 51.6 × 51.6 m. This was done to allow two additional treatment
effects to be tested: plot size (1 mu vs. 4 mu) and species richness of understory shrubs. For the
latter treatment the four plots in a larger quadrat were planted with 0, 2, 4 or 8 shrub species,
randomly selected from a pool of 18 species, at the same total density as the trees; each individual
shrub was planted in the center between four adjacent trees. Species compositions of pools A2, A3,
B2 and B3 were planted in only one plot each and without shrub addition. Of the 396 plots, nine
had to be excluded because these were not established due to a lack of sapling material or high
initial mortality. All plots were weeded annually to remove emerging herbs and woody species that
were not part of the planting design.
3
Tree measurements
We studied how changing tree species richness along the different functional-trait trajectories
of simulated extinction scenarios in BEF-China affected the stand-level development of tree basal
area, aboveground volume and aboveground carbon from 2013–2017. These productivity-related
variables were derived from direct measurements of tree basal diameter and height, using
allometric equations determined by complete aboveground harvests of young trees in a forest near
the experimental area. The direct measurements were taken for all survivors of the 16 central trees
in each plot of 400 trees in September/October within at most 23 consecutive days per site per year.
Because we analyzed plot-level rather than individual-level data, we focused on a comparatively
large number of replicates at plot rather than individual level within the given time available for
measurements. A larger number of trees sampled within plots would have increased the precision
of the plot-level data and the chances to obtain higher significance levels for treatment effects.
Data from 154 separately harvested trees near the experimental site were used to obtain
conversion factors to calculate aboveground tree volume and biomass from the direct
measurements of basal diameter and height (see section “Conversion factors for individual tree
volume and biomass” below) and aggregated the individual data to the stand level. Biomass was
converted to carbon content (38) by multiplying with 0.474 g C g-1. To characterize annual stand
growth, we further derived yearly increments of stand basal area, stand volume and stand carbon
from successive inventories. We determined the same metrics at the population level (stand-level
data separated into species). We used a Bayesian approach to estimate the uncertainties of plot-
level carbon arising from the use of different allometric equations for experimental species and the
two commercial monoculture species (see section “Estimation of aboveground carbon and its
uncertainties” below).
Conversion factors for individual tree volume and biomass
We harvested 154 trees in a natural forest in 2010 near the experimental sites to determine
conversion factors from cylindrical volume (tree basal area × height) to true volume and biomass
(39). The trees belonged to eight common species and three life forms (evergreen angiosperms,
deciduous angiosperms and conifers) and were chosen to represent a naturally occurring size span
of young trees.
Trees were separated into large woody parts (stems and large branches with a diameter
!
3
cm), twigs (the apical part of the stem and large branches plus side branches with a diameter < 3
cm) and dead attached material (large dead branches or twigs). Branches were divided into
segments of about 1 m length. The volume of large woody parts and twigs was determined
geometrically, approximating the parts as truncated cones (large woody parts, V =
1
3π
"
(r1
2+r1r2+r22) L where L is the length and r1 and r2 are the end radii), or cones (twigs, as above
but r2 = 0). The density of these fractions was determined by oven-drying a representative
subsample of stem and branch discs or twigs. These geometric and density data were then scaled up
to total aboveground tree biomass, modeling twig mass and density in dependence of branch
positions within tree crowns (39).
Conversion factors from cylindrical volume to true volume (and mass) were determined as
total tree volume (and tree mass, including leaves) divided by cylindrical volume. We analyzed the
variation of these conversion factors with tree size and species life form using mixed-effects
models with species identity as random term. We found that large trees deviated from the linear
relationship between form factor and cylindrical volume, and we therefore removed trees with a
4
cylindrical volume 0.5 m3 from the calibration, leaving a set of 119 trees. Within this set, there
was only a small variance among species and no significant effect of life form on the form factor;
the form factor decreased linearly with the cylindrical volume of harvested trees. We therefore used
a form factor of 0.5412 m3 m-3
#
0.1985 m-3 BA h (with basal area BA in m2 and height h in m). The
intercept of 0.5412 m3 m-3 is the weighted average form factor of evergreen and deciduous species
at size zero (in our study, 19 of 40 species were evergreen and 21 deciduous). Biomass factors were
determined similarly, yielding a conversion factor of 269.13 kg m-3
#
141.96 kg m-6 BA h. For the
two coniferous species, Pinus massoniana and Cunninghamia lanceolata, we used separate
equations obtained from the harvested trees of these two species. Here the form factor was 0.5083
m3 m-3
#
0.1985 m-6 BA h and the biomass factor was 216.79 kg m-3
#
141.96 kg m-6 BA h.
Using the conversion factors obtained from the harvested trees in allometric equations we
estimated their volume and biomass from basal area and height; the estimated volume was strongly
correlated with the real volume of the trees (r2 = 0.907 for angiosperms and r2 = 0.891 for
gymnosperms) and the same was the case for estimated and real biomass (r2 = 0.913 for
angiosperms and r2 = 0.830 for gymnosperms).
Estimation of aboveground carbon and its uncertainties
We used Bayesian statistical techniques to estimate plot-level carbon after eight years of
growth (site A in 2017) and its uncertainty arising from the allometry models. First, we re-fitted the
allometric models with the data of harvested trees as we did in the non-Bayesian approach:
$%&'&()*
+
,(- ."/0%&'&( - 1'& 2
3
""""""""""""""""""""""""""""""""
4
56
7
&
where
$%&'&(
is the conversion factor for biomass derived from harvested tree i of species j and life
form k;
/0%&'&(
is the raw volume of harvested trees calculated as cylindrical volume,
,(
is life-form
specific intercept,
.
is the slope,
1'
is the species-level random effect and
2
is the model error. The
Bayesian models produced similar results as those from non-Bayesian methods using linear mixed-
effects models.
Then we used the fitted Bayesian allometric models (eqn. S1) to predict the posterior
distribution of the conversion factors for each tree in the main experiment:
$
8
9)*
4
,
:
- ."/09& 2
7
""""""""""""""""""""""""""""""""""""""""""""""""""""
4
5;
7
&
where
$
8
9
is the predicted conversion factor for biomass of tree t from the main experiment,
,
: is the
average intercept weighted by the proportion of the corresponding life form in the experiment and
/09
is the raw volume of tree t. We omitted species-level random effects in equation S2 because we
found only a small variance among species. The posterior distribution of biomass was derived as
$
8
9</09
. We further derived the posterior distribution of carbon as the product of predicted
biomass and carbon density. Finally, we calculated the posterior distribution of plot-level carbon by
summing the tree-specific carbon values within each plot. Therefore, the plot-level carbon
contained the uncertainties from both the allometric parameters (
,
: and
.
) and model error (
2
).
We fitted equations S1 and S2 together in a single Bayesian model. We ran the Bayesian
models in JAGS 4.2.0 using the rjags package (40) with three parallel chains. We set diffuse priors
for each parameter and assessed the parameter convergence with Gelman and Rubin’s convergence
diagnostics (with a threshold value < 1.05) (41).
Additive partitioning of net biodiversity effects into complementarity and selection effects
We used the additive partitioning method of Loreau and Hector (10) to decompose net
biodiversity effects (NEs) of productivity measures into complementarity effects (CEs) and
selection effects (SEs), separately for each year and diversity level. CEs and SEs depend on relative
5
yields of species, which we calculated using monoculture volume as denominator (10). This
method has the problem that very small monoculture values lead to unrealistically large relative
yield values and these values should therefore be excluded from additive partitioning calculations
(42). We used a stand-level tree volume of 0.2 m3 ha–1 in monoculture as cut-off point to avoid
extreme relative yield values. Formally, CEs and SEs are related to (co-)variances and therefore
were square-root transformed with sign reconstruction (sign(y)=|y|) prior to analysis, which
improved the normality of residuals (10).
We used the following equation to calculate individual species SEs:
5>%?"4@AB
%# @AB"7 <
+
C%# C
3
Here
@AB
%
is the deviation from expected relative yield of species
D
in the mixture and
C%
is the
yield of species
D
in monoculture.
Functional and phylogenetic diversity measures
We used seven functional traits to calculate measures of functional diversity for all
communities in the experiment. These traits were determined in plots that were part of the
experiment (43): leaf duration (LD, deciduous or evergreen), specific leaf area (SLA), branch-wood
density (WD), leaf dry matter content, leaf nitrogen, leaf phosphorus and leaf thickness. We used
species means to calculate functional diversity (FD) (44) and functional dispersion (FDis) (45).
After testing the predictive power of these functional diversity measures in explaining variation in
stand-level productivity measures for all traits individually and in different combinations, we
retained those combining all seven traits or combining the three traits LD, SLA and WD. These
three functional traits are known to be associated with a trade-off between rapid resource
acquisition and fast growth vs. high tolerance to environmental stress and slower growth (46, 47).
For these three traits we further calculated individual and multivariate Euclidean trait distance (TD)
in two-species mixtures to assess its relationship with the net biodiversity effect and its components
(NE, CE and SE). For the multivariate TD the three traits were first standardized with a z-
transformation. We calculated two phylogenetic diversity measures (48) with a node age-calibrated
phylogenetic tree (49): Faith’s phylogenetic diversity (PD) and mean pairwise phylogenetic
distance (MPD). FD and PD are dependent on species richness and thus encompass functional and
phylogenetic diversity both among and within species-richness levels. Distance measures (FDis,
TD and MPD) are independent of species richness (48).
Vertical crown extent and overlap
In 2016 and 2017 we measured the crown extent of each surviving tree as interval between the
lowest side branch and the top of the tree. Species means per plot were then used to calculate
vertical crown overlap between species as proportional similarity (30), PSA,B = AÇB / AÈB, where
AÇB is the vertical extent that is occupied by both species and AÈB is the extent occupied by at
least one of the two species A or B. For mixtures with more than two species we used the mean
proportional similarity between all possible pairs as measure for vertical crown overlap.
Functional-trait trajectories of extinction sequences
We derived trait-based trajectories of extinction sequences from the changes of FDs and
community-weighted mean trait values of the three functional traits LD, SLA and WD across
richness gradients (Fig. S2). Community-weighted means were calculated with equal weights for
all planted species in a plot. For each species pool, a single 16-species mixture leads to 16
extinction trajectories ending in different monocultures.
6
Statistical analysis
We used analysis of variance based on type-I sum of squares in linear mixed-effects models to
assess the effects of tree species richness and other explanatory variables on productivity (50). All
analyses were done in R 3.3.2 and ASReml-R (51). The models included the fixed effects site, tree
species richness (log2-transformed), year (as continuous variable, i.e. linear term—followed by
annual precipitation in tests for climatic effects, or as multi-level factor), the interaction log2(tree
species richness) × year, and the interaction site × year. Random effects were species composition
(with a separate variance component for each site), large plot (set of four plots arranged in a
quadrat for pool A1 and B1, see Table S2; with a separate variance component for each site), plot
and their interactions with year. The interaction of year and site and the site-specific variance terms
estimated for some random terms accounted for the fact that site B was established one year after
site A and that trees at site B were therefore smaller. Model residuals were checked for normality
and homogeneity of variances; these assumptions were fulfilled without transformations of
dependent variables. For the analyses of shrub diversity effects, the model contained the additional
fixed effects shrub presence (a two-level factor: 0 vs. 2, 4 or 8 shrub species), log2 of shrub species
richness (for shrub-species richness > 0), plot size (a two-level factor: 1 vs. 4 mu) and the
interactions of all these terms with log2(tree species richness) and with year.
To assess whether functional or phylogenetic diversity measures explained effects of species
richness or additional variation in dependent variables these measures were fitted as covariates
before (model 1) or after species richness (model 2) in separate analyses of variance (Tables S3 and
S4). The effects of vertical crown overlap were tested in the same way but not presented in tables
because they were not significant. We further focused on two-species mixtures to assess the effects
of trait distances (TD) on biodiversity effects (NE, CE and SE) using linear mixed-effects models.
We set site, TD, year (factor) and the interaction between TD and year as fixed effects, community
composition, plot size and their interactions with year as random effects. For these analyses
diversity effects were based on the tree stand volume from 2013–2017 and were square-root
transformed with sign reconstruction to improve normality of model residuals. We excluded one
plot with extreme values of CE and SE (absolute value > 2000) in 2017 before model fitting.
However, including this extreme plot produced qualitatively similar results.
We separated FD-based extinction scenarios (Fig. S2A) into four steps by halving species
richness from 16 to 8, 8 to 4 etc. and assessed the relationship between changes in FD and volume
(average across years) for each step with simple linear regression. Then, we regressed the four
slopes of these regressions against step, 1 referring to extinction step 2 à 1, 2 to step 4 à 2 etc.
(Fig. S8A). For the other three types of functional-trait trajectories (Fig. S2, B to D) we plotted
species-richness effects (average slopes across years of stand volume vs. log2-transformed tree
species richness for each extinction sequence) against changes in functional trait means (slopes of
functional trait mean vs. log2-transformed tree species richness for each extinction sequence) and
tested the relationship using simple linear regression (Fig. S7, B to D).
7
Supplementary Text
Productivity decrease due to 10 % loss of tree species richness
Liang et al. (3) used global forest observational data to estimate the effect of 10 % loss of tree
species richness on productivity (measured as volume increment). To compare the diversity effect
of our experiment with the global estimate, we predicted the productivity loss under the same
pressure of species loss. First, we predicted the volume increment from 2016 to 2017 at the planted
species richness (1, 2, 4, 8, 16) and their corresponding 10 % lower richness levels using our fitted
mixed-effects models (Table 1). Then we took the average productivity decrease due to 10 % loss
of tree richness across planted richness levels. The productivity drops would be 3.0 %, 2.5 %, 2.2
%, and 1.9 % when losing on average 10 % of the species, starting from species richness 2, 4, 8 and
16, respectively. These values correspond closely to the 2.1–3.1 % drop obtained with the power
function in Liang et al. (3), which they converted to a potential yearly commercial forest
productivity valued at 20 billion US$.
8
Fig. S1. Map of BEF-China experimental plots analyzed in this paper. Each species
composition of pools A1 (yellow) and B1 (light blue) is replicated across five plots of which four
are arranged into a larger quadrat of 51.6 × 51.6 m; species compositions of pools A2, A3, B2 and
B3 (other colors) are not replicated (see Tables S1, S2). Monocultures of two commercial conifers,
Pinus massoniana and Cunninghamia lanceolata, are replicated at both sites in five plots each
(dark blue). Note that in some cases two pools share a monoculture species (plots with diagonal
coloring). Empty plots belong to the overall BEF-China experiment but not to the treatments
analyzed in the present paper.
Site A Site B
N
1
4
1
2
1
4
8
1
2
2
4
1
1
2
4
1
1
4
1
1
2
4
1
2
1
2
8
4
2
2
4
4
2
1
2
1
1
1
1
16
1
2
2
16
1
2
1
4
8
1
1
1
2
2
2
8
4
1
1
1
1
1
2
8
1
24
2
1
8
2
2
2
2
1
16
1
4
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
4
4
4
4
8
8
16
24
4
1
8
8
1
4
16
2
4
41
2
1
1
1
1
2
1
1
1
1
2
16
16
2
1
1
2
2
4
2
8
1
1
2
2
2
4
1
2
1
8
24
1
1
1
1
1
1
2
4
2
1
4
2
1
8
4
1
1
4
2
2
1
2
1
8
2
1
1
1
2
2
2
2
4
4
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
4
4
4
4
8
8
16
24
Random extinction: pool A1 pool A2 pool A3
pool B1 pool B2 pool B3
economic speciesReference plots:
100 m
sites 5 km
apart
9
Fig. S2. Functional-trait trajectories of the different extinction sequences based on functional-
trait diversity FD (A), mean leaf-duration class (proportion of evergreen species) (B), mean
specific leaf area (SLA) (C) and mean wood density (D). Extinction sequences are shown from
right to left as function of species richness (log2-scale). Different colors represent different species
pools. FD was calculated with the three functional traits shown in B–D. Means in B–D are
community-weighted means with species weighted equally.
0
5
10
15
1248 16
FD
A
0.0
0.5
1.0
1248 16
Proportion of evergreen sp.
B
8
12
16
20
1248 16
SLA (mm2mg1)
C
0.3
0.4
0.5
0.6
0.7
1248 16
WD (g cm3)
D
Tree species richness
10
Fig. S3. Stand-level tree basal area (A, B) and its annual increment (C, D) as a function of tree
species richness from 2013–2017. In panels A and C, raw data points and regression lines are
shown for each year. Panels B and D show predicted means ± standard errors based on mixed
models (Table 1). Standard deviations of species compositions (square root of corresponding
between-composition variance components) are shown as black error bars above the raw data.
Site A
Site B
1 2 4 8 16 24 1 2 4 8 16 24
0
20
40
60
Stand basal area
(m2 ha1)
A
0
5
10
15
20
25
1 2 4 8 16 24
2017
2016
2015
2014
2013
B
Site A
Site B
1 2 4 8 16 24 1 2 4 8 16 24
10
0
10
20
Tree species richness
Stand basal area increment
(m2 ha1 yr1)
C
0
1
2
3
4
5
6
1 2 4 8 16 24
Tree species richness
20162017
20152016
20142015
20132014
D
11
Fig. S4. Stand-level tree volume (A) and its increment (B) as a function of tree species
richness from 2013–2017. The figure panels A and B correspond to panels A and C in Fig. 2 of the
main text, respectively, but with dashed quantile regression lines for the largest 10% of values at
each diversity level added to the solid regression lines fitted across all values for each year.
Standard deviations of species compositions (square root of corresponding between-composition
variance components) are shown as black error bars.
Site A
Site B
1 2 4 8 16 24 1 2 4 8 16 24
0
100
200
300
Stand volume
(m3 ha1)
2017
2016
2015
2014
2013
A
Site A
Site B
1 2 4 8 16 24 1 2 4 8 16 24
40
0
40
80
Tree species richness
Stand volume increment
(m3 ha1 yr1)
20162017
20152016
20142015
20132014
B
12
Fig. S5. Aboveground stand-level tree carbon (A, B) and its annual increment (C, D) as a
function of tree species richness from 2013–2017. In panels A and C, raw data points and
regression lines are shown for each year. On the left of each panel, means ± standard errors for the
two economic tree species are shown (PiMa = Pinus massoniana; CuLa = Cunninghamia
lanceolata). Panels B and D show predicted means ± standard errors based on mixed models (Table
1). Standard deviations of species compositions (square root of corresponding between-
composition variance components) are shown as black error bars above the raw data.
Site A
Site B
PiMa
CuLa
1 2 4 8 16 24
PiMa
CuLa
1 2 4 8 1624
0
20
40
60
Aboveground carbon
(Mg C ha1)
A
0
5
10
15
20
1 2 4 8 16 24
2017
2016
2015
2014
2013
B
Site A
Site B
PiMa
CuLa
1 2 4 8 16 24
PiMa
CuLa
1 2 4 8 1624
0
10
20
Tree species richness
Aboveground carbon increment
(Mg C ha1yr1)
C
0
1
2
3
4
5
6
7
1 2 4 8 16 24
Tree species richness
20172016
20162015
20152014
20142013
D
13
Fig. S6. Stand density as a function of tree species richness from 2013–2017. Raw data points
are shown together with regression lines (for guidance only, effects of species richness and year by
species richness interactions were not significant in mixed-model analysis: P > 0.2). Density
indicates the number of surviving trees out of 16 planted in the central area of each plot.
Site A
Site B
1 2 4 8 16 24 1 2 4 8 16 24
0
5
10
15
Tree species richness
Stand density (number of surviving trees)
2013
2014
2015
2016
2017
14
Fig. S7. Monoculture stand-level tree volume of species in 2017 (A) and species-specific
selection effects (SEs) from 2013–2017 on stand-level tree volume (B). Plotted are means ±
standard errors. The y-axis in B is square-root scaled to reflect the quadratic nature of selection
effects (10).
NA
Site A (planted in 2009)
Site B (planted in 2010)
0
50
100
150
Monoculture volume in 2017
(m3ha1)
A
Schima superba
Choerospondias axillaris
Cinnamomum camphora
Castanea henryi
Nyssa sinensis
Liquidambar formosana
Quercus acutissima
Triadica sebifera
Sapindus saponaria
Lithocarpus glaber
Triadica cochinchinensis
Diospyros japonica
Cyclobalanopsis myrsinifolia
Melia azedarach
Cyclobalanopsis glauca
Daphniphyllum oldhamii
Castanopsis carlesii
Castanopsis sclerophylla
Quercus fabri
Quercus serrata
Castanopsis eyrei
Koelreuteria bipinnata
Rhus chinensis
Acer davidii
Alniphyllum fortunei
Elaeocarpus glabripetalus
Elaeocarpus chinensis
Elaeocarpus japonicus
Lithocarpus glaber
Manglietia fordiana
Schima superba
Cinnamomum camphora
Daphniphyllum oldhamii
Castanopsis fargesii
Castanopsis sclerophylla
Cyclobalanopsis glauca
Diospyros japonica
Betula luminifera
Idesia polycarpa
Phoebe bournei
Machilus thunbergii
Ailanthus altissima
Quercus phillyreoides
Machilus leptophylla
Machilus grijsii
Castanopsis eyrei
Celtis biondii
Meliosma flexuosa
36
16
4
0
4
16
Species
Selection effect
(m3ha1)
2017
2016
2015
2014
2013
B
15
Fig. S8. Relationship between functional-trait trajectories of extinction sequences and
reduction in stand volume when communities lose half of their species. In A, changes in
functional trait diversity (FD) and volumes are shown separately for each step along the extinction
sequences whereas in B–D slopes of stand volumes vs. log2-transformed species richness are shown
for each extinction sequence with the given changes in community-weighted mean traits (Fig. S2).
In A, slopes fitted for the different extinction steps decrease linearly from positive at low to
negative at high richness (F1,2 = 120.233, P = 0.008), that is, high FD tends to increase volume
gains at low species richness but to decrease them at high species richness. The slopes of the
regression lines in B–D are not significantly different from zero (P > 0.2). Each point represents an
average value across five years.
100
50
0
50
100
0246
FD difference
Volume difference (m3ha1)
step
21
42
84
168
A
10
0
10
20
0.2 0.1 0.0 0.1 0.2
Change in proportions of evergreen sp.
Diversity effect
(m3ha1per doubling sp.)
B
10
0
10
20
210 1 2
Change in SLA
(mm2mg1per doubling sp.)
Diversity effect
(m3ha1per doubling sp.)
C
10
0
10
20
0.06 0.03 0.00 0.03 0.06
Change in WD
(g cm3per doubling sp.)
Diversity effect
(m3ha1per doubling sp.)
D
16
Fig. S9. Mean basal area of the longest ramet of the 18 shrub species used to assemble shrub-
diversity treatments (A) and effects of shrub species richness on average stand-level tree basal
area and shrub basal area (B). In B, stand-level basal areas of trees (circles, left y-axis) and
understory shrubs (triangles, right y-axis; only data for 2017) are shown with means ± standard
errors. Data are from species pools A1 and B1 (see Tables S1 and S2).
Site A (planted in 2009)
Site B (planted in 2010)
Distylium buxifolium
Eurya muricata
Syzygium buxifolium
Distylium myricoides
Gardenia jasminoides
Loropetalum chinense
Camellia chekiangoleosa
Rhaphiolepis indica
Ardisia crenata
Itea chinensis
Distylium buxifolium
Syzygium buxifolium
Itea chinensis
Viburnum setigerum
Hydrangea chinensis
Rhododendron ovatum
Phyllanthus glaucus
Rhododendron simsii
Photinia hirsuta
Ficus erecta
0
1
2
3
4
5
Shrub species
Shrub mean basal area in 2017
(cm2 individual1)
A
2013
2014
2015
2016
2017
0248 0248 0248 0248 0248
0
5
10
15
0.0
0.5
1.0
1.5
Shrub diversity level
Tree basal area (m2 ha1)
Shrub basal area (m2 ha1)
B
17
Table S1. Characteristics of tree species used in the BEF-China experiment. Information about
sites and species pools refers to the design presented in Fig. S1 and Table S2. Information about
other species characters was extracted from references (36, 52-57). Initial height in m (means ±
standard errors) was directly measured in the year of planting.
Species
Abbre-
viation
Site
Pool
Leaf
dur.
Shade
Succ.
state
Habitat by description from floras
Initial
height
Acer davidii Franchet
AcDa
A
A2, A3
D
I
I
streams, roadsides and sparse forest
25.8 ±3.7
Ailanthus altissima (Miller)
Swingle
AiAl
B
B1, B3
D
I
E/I
62.4 ±5.0
Alniphyllum fortunei
(Hemsley) Makino
AlFo
B
B1, B3
D
I
E
southern slopes of weed forest
32 ±1.2
Betula luminifera H.
Winkler in Engler
BeLu
B
B1, B2
D
I
E
valleys, streams, piedmont and
sunny mountain slopes
29.5 ±1.4
Castanea henryi (Skan)
Rehd. et Wils.
CaHe
A
A1, A3
D
T
E
70.4 ±3.3
Castanopsis carlesii
(Hemsley) Hayata
CaCa
A
A2, A3
E
T
L
mixed and evergreen broadleaf
forest
12.5 ±4.4
Castanopsis eyrei
(Champion ex Bentham)
Tutcher
CaEy
AB
A1, A2,
B1, B2
E
T
L
evergreen broadleaf forest or
mixed coniferous and broadleaf
forest, hills, dense or sparse
montane forest
13.8 ±0.8
Castanopsis fargesii
Franchet
CaFa
B
B1, B2
E
T
I/L
slopes and valleys
14.4 ±0.6
Castanopsis sclerophylla
(Lindley & Paxton) Schottky
CaSc
AB
A1, A3,
B2, B3
E
T
E/I/L
16.4 ±0.8
Celtis biondii Pampanini
CeBi
B
B1, B2
D
T
E/I
26.1 ±1.7
Choerospondias axillaris
(Roxb.) Burtt et Hill
ChAx
A
A1, A3
D
I
E
lowland, hills and mountain forest
107.6 ±3.1
Cinnamomum camphora
(Linnaeus) J. Presl in
Berchtold & J. Presl
CiCa
AB
A2, A3,
B2, B3
E
T
E/I/L
25.9 ±1.5
Cunninghamia lanceolata
(Lamb.) Hook.
CuLa
AB
E
I
E
25.6 ±0.9
Cyclobalanopsis glauca
(Thunberg) Oersted
CyGl
AB
A1, A3,
B2, B3
E
T
I/L
slopes, streams and valleys,
evergreen broadleaf forest or
mixed mesophytic forest
12.3 ±0.7
Cyclobalanopsis myrsinifolia
(Blume) Oersted
CyMy
A
A1, A2
E
T
I/L
lower montane broadleaf forest,
mixed mesophytic forest in valleys
11.6 ±0.6
Daphniphyllum oldhamii
(Hemsley) K. Rosenthal in
Engler
DaOl
AB
A2, A3,
B2, B3
E
T
L
slopes of broadleaf forest
19.4 ±1.1
Diospyros japonica Siebold
& Zuccarini
DiJa
AB
A2, A3,
B2, B3
D
I
E
valleys, slopes, mixed forest by
streams in ravines
38.2 ±2.2
Elaeocarpus chinensis
(Gardner & Champion) J. D.
Hooker ex Bentham
ElCh
B
B1, B3
E
T
I/L
weed forest of mountain slopes,
evergreen forest
26 ±1.3
Elaeocarpus glabripetalus
Merrill
ElGl
B
B1, B3
E
T
I/L
31.9 ±1.3
Elaeocarpus japonicus
Siebold & Zuccarini
ElJa
B
B1, B2
E
T
I/L
valleys, mountain slopes,
streamsides, evergreen broadleaf
forest
26.6 ±1.0
18
Idesia polycarpa
Maximowicz
IdPo
B
B1, B3
D
I
E
sunny slopes, streamsides in the
forest, deciduous broadleaf forest,
mixed coniferous and broadleaf
forest
29.3 ±1.4
Koelreuteria bipinnata
Franch.
KoBi
A
A1, A2
D
I
E
slopes and streamsides, sparse
forest
31.5 ±1.2
Liquidambar formosana
Hance
LiFo
A
A1, A3
D
I
I
43 ±1.5
Lithocarpus glaber (Thunb.)
Nakai
LiGl
AB
A1, A2,
B1, B2
E
T
I/L
weed forest, mixed mesophytic
forest, frequent on sunny slopes
18.1 ±0.6
Machilus grijsii Hance
MaGr
B
B1, B2
E
T
E/I
montane shrubland, dense forest or
margins of forest, thickets
11.7 ±1.8
Machilus leptophylla
Handel-Mazzetti
MaLe
B
B1, B2
E
T
I/L
11.1 ±0.5
Machilus thunbergii Siebold
& Zuccarini
MaTh
B
B1, B3
E
T
I/L
mountain slopes or valleys,
evergreen broadleaf forest
9 ±0.9
Manglietia fordiana Oliver
MaFo
B
B1, B2
E
T
I/L
hills, forest between rivers
18.6 ±0.7
Melia azedarach Linnaeus
MeAz
A
A2, A3
D
I
I
64 ±3.4
Meliosma flexuosa
Pampanini
MeFl
B
B1, B3
D
I
E/I
submontane broadleaf forest or
mixed coniferous and broadleaf
forest
23.4 ±3.4
Nyssa sinensis Oliver
NySi
A
A1, A2
D
I
E
valleys, sunny slopes of wet
broadleaf forest; wet mixed forest
along valleys and streams
55.5 ±2.9
Phoebe bournei (Hemsley)
Yen C. Yang
PhBo
B
B1, B3
E
T
I/L
mountain valleys, evergreen
broadleaf forest
12.9 ±0.6
Pinus massoniana Lamb.
PiMa
AB
E
I
E/I
22.2 ±1.1
Quercus acutissima
Carruthers
QuAc
A
A2, A3
D
I
E
42.4 ±4.6
Quercus fabri Hance
QuFa
A
A1, A2
D
I
E
25.6 ±1.5
Quercus phillyreoides A.
Gray
QuPh
B
B1, B2
E
T
I/L
hills, submontane, bare rocks and
cliffs, mixed mesophytic forest
17.6 ±3.3
Quercus serrata Murray
QuSe
A
A1, A3
D
I
E
37.7 ±1.6
Rhus chinensis Mill.
RhCh
A
A1, A3
D
I
E
47.7 ±2.2
Sapindus saponaria
Linnaeus
SaSa
A
A1, A2
D
I
E
slopes, streamsides and ravines,
margins of forest
52.5 ±1.9
Schima superba Gardn. et
Champ.
ScSu
AB
A1, A2,
B1, B2
E
T
E/I/L
33.1 ±1.1
Triadica cochinchinensis
Loureiro
TrCo
A
A2, A3
D
I
E
moist evergreen broadleaf forest,
montane forest or brushwood
85.1 ±13.7
Triadica sebifera (L.) Small
TrSe
A
A1, A3
D
I
E
forest on limestone
71.4 ±2.3
Notes: leaf dur(ation) D = deciduous, E = evergreen; shade (tolerance) I = intolerant, T = tolerant,
succ(essional) stage E = early, I = intermediate, L = late.
19
Table S2. Experimental design.
Site
Pool
Species
richness
Plot size
Shrub
treatment
Species composition
A
A1
16
4mu/1mu
yes
CyGl
QuFa
RhCh
ScSu
CaEy
CyMy
KoBi
LiGl
CaHe
NySi
LiFo
SaSa
CaSc
QuSe
ChAx
TrSe
8
4mu/1mu
yes
CyGl
QuFa
RhCh
ScSu
CaEy
CyMy
KoBi
LiGl
CaHe
NySi
LiFo
SaSa
CaSc
QuSe
ChAx
TrSe
4
4mu/1mu
yes
CyGl
QuFa
RhCh
ScSu
CaEy
CyMy
KoBi
LiGl
CaHe
NySi
LiFo
SaSa
CaSc
QuSe
ChAx
TrSe
2
4mu/1mu
yes
CyGl
QuFa
RhCh
ScSu
CaEy
CyMy
KoBi
LiGl
CaHe
NySi
LiFo
SaSa
CaSc
QuSe
ChAx
TrSe
1
4mu/1mu
yes
CyGl
QuFa
RhCh
ScSu
CaEy
CyMy
KoBi
LiGl
CaHe
NySi
LiFo
SaSa
CaSc
QuSe
ChAx
TrSe
A2
16
1mu
no
CaCa
LiGl
AcDa
MeAz
CaEy
KoBi
CiCa
CyMy
DiJa
NySi
TrCo
ScSu
DaOl
QuFa
QuAc
SaSa
8
1mu
no
CaCa
LiGl
AcDa
MeAz
CaEy
KoBi
CiCa
CyMy
DiJa
NySi
TrCo
ScSu
DaOl
QuFa
QuAc
SaSa
4
1mu
no
CaCa
LiGl
AcDa
MeAz
CaEy
KoBi
CiCa
CyMy
DiJa
NySi
TrCo
ScSu
DaOl
QuFa
QuAc
SaSa
2
1mu
no
CaCa
LiGl
AcDa
MeAz
CaEy
KoBi
CiCa
CyMy
DiJa
NySi
TrCo
ScSu
DaOl
QuFa
QuAc
SaSa
1
1mu
no
CaCa
LiGl
AcDa
MeAz
CaEy
KoBi
CiCa
CyMy
DiJa
NySi
TrCo
ScSu
DaOl
QuFa
QuAc
SaSa
A3
16
1mu
no
AcDa
QuAc
CaHe
RhCh
CaSc
CiCa
LiFo
MeAz
CaCa
CyGl
TrCo
TrSe
ChAx
DiJa
DaOl
QuSe
8
1mu
no
AcDa
QuAc
CaHe
RhCh
CaSc
CiCa
LiFo
MeAz
CaCa
CyGl
TrCo
TrSe
ChAx
DiJa
DaOl
QuSe
4
1mu
no
AcDa
QuAc
CaHe
RhCh
CaSc
CiCa
LiFo
MeAz
CaCa
CyGl
TrCo
TrSe
ChAx
DiJa
DaOl
QuSe
2
1mu
no
AcDa
QuAc
CaHe
RhCh
CaSc
CiCa
LiFo
MeAz
CaCa
CyGl
TrCo
TrSe
ChAx
DiJa
DaOl
QuSe
1
1mu
no
AcDa
QuAc
CaHe
RhCh
CaSc
CiCa
LiFo
MeAz
CaCa
CyGl
TrCo
TrSe
ChAx
DiJa
DaOl
QuSe
B
B1
16
4mu/1mu
yes
AiAl
MeFl
IdPo
MaGr
CeBi
ElGl
ElJa
PhBo
BeLu
CaFa
MaFo
QuPh
ElCh
MaTh
AlFo
MaLe
8
4mu/1mu
yes
AiAl
MeFl
IdPo
MaGr
CeBi
ElGl
ElJa
PhBo
BeLu
CaFa
MaFo
QuPh
ElCh
MaTh
AlFo
MaLe
4
4mu/1mu
yes
AiAl
MeFl
IdPo
MaGr
CeBi
ElGl
ElJa
PhBo
BeLu
CaFa
MaFo
QuPh
ElCh
MaTh
AlFo
MaLe
2
4mu/1mu
yes
AiAl
MeFl
IdPo
MaGr
CeBi
ElGl
ElJa
PhBo
BeLu
CaFa
MaFo
QuPh
ElCh
MaTh
AlFo
MaLe
1
4mu/1mu
yes
AiAl
MeFl
IdPo
MaGr
CeBi
ElGl
ElJa
PhBo
BeLu
CaFa
MaFo
QuPh
ElCh
MaTh
AlFo
MaLe
B2
16
1mu
no
CaEy
CeBi
MaLe
PhBo
DiJa
LiGl
ElGl
MaTh
AiAl
AlFo
CaFa
CaSc
CyGl
ScSu
CiCa
DaOl
8
1mu
no
CaEy
CeBi
MaLe
PhBo
DiJa
LiGl
ElGl
MaTh
AiAl
AlFo
CaFa
CaSc
CyGl
ScSu
CiCa
DaOl
4
1mu
no
CaEy
CeBi
MaLe
PhBo
DiJa
LiGl
ElGl
MaTh
AiAl
AlFo
CaFa
CaSc
CyGl
ScSu
CiCa
DaOl
2
1mu
no
CaEy
CeBi
MaLe
PhBo
DiJa
LiGl
ElGl
MaTh
AiAl
AlFo
CaFa
CaSc
CyGl
ScSu
CiCa
DaOl
1
1mu
no
CaEy
CeBi
MaLe
PhBo
DiJa
LiGl
ElGl
MaTh
AiAl
AlFo
CaFa
CaSc
CyGl
ScSu
CiCa
DaOl
B3
16
1mu
no
BeLu
DaOl
CaEy
QuPh
CyGl
MaGr
ElJa
LiGl
CaSc
IdPo
ElCh
MaFo
CiCa
DiJa
MeFl
ScSu
8
1mu
no
BeLu
DaOl
CaEy
QuPh
CyGl
MaGr
ElJa
LiGl
CaSc
IdPo
ElCh
MaFo
CiCa
DiJa
MeFl
ScSu
4
1mu
no
BeLu
DaOl
CaEy
QuPh
CyGl
MaGr
ElJa
LiGl
CaSc
IdPo
ElCh
MaFo
CiCa
DiJa
MeFl
ScSu
2
1mu
no
BeLu
DaOl
CaEy
QuPh
CyGl
MaGr
ElJa
LiGl
CaSc
IdPo
ElCh
MaFo
CiCa
DiJa
MeFl
ScSu
1
1mu
no
BeLu
DaOl
CaEy
QuPh
CyGl
MaGr
ElJa
LiGl
CaSc
IdPo
ElCh
MaFo
CiCa
DiJa
MeFl
ScSu
Note: See Table S1 for species abbreviations.
20
Table S3. Summary statistics from mixed-effects models assessing the effects of functional-
trait diversity measures (44, 45) and species richness (logSR) on stand volume from 2013–
2017. Left side measures based on seven functional traits, right side measures based on three
functional traits.
Seven traits
df
ddf
F
P
Three traits
df
ddf
F
P
Model 1
Site
1
99.7
20.68
<0.001
Site
1
99.9
20.71
<0.001
FD
1
81.6
6.59
0.012
FD
1
81.6
6.33
0.014
LogSR
1
107.3
0.10
0.755
LogSR
1
113.6
0.36
0.550
Year
4
393.8
196.9
<0.001
Year
4
396.6
196.6
<0.001
Site × year
4
407.2
20.80
<0.001
Site ×year
4
408.9
20.84
<0.001
FD × year
4
335.1
12.48
<0.001
FD ×year
4
335.9
11.73
<0.001
LogSR ×year
4
441.4
0.08
0.988
LogSR ×year
4
470.0
0.58
0.678
Model 2
Site
1
99.7
20.68
<0.001
Site
1
99.7
20.71
<0.001
LogSR
1
87.5
6.58
0.012
LogSR
1
87.8
6.59
0.012
FD
1
99.8
0.11
0.746
FD
1
104.6
0.09
0.760
Year
4
393.8
196.9
<0.001
Year
4
396.6
196.6
<0.001
Site ×year
4
407.2
20.80
<0.001
Site ×year
4
408.9
20.84
<0.001
LogSR ×year
4
360.8
11.95
<0.001
LogSR ×year
4
363.5
11.94
<0.001
FD ×year
4
410.3
0.61
0.654
FD ×year
4
429.7
0.37
0.829
Model 1
Site
1
99.9
20.71
<0.001
Site
1
99.7
20.68
<0.001
FDis
1
85.9
3.77
0.055
FDis
1
80.8
4.66
0.034
LogSR
1
93.7
2.82
0.096
LogSR
1
88.8
2.05
0.156
Year
4
398.7
195.7
<0.001
Year
4
396.6
196.2
<0.001
Site × year
4
409.9
20.82
<0.001
Site ×year
4
408.9
20.83
<0.001
FDis × year
4
357.3
6.80
<0.001
FDis ×year
4
335.1
8.52
<0.001
LogSR ×year
4
389.9
5.14
<0.001
LogSR ×year
4
368.1
3.68
0.006
Model 2
Site
1
99.9
20.71
<0.001
Site
1
99.7
20.68
<0.001
LogSR
1
88.0
6.58
0.012
LogSR
1
87.5
6.58
0.012
FDis
1
91.5
0.010
0.919
FDis
1
81.8
0.12
0.729
Year
4
398.7
195.7
<0.001
Year
4
396.6
196.2
<0.001
Site ×year
4
409.9
20.82
<0.001
Site ×year
4
408.9
20.83
<0.001
LogSR ×year
4
365.7
11.89
<0.001
LogSR ×year
4
363.7
11.92
<0.001
FDis ×year
4
380.9
0.041
0.997
FDis ×year
4
338.1
0.27
0.894
Note: FD is Petchey and Gaston’s functional diversity (44); FDis is functional dispersion (45).
Abbreviations: df = numerator degrees of freedom; ddf = denominator degrees of freedom. F and P
indicate F-ratios and P-values of the significance tests.
21
Table S4. Summary statistics from mixed-effects models assessing the effects of phylogenetic
diversity measures (48) and species richness (logSR) on stand volume from 2013–2017.
df
ddf
F
P
Model 1
Site
1
99.7
20.67
<0.001
PD
1
78.6
5.79
0.018
LogSR
1
97.9
0.79
0.375
Year
4
395.3
196.0
<0.001
Site × year
4
408.3
20.80
<0.001
PD × year
4
324.8
10.20
<0.001
LogSR ×year
4
405.6
1.85
0.118
Model 2
Site
1
99.7
20.67
<0.001
LogSR
1
87.4
6.57
0.012
PD
1
88.1
0.02
0.902
Year
4
395.3
196.0
<0.001
Site ×year
4
408.3
20.80
<0.001
LogSR ×year
4
362.4
11.90
<0.001
PD ×year
4
364.7
0.16
0.959
Model 1
Site
1
99.4
20.93
<0.001
MPD
1
84.2
8.61
0.004
LogSR
1
93.8
0.55
0.459
Year
4
400.2
198.1
<0.001
Site × year
4
408.6
21.06
<0.001
MPD × year
4
348.7
11.71
<0.001
LogSR ×year
4
390.7
2.31
0.057
Model 2
Site
1
99.4
20.93
<0.001
LogSR
1
88.2
6.66
0.011
MPD
1
89.4
2.50
0.117
Year
4
400.2
198.1
<0.001
Site ×year
4
408.6
21.06
<0.001
LogSR ×year
4
367.0
12.05
<0.001
MPD ×year
4
370.8
1.97
0.098
Notes: PD is Faith’s phylogenetic diversity; MPD is mean pairwise phylogenetic distance.
Abbreviations: df = numerator degrees of freedom; ddf = denominator degrees of freedom. F and P
indicate F-ratios and P-values of the significance tests.
22
Table S5. Summary statistics from mixed-effects models assessing the temporal change of the
relationships between trait distance (TD) and biodiversity effects on stand volume in two-
species mixtures. TDs were calculated with leaf duration, specific leaf area (SLA) and wood
density (WD), and jointly with the three z-transformed traits (multivariate TD). Biodiversity effects
obtained by additive partitioning were analyzed: net biodiversity effect (NE), complementarity
effect (CE) and selection effect (SE).
NE
CE
SE
df
ddf
F
P
df
ddf
F
P
df
ddf
F
P
Leaf duration
TD
1
26.3
0.070
0.793
1
23.6
0.492
0.490
1
25.9
2.717
0.111
Year
4
74.4
1.054
0.385
4
100.2
0.587
0.673
4
107.2
0.393
0.813
TD × year
4
68.9
1.757
0.148
4
93.0
3.921
0.006
4
99.4
3.378
0.012
SLA
TD
1
34.8
0.841
0.366
1
32.2
2.411
0.130
1
35.0
0.922
0.344
Year
4
82.9
0.924
0.454
4
86.8
0.718
0.582
4
102.5
0.433
0.785
TD × year
4
83.8
0.824
0.514
4
88.5
6.745
<0.001
4
103.4
2.275
0.066
WD
TD
1
33.7
0.943
0.338
1
38.0
0.706
0.406
1
33.2
0.010
0.919
Year
4
74.4
0.970
0.429
4
104.6
0.580
0.678
4
110.6
0.363
0.835
TD × year
4
102.0
1.673
0.162
4
136.5
1.359
0.251
4
138.6
0.406
0.804
Multivariate TD
TD
1
36.4
1.540
0.223
1
36.3
2.609
0.115
1
33.4
1.119
0.298
Year
4
76.4
1.069
0.378
4
84.8
0.844
0.501
4
102.5
0.432
0.785
TD × year
4
101.2
3.325
0.013
4
118.1
9.300
<0.001
4
123.4
3.571
0.009
Notes: biodiversity effects were square-root transformed with sign reconstruction (
E
ign(y)=|y|).
TD, year and their interaction were fitted after site [for random terms see (24)]. Abbreviations: df =
numerator degrees of freedom; ddf = denominator degrees of freedom. F and P indicate F-ratios
and P-values of the significance tests.
References and Notes
1. A. S. Mori, Environmental controls on the causes and functional consequences of tree species
diversity. J. Ecol. 106, 113–125 (2018). doi:10.1111/1365-2745.12851
2. L. Gamfeldt, T. Snäll, R. Bagchi, M. Jonsson, L. Gustafsson, P. Kjellander, M. C. Ruiz-Jaen,
M. Fröberg, J. Stendahl, C. D. Philipson, G. Mikusiński, E. Andersson, B. Westerlund, H.
Andrén, F. Moberg, J. Moen, J. Bengtsson, Higher levels of multiple ecosystem services
are found in forests with more tree species. Nat. Commun. 4, 1340 (2013).
doi:10.1038/ncomms2328 Medline
3. J. Liang, T. W. Crowther, N. Picard, S. Wiser, M. Zhou, G. Alberti, E.-D. Schulze, A. D.
McGuire, F. Bozzato, H. Pretzsch, S. de-Miguel, A. Paquette, B. Hérault, M. Scherer-
Lorenzen, C. B. Barrett, H. B. Glick, G. M. Hengeveld, G.-J. Nabuurs, S. Pfautsch, H.
Viana, A. C. Vibrans, C. Ammer, P. Schall, D. Verbyla, N. Tchebakova, M. Fischer, J. V.
Watson, H. Y. H. Chen, X. Lei, M.-J. Schelhaas, H. Lu, D. Gianelle, E. I. Parfenova, C.
Salas, E. Lee, B. Lee, H. S. Kim, H. Bruelheide, D. A. Coomes, D. Piotto, T. Sunderland,
B. Schmid, S. Gourlet-Fleury, B. Sonké, R. Tavani, J. Zhu, S. Brandl, J. Vayreda, F.
Kitahara, E. B. Searle, V. J. Neldner, M. R. Ngugi, C. Baraloto, L. Frizzera, R. Bałazy, J.
Oleksyn, T. Zawiła-Niedźwiecki, O. Bouriaud, F. Bussotti, L. Finér, B. Jaroszewicz, T.
Jucker, F. Valladares, A. M. Jagodzinski, P. L. Peri, C. Gonmadje, W. Marthy, T.
O’Brien, E. H. Martin, A. R. Marshall, F. Rovero, R. Bitariho, P. A. Niklaus, P. Alvarez-
Loayza, N. Chamuya, R. Valencia, F. Mortier, V. Wortel, N. L. Engone-Obiang, L. V.
Ferreira, D. E. Odeke, R. M. Vasquez, S. L. Lewis, P. B. Reich, Positive biodiversity-
productivity relationship predominant in global forests. Science 354, eaaf8957 (2016).
doi:10.1126/science.aaf8957 Medline
4. A. Hector, B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G. Dimitrakopoulos,
J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P. Hogberg, K. Huss-Danell, J.
Joshi, A. Jumpponen, C. Korner, P. W. Leadley, M. Loreau, A. Minns, C. P. Mulder, G.
O’Donovan, S. J. Otway, J. S. Pereira, A. Prinz, D. J. Read, l. et, Plant diversity and
productivity experiments in european grasslands. Science 286, 1123–1127 (1999).
doi:10.1126/science.286.5442.1123 Medline
5. P. B. Reich, D. Tilman, F. Isbell, K. Mueller, S. E. Hobbie, D. F. B. Flynn, N. Eisenhauer,
Impacts of biodiversity loss escalate through time as redundancy fades. Science 336,
589–592 (2012). doi:10.1126/science.1217909 Medline
6. D. Tilman, D. Wedin, J. Knops, Productivity and sustainability influenced by biodiversity in
grassland ecosystems. Nature 379, 718–720 (1996). doi:10.1038/379718a0
7. D. Tilman, C. L. Lehman, K. T. Thomson, Plant diversity and ecosystem productivity:
Theoretical considerations. Proc. Natl. Acad. Sci. U.S.A. 94, 1857–1861 (1997).
doi:10.1073/pnas.94.5.1857 Medline
8. S. A. Schnitzer, J. N. Klironomos, J. Hillerislambers, L. L. Kinkel, P. B. Reich, K. Xiao, M. C.
Rillig, B. A. Sikes, R. M. Callaway, S. A. Mangan, E. H. van Nes, M. Scheffer, Soil
microbes drive the classic plant diversity-productivity pattern. Ecology 92, 296–303
(2011). doi:10.1890/10-0773.1 Medline
9. M. A. Huston, Hidden treatments in ecological experiments: Re-evaluating the ecosystem
function of biodiversity. Oecologia 110, 449–460 (1997). doi:10.1007/s004420050180
Medline
10. M. Loreau, A. Hector, Partitioning selection and complementarity in biodiversity
experiments. Nature 412, 72–76 (2001). doi:10.1038/35083573 Medline
11. H. Bruelheide, M. Böhnke, S. Both, T. Fang, T. Assmann, M. Baruffol, J. Bauhus, F. Buscot,
X.-Y. Chen, B.-Y. Ding, W. Durka, A. Erfmeier, M. Fischer, C. Geißler, D. Guo, L.-D.
Guo, W. Härdtle, J.-S. He, A. Hector, W. Kröber, P. Kühn, A. C. Lang, K. Nadrowski, K.
Pei, M. Scherer-Lorenzen, X. Shi, T. Scholten, A. Schuldt, S. Trogisch, G. von Oheimb,
E. Welk, C. Wirth, Y.-T. Wu, X. Yang, X. Zeng, S. Zhang, H. Zhou, K. Ma, B. Schmid,
Community assembly during secondary forest succession in a Chinese subtropical forest.
Ecol. Monogr. 81, 25–41 (2011). doi:10.1890/09-2172.1
12. S. P. Hubbell, Neutral theory and the evolution of ecological equivalence. Ecology 87, 1387–
1398 (2006). doi:10.1890/0012-9658(2006)87[1387:NTATEO]2.0.CO;2 Medline
13. X. Wang, T. Wiegand, N. J. B. Kraft, N. G. Swenson, S. J. Davies, Z. Hao, R. Howe, Y. Lin,
K. Ma, X. Mi, S.-H. Su, I.-F. Sun, A. Wolf, Stochastic dilution effects weaken
deterministic effects of niche-based processes in species rich forests. Ecology 97, 347–
360 (2016). doi:10.1890/14-2357.1 Medline
14. M. Scherer-Lorenzen, “The functional role of biodiversity in the context of global change” in
Forests and Global Change, D. A. Coomes, D. F. R. P. Burslem, W. D. Simonson, Eds.
(Cambridge Univ. Press, 2014), pp. 195–238.
15. K. Verheyen, M. Vanhellemont, H. Auge, L. Baeten, C. Baraloto, N. Barsoum, S. Bilodeau-
Gauthier, H. Bruelheide, B. Castagneyrol, D. Godbold, J. Haase, A. Hector, H. Jactel, J.
Koricheva, M. Loreau, S. Mereu, C. Messier, B. Muys, P. Nolet, A. Paquette, J. Parker,
M. Perring, Q. Ponette, C. Potvin, P. Reich, A. Smith, M. Weih, M. Scherer-Lorenzen,
Contributions of a global network of tree diversity experiments to sustainable forest
plantations. Ambio 45, 29–41 (2016). doi:10.1007/s13280-015-0685-1 Medline
16. J. J. Grossman, J. Cavender-Bares, S. E. Hobbie, P. B. Reich, R. A. Montgomery, Species
richness and traits predict overyielding in stem growth in an early-successional tree
diversity experiment. Ecology 98, 2601–2614 (2017). doi:10.1002/ecy.1958 Medline
17. C. Potvin, N. J. Gotelli, Biodiversity enhances individual performance but does not affect
survivorship in tropical trees. Ecol. Lett. 11, 217–223 (2008). doi:10.1111/j.1461-
0248.2007.01148.x Medline
18. J. Sapijanskas, A. Paquette, C. Potvin, N. Kunert, M. Loreau, Tropical tree diversity
enhances light capture through crown plasticity and spatial and temporal niche
differences. Ecology 95, 2479–2492 (2014). doi:10.1890/13-1366.1
19. C. M. Tobner, A. Paquette, D. Gravel, P. B. Reich, L. J. Williams, C. Messier, Functional
identity is the main driver of diversity effects in young tree communities. Ecol. Lett. 19,
638–647 (2016). doi:10.1111/ele.12600 Medline
20. T. Van de Peer, K. Verheyen, Q. Ponette, N. N. Setiawan, B. Muys, Overyielding in young
tree plantations is driven by local complementarity and selection effects related to shade
tolerance. J. Ecol. 106, 1096–1105 (2017). doi:10.1111/1365-2745.12839
21. L. J. Williams, A. Paquette, J. Cavender-Bares, C. Messier, P. B. Reich, Spatial
complementarity in tree crowns explains overyielding in species mixtures. Nat. Ecol.
Evol. 1, 63 (2017). doi:10.1038/s41559-016-0063 Medline
22. D. A. Clarke, P. H. York, M. A. Rasheed, T. D. Northfield, Does biodiversity–ecosystem
function literature neglect tropical ecosystems? Trends Ecol. Evol. 32, 320–323 (2017).
doi:10.1016/j.tree.2017.02.012 Medline
23. H. Bruelheide, K. Nadrowski, T. Assmann, J. Bauhus, S. Both, F. Buscot, X.-Y. Chen, B.
Ding, W. Durka, A. Erfmeier, J. L. M. Gutknecht, D. Guo, L.-D. Guo, W. Härdtle, J.-S.
He, A.-M. Klein, P. Kühn, Y. Liang, X. Liu, S. Michalski, P. A. Niklaus, K. Pei, M.
Scherer-Lorenzen, T. Scholten, A. Schuldt, G. Seidler, S. Trogisch, G. von Oheimb, E.
Welk, C. Wirth, T. Wubet, X. Yang, M. Yu, S. Zhang, H. Zhou, M. Fischer, K. Ma, B.
Schmid, Designing forest biodiversity experiments: General considerations illustrated by
a new large experiment in subtropical China. Methods Ecol. Evol. 5, 74–89 (2014).
doi:10.1111/2041-210X.12126
24. Materials and methods are available as supplementary materials.
25. P. B. Adler, A. Fajardo, A. R. Kleinhesselink, N. J. B. Kraft, Trait-based tests of coexistence
mechanisms. Ecol. Lett. 16, 1294–1306 (2013). doi:10.1111/ele.12157 Medline
26. S. Greenwood, P. Ruiz-Benito, J. Martínez-Vilalta, F. Lloret, T. Kitzberger, C. D. Allen, R.
Fensham, D. C. Laughlin, J. Kattge, G. Bönisch, N. J. B. Kraft, A. S. Jump, Tree
mortality across biomes is promoted by drought intensity, lower wood density and higher
specific leaf area. Ecol. Lett. 20, 539–553 (2017). doi:10.1111/ele.12748 Medline
27. E. Marquard, A. Weigelt, C. Roscher, M. Gubsch, A. Lipowsky, B. Schmid, Positive
biodiversity-productivity relationship due to increased plant density. J. Ecol. 97, 696–704
(2009). doi:10.1111/j.1365-2745.2009.01521.x
28. D. F. B. Flynn, N. Mirotchnick, M. Jain, M. I. Palmer, S. Naeem, Functional and
phylogenetic diversity as predictors of biodiversity–ecosystem-function relationships.
Ecology 92, 1573–1581 (2011). doi:10.1890/10-1245.1 Medline
29. T. Jucker, O. Bouriaud, D. A. Coomes, Crown plasticity enables trees to optimize canopy
packing in mixed-species forests. Funct. Ecol. 29, 1078–1086 (2015). doi:10.1111/1365-
2435.12428
30. P. A. Niklaus, M. Baruffol, J.-S. He, K. Ma, B. Schmid, Can niche plasticity promote
biodiversity-productivity relationships through increased complementarity? Ecology 98,
1104–1116 (2017). doi:10.1002/ecy.1748 Medline
31. R. G. Wagner, K. M. Little, B. Richardson, K. McNabb, The role of vegetation management
for enhancing productivity of the world’s forests. Forestry 79, 57–79 (2006).
doi:10.1093/forestry/cpi057
32. F. Isbell, A. Gonzalez, M. Loreau, J. Cowles, S. Díaz, A. Hector, G. M. Mace, D. A. Wardle,
M. I. O’Connor, J. E. Duffy, L. A. Turnbull, P. L. Thompson, A. Larigauderie, Linking
the influence and dependence of people on biodiversity across scales. Nature 546, 65–72
(2017). doi:10.1038/nature22899 Medline
33. Food and Agriculture Organization of the United Nations (FAO), Global Forest Resources
Assessment (FAO, 2015).
34. R. J. Keenan, G. A. Reams, F. Achard, J. V. de Freitas, A. Grainger, E. Lindquist, Dynamics
of global forest area: Results from the FAO Global Forest Resources Assessment 2015.
For. Ecol. Manage. 352, 9–20 (2015). doi:10.1016/j.foreco.2015.06.014
35. F. Hua, X. Wang, X. Zheng, B. Fisher, L. Wang, J. Zhu, Y. Tang, D. W. Yu, D. S. Wilcove,
Opportunities for biodiversity gains under the world’s largest reforestation programme.
Nat. Commun. 7, 12717 (2016). doi:10.1038/ncomms12717 Medline
36. X. Yang, J. Bauhus, S. Both, T. Fang, W. rdtle, W. Kröber, K. Ma, K. Nadrowski, K. Pei,
M. Scherer-Lorenzen, T. Scholten, G. Seidler, B. Schmid, G. von Oheimb, H. Bruelheide,
Establishment success in a forest biodiversity and ecosystem functioning experiment in
subtropical China (BEF-China). Eur. J. For. Res. 132, 593–606 (2013).
doi:10.1007/s10342-013-0696-z
37. Climate data from National Meteorological Information Center, China Meteorological
Administration; http://data.cma.cn.
38. A. R. Martin, S. C. Thomas, A reassessment of carbon content in tropical trees. PLOS ONE
6, e23533 (2011). doi:10.1371/journal.pone.0023533 Medline
39. M. Brezzi, thesis, University of Zurich, 2015.
40. M. Plummer, rjags-package: Bayesian graphical models using MCMC. R package version 4-
6. (2016).
41. A. Gelman, D. B. Rubin, Inference from iterative simulation using multiple sequences. Stat.
Sci. 7, 457–472 (1992). doi:10.1214/ss/1177011136
42. E. Marquard, A. Weigelt, V. M. Temperton, C. Roscher, J. Schumacher, N. Buchmann, M.
Fischer, W. W. Weisser, B. Schmid, Plant species richness and functional composition
drive overyielding in a six-year grassland experiment. Ecology 90, 3290–3302 (2009).
doi:10.1890/09-0069.1 Medline
43. W. Kröber, Y. Li, W. Härdtle, K. Ma, B. Schmid, K. Schmidt, T. Scholten, G. Seidler, G.
von Oheimb, E. Welk, C. Wirth, H. Bruelheide, Early subtropical forest growth is driven
by community mean trait values and functional diversity rather than the abiotic
environment. Ecol. Evol. 5, 3541–3556 (2015). doi:10.1002/ece3.1604 Medline
44. O. L. Petchey, K. J. Gaston, Functional diversity (FD), species richness and community
composition. Ecol. Lett. 5, 402–411 (2002). doi:10.1046/j.1461-0248.2002.00339.x
45. E. Laliberté, P. Legendre, A distance-based framework for measuring functional diversity
from multiple traits. Ecology 91, 299–305 (2010). doi:10.1890/08-2244.1 Medline
46. B. F. Chabot, D. J. Hicks, The ecology of leaf life spans. Annu. Rev. Ecol. Syst. 13, 229–259
(1982). doi:10.1146/annurev.es.13.110182.001305
47. S. J. Wright, K. Kitajima, N. J. B. Kraft, P. B. Reich, I. J. Wright, D. E. Bunker, R. Condit, J.
W. Dalling, S. J. Davies, S. Díaz, B. M. J. Engelbrecht, K. E. Harms, S. P. Hubbell, C. O.
Marks, M. C. Ruiz-Jaen, C. M. Salvador, A. E. Zanne, Functional traits and the growth-
mortality trade-off in tropical trees. Ecology 91, 3664–3674 (2010). doi:10.1890/09-
2335.1 Medline
48. M. Vellend, W. K. Cornwell, K. Magnuson-Ford, A. Ø. Mooers, Measuring phylogenetic
biodiversity, in Biological Diversity: Frontiers in Measurement and Assessment, A. E.
Magurran, B. J. McGill, Eds. (Oxford Univ. Press, 2011), pp. 194–207.
49. O. Purschke, S. G. Michalski, H. Bruelheide, W. Durka, Phylogenetic turnover during
subtropical forest succession across environmental and phylogenetic scales. Ecol. Evol. 7,
11079–11091 (2017). doi:10.1002/ece3.3564 Medline
50. B. Schmid, M. Baruffol, Z. Wang, P. A. Niklaus, A guide to analyzing biodiversity
experiments. J. Plant Ecol. 10, 91–110 (2017). doi:10.1093/jpe/rtw107
51. D. Butler, B. Cullis, A. Gilmour, B. Gogel, Analysis of Mixed Models for S language
Environments: ASReml–R Reference Manual (Queensland DPI, 2007);
www.vsni.co.uk/downloads/asreml/release2/doc/asreml-R.pdf.
52. eFloras (Missouri Botanical Garden, Harvard University Herbaria, 2017); www.efloras.org,
http://frps.eflora.cn (accessed 22 December 2017).
53. Editorial Committee of Flora of Zhejiang, Flora of Zhejiang (Zhejiang Science and
Technology Press, 1993).
54. G. Hu, K. J. Feeley, J. Wu, G. Xu, M. Yu, Determinants of plant species richness and
patterns of nestedness in fragmented landscapes: Evidence from land-bridge islands.
Landsc. Ecol. 26, 1405–1417 (2011). doi:10.1007/s10980-011-9662-7
55. M. Yu, G. Hu, K. J. Feeley, J. Wu, P. Ding, Richness and composition of plants and birds on
land-bridge islands: Effects of island attributes and differential responses of species
groups. J. Biogeogr. 39, 1124–1133 (2012). doi:10.1111/j.1365-2699.2011.02676.x
56. Y. Jin, H. Qian, M. Yu, Phylogenetic structure of tree species across different life stages
from seedlings to canopy trees in a subtropical evergreen broad-leaved forest. PLOS ONE
10, e0131162 (2015). doi:10.1371/journal.pone.0131162 Medline
57. Y. Jin, S. E. Russo, M. Yu, Effects of light and topography on regeneration and coexistence
of evergreen and d