You are on page 1of 13

Received: 27 October 2021 | Revised: 28 February 2023 | Accepted: 20 March 2023

DOI: 10.1111/ele.14220

LETTER

Biodiversity mitigates trade-­offs among species functional traits


underpinning multiple ecosystem services

Emelie Waldén1 | Cibele Queiroz2,3 | Jan Plue1,4 | Regina Lindborg1,5

1
Department of Physical Geography, Abstract
Stockholm University, Stockholm, Sweden
2
Biodiversity loss and its effects on humanity is of major global concern. While
Stockholm Resilience Centre, Stockholm
University, Stockholm, Sweden
a growing body of literature confirms positive relationships between biodiversity
3
Global Resilience Partnership, Stockholm,
and multiple ecological functions, the links between biodiversity, ecological
Sweden functions and multiple ecosystem services is yet unclear. Studies of biodiversity–­
4
IVL Swedish Environmental Institute, functionality relationships are mainly based on computer simulations or controlled
Stockholm, Sweden field experiments using only few species. Here, we use a trait-­based approach to
5
The Bolin Centre for Climate Research, integrate plant functions into an ecosystem service assessment to address impacts
Stockholm University, Stockholm, Sweden
of restoration on species-­r ich grasslands over time. We found trade-­offs among
Correspondence functions and services when analysing contributions from individual species. At
Regina Lindborg, Department of Physical the community level, these trade-­offs disappeared for almost all services with
Geography, Stockholm University, SE-­106
91 Stockholm, Sweden.
time since restoration as an effect of increased species diversity and more evenly
Email: regina.lindborg@natgeo.su.se distributed species. Restoration to enhance biodiversity also in species-­ r ich
communities is therefore essential to secure higher functional redundancy towards
Editor: Akira S Mori disturbances and sustainable provision of multiple ecosystem services over time.

K EY WOR DS
grassland, long-­term in situ experiments, multifunctionality, plants, restoration, species richness

I N T RODUC T ION number of ecosystem functions even at low levels of spe-


cies richness (Flynn et al., 2009; Wardle et al., 1997). The
How to protect biodiversity and manage for multiple eco- contribution of biodiversity to ecosystem service provi-
system services provision in the Anthropocene are top sion is less clear, as links between species richness, eco-
priority tasks for policies worldwide (IPBES, 2019; Maes system functions and services are complex (Harrison
et al., 2014). Reports of decreasing global biodiversity et al., 2014; Lavorel, 2013; Mace et al., 2012). For exam-
(Newbold et al., 2015) have re-­fuelled the debate on how ple, each individual service relies on several ecosystem
biodiversity loss can affect ecosystem functions and ser- functions and the degree of this dependency varies sig-
vices essential for human well-­being. The link between nificantly among services (Cardinale et al., 2012; de Bello
biodiversity and ecosystem functioning has been debated et al., 2010).
in the literature for decades (Cardinale et al., 2012; Mace Although a wide diversity of studies have explored
et al., 2012), and positive relations have been shown in the relationships between biodiversity and ecosystem
a growing number of experiments and data syntheses functioning (BEF), the majority are experimental in
(Hector & Bagchi, 2007; Hooper et al., 2012; Lefcheck situ studies (Balvanera et al., 2014) encompassing tar-
et al., 2015; Tilman et al., 2006). In contrast, other stud- geted species manipulations (Fanin et al., 2018; Finn
ies show species functional redundancy in a community et al., 2013; Meyer et al., 2018; Zavaleta et al., 2010), or
context (Johnson et al., 1996; Loreau, 1998; Naeem, 2002) in silico simulations, including communities with both
and many theoretical models predict saturation of the low (Gamfeldt & Roger, 2017) and high species diversity

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2023 The Authors. Ecology Letters published by John Wiley & Sons Ltd.

Ecology Letters. 2023;26:929–941.  wileyonlinelibrary.com/journal/ele | 929


14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
930 |    BIODIVERSITY ENHANCES ECOSYSTEM SERVICES

(Dee et al., 2017; Solan et al., 2004). Experimental in situ species with traits contributing to particular functions,
and in silico studies are complementary in order to un- should increase the expression of those functions, as well
derstand underlying ecological processes, and these pro- as the generation of underpinned services. Species co-­
cesses might change when studying natural dynamics existing in high-­diversity systems often have traits that
through time (Cardinale et al., 2011; Loreau et al., 2002; are not equally important underpinning the various func-
Manning et al., 2019). The impact of biodiversity may tions, hence complementing each other at the community
also be under-­or overestimated due to the often short-­ level (Hector & Bagchi, 2007). Following habitat resto-
term and/or small-­scale nature of in situ experiments ration, communities usually shift in species composition,
(Cardinale et al., 2012; Duffy, 2009; Isbell et al., 2018; both in terms of species identity and abundance (Carlucci
Manning et al., 2018) and potential risks of missing im- et al., 2020; Jones et al., 2018). Our selected restored sites
portant unknown factors in simulation (Gamfeldt & are subjected to passive restoration, where plant commu-
Roger, 2017). These limitations make it difficult to apply nity recovery relies on natural processes such as spatial
findings from experimental BEF studies to real-­world and temporal seed dispersal, rendering them particularly
ecosystem service generation, as provision of most ser- interesting for analysing temporal relationships between
vices are underpinned by ecological functions at larger community change, biodiversity, ecosystem functions
spatial and temporal scales (Isbell et al., 2017; Lindborg and services.
et al., 2017), although recent work has argued that the In this article, we assess changes in biodiversity and
findings drawn from BEF experiments are generally ro- ecosystem services provision in semi-­natural grasslands
bust (Jochum et al., 2020). over a restoration period of 20 years. Such long-­term in
The link between biodiversity and ecosystem ser- situ studies assessing restoration of species richness and
vices becomes even more complex when considering links to multiple ecosystem services have, to the best of
multiple services. Services can interact in various ways our knowledge, never been conducted. Based on theory
in form of trade-­offs and synergies, and the type and (Hector & Bagchi, 2007; Loreau, 1998), experimental
strength of these relationships may vary over space and studies (Tilman et al., 2006) and modelling (Gamfeldt
time (Lindborg et al., 2017; Queiroz et al., 2015; Ross & Roger, 2017), we expected that an increase in plant
et al., 2021). They can also form interdependent bun- diversity over time would increase the variety of traits
dles, that is, common sets of correlated services (Bennett at the community level underpinning a greater variety
et al., 2009). Understanding these complex relationships of ecological functions. Therefore, higher plant species
is important for production of multiple benefits in a par- richness would, through their functional traits, increase
ticular setting and time frame. While several recent BEF the possibility to generate multiple ecosystem services
meta-­studies found a positive effect of biodiversity when (Cresswell et al., 2018). In a step-­w ise procedure, we first
considering multiple functions (Balvanera et al., 2014; identified five services for which semi-­natural grasslands
Hector & Bagchi, 2007; Isbell et al., 2011), the link be- are important providers: meat production, pollination,
tween biodiversity and multiple services is still unclear water retention, temperature regulation and cultural
(Cardinale et al., 2012; but see Dainese et al., 2019). heritage (Table 1). Second, we coupled a set of relevant
The need for customized management of multiple ser- species functional traits with each of the selected ser-
vices is currently recognized as a vital aspect when re- vices. These different sets of traits were then used to
storing degraded natural habitats (Kollmann et al., 2016; calculate the contribution of each species to individual
McDonald et al., 2016). Since many biodiversity-­rich eco- services. Finally, we paired these individual contribu-
systems are degraded (Chase et al., 2020) and/or face fur- tions with data on community change in the restored
ther degradation due to multiple environmental pressures habitats and compared to reference grassland sites (con-
(Watson et al., 2016), there is an urgent need for long-­term tinuously grazed species-­r ich grasslands), to understand
in situ studies that evaluate the effects of restoration how restoration measures have affected the distribution
measures not only on species richness but also on the re- of multiple services.
covery of multiple ecosystem services in high-­diversity
systems (Balvanera et al., 2014; Fiedler et al., 2018). Here,
we use long-­term restoration data from species-­rich semi-­ M AT E R I A L S A N D M ET HOD S
natural grasslands to examine the relationships between
species diversity, functions and services in plant commu- Study sites and field inventory
nities under non-­experimental conditions. To understand
different plant species' contribution to ecosystem ser- The study was conducted in restored and reference semi-­
vices, we use a trait-­based approach, that is, linking plant natural grasslands situated in multiple agricultural land-
functional traits (Violle et al., 2007) to specific functions scapes in south-­c entral Sweden (Table S2). In terms of
and services (Cresswell et al., 2018; Díaz et al., 2007). biodiversity, traditionally managed semi-­natural grass-
According to the mass ratio theory (Grime, 1998), eco- lands are considered Europe's equivalent to tropical
system functioning is primarily driven by the dominant rainforests (Wilson et al., 2012), meaning high biodiver-
species in a community. Thus, increased abundance of sity systems at small spatial scales. To uphold grassland
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
WALDÉN et al.     | 931

T A B L E 1 Ecosystem services in semi-­natural grasslands coupled to plant species functional effect traits, trait unit measured, data type and
data source (see also Tables S1, S3 and S4). Categorical values were translated to a nominal scale. All traits were then standardized into a 0–­1
scale based on the community maximum. More information regarding data source methods for collecting functional trait data can be found in
each data source reference.

Ecosystem service Functional traits Unit measured in data source Data type Data source
Meat production Leaf dry mass mg Cont. TRY
Specific Leaf Area (SLA) mm 2 mg−1 Cont. TRY
Foraging value Ranging from poisonous to best forage Ord. BiolFlor
value (scale 0–­9)
Pollination Pollination syndrome Insect pollinated or not (scale 0–­1) Bin. BiolFlor
Flowering duration No. of months Ord. Den nya nordiska floran
Nectar quantity Ranging from none to plenty (scale 0–­3) Ord. BiolFlor, Imkerbond
Zoersel
Pollen quantity Ranging from none to plenty (scale 0–­5) Ord. Imkerbond Zoersel,
BiolFlor
Temperature Height m Cont. TRY
regulation SLA mm mg2 −1
Cont. TRY
Root architecture type Tap root, adventitious, fibrous (scale 1–­3) Cat. TRY
Lifespan Annual, biennial, 2–­5 years, 5–­50 years and Ord. LEDA
>50 years (scale 1–­5)
Water retention SLA mm 2 mg−1 Cont. TRY
Height m Cont. TRY
Clonal lateral spread rate Non- ­clonal, <0.01, 0.01–­0.25 and Ord. CLO-­PLA
>0.25 m year−1 (scale 0–­3)
Root architecture type Tap root, adventitious, fibrous (scale 1–­3) Cat. TRY
Cultural heritage Grassland specialist Specialist or not (scale 0–­1) Bin. Om hävden upphör
Mentions in traditional music No. of songs Cont. Svenskt Visarkiv
and older popular songs
Abbreviations: Bin., Binary variable; Cat., Categorical value; Cont., Continuous variable; Ord., Ordinal value.

biodiversity, traditional management by grazing or number of units studied (i.e. frequency), or as biomass.
mowing, without the use of pesticides or fertilizers, is In this study, we used the frequency measure based on
necessary. However, as a consequence of agricultural in- presence–­absence in 10 plots per grassland, scale 0–­10
tensification, up to 90% of these habitats have been lost (hereafter referred to as abundance).
during the last century (Strijker, 2005). Restoration of
semi-­natural grasslands and other degraded or damaged
habitats is highlighted as a key feature in the European Ecosystem services selection and trait
Union Biodiversity Strategy for 2030. data collection
We used data from plant inventories of 16 restored
semi-­natural grasslands collected at two time steps; The number of studies addressing ecosystem services (ES)
1–­9 years (T1) and 12–­20 years (T2) post-­restoration, in from grasslands (natural, semi-­ natural or improved) is
a total of 260 species. The total number of species in- generally low compared to other ecosystems like, for exam-
creased significantly between T1 and T2 in the restored ple, forests (Bardgett et al., 2021). However, there are more
grasslands, from 51.6 to 62.9 species per grassland site than 20 different services acknowledged from grasslands
in average, although the average number of species related to fodder production, water, soil health, carbon
per grassland was still significantly higher in the refer- sequestration, habitat provision, pollination and biologi-
ence grasslands (76.2 species) (Figure S1; Waldén and cal control (Bengtsson et al., 2019; Zhao et al., 2020). We
Lindborg (2016)). Vascular plant species richness and selected five of these: ‘meat production’, ‘pollination’,
individual species abundance was inventoried using 10 ‘temperature regulation’, ‘water retention’ and ‘cultural
plots (1 m × 1 m) equally distributed in two transects per heritage’ (Table 1) (for references, see Table S1), as (1) they
grassland. Abundance in its broader sense can be mea- are the most strongly related to traditionally managed
sured in different ways, either as the total number of an semi-­natural grasslands in agricultural landscapes, and (2)
individual species per unit area at which data are col- their underpinning of ecological functions is, through lit-
lected (i.e. abundance/density), as the number of unit erature, linked to specific plant traits. Since these habitats
areas in which the species occurred in relation to total are grazed by livestock, the grassland's fodder production
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
932 |    BIODIVERSITY ENHANCES ECOSYSTEM SERVICES

quality is important for meat production. Grasslands are All traits were then standardized into a 0–­1 scale
also able to act as a food resource for crop-­pollinating in- (based on the community maximum), where higher
sects, through its diverse supply of flowering plants dur- value equals higher contribution to the ES. Correlation
ing the season. Being a permanent semi-­natural habitat in between traits within each ES trait group was tested in
agricultural landscapes, grasslands may also sequestrate multiple Spearman correlation tests (Table S5). Most
carbon and prevent flooding in such landscapes. Finally, traits were uncorrelated, however, for example, polli-
through its historical origin, semi-­natural grasslands are nation syndrome and nectar/pollen quantity were cor-
important carriers of cultural heritage. related (ρSpearman > 0.5 at p < 0.001). Nevertheless, due to
The selection of plant traits is fundamental, as dif- the ES calculation method described below and com-
ferent traits play different roles in the ecological func- parably low correlations between other traits, all traits
tions underpinning services (Cresswell et al., 2018). In were included during data preparation. For community
this study, we assessed the services indirectly by assign- average trait and ES-­values, see Table S6.
ing scores to each species based on the extent to which
species-­level functional traits contribute to providing ES.
This is a commonly used approach when measuring mul- Data preparation to estimate ecosystem services
tiple services in species-­rich systems (Díaz et al., 2007; from species traits
Zhao et al., 2020). For each ES, we selected a set of func-
tional traits (Table 1) that were identified in the literature The five selected ES were treated separately in a step-­wise
as being important for the ecological functions under- procedure (Figure 1). After selecting functional traits for
pinning that specific ES (Table S1). Another important each ES and collecting trait data for each species, species-­
criterion for trait selection was whether trait data were specific ES-­values was calculated, that is, each species
publicly available and accessible for most species in our received a separate ES value which reflected the species'
study. In a few cases, other and/or additional traits would contribution to the provision of each service; ‘meat pro-
have been preferred than those selected, but they could duction’, ‘pollination’, ‘temperature regulation’, ‘water re-
not be included due to high levels of missing data. In some tention’ and ‘cultural heritage’. In theory, a typical plant
cases, especially for temperature regulation and water re- with high ES value for, (1) ‘meat production’ would have
tention, many of the traits identified in the literature as a high forage value, high leaf dry mass and high SLA, (2),
important to contribute to those particular ES were the ‘pollination’ would be insect pollinated, have long flower
same. This is a consequence of the fact that some ES are duration and plenty of nectar and pollen, (3) ‘cultural her-
underpinned by similar ecological processes, traits and itage’ would be a grassland specialist plant and be men-
species (Bennett et al., 2009; Rullens et al., 2019), influ- tioned frequently in traditional music, (4) ‘temperature
encing the correlation among ES, found in our analysis. regulation’ would be tall and long-­lived plants with high
All final traits were carefully selected to match the ES (for SLA and fibrous roots, and (5) ‘water retention’ would be
description of trait selection and references, see Tables S1 tall plants with a high clonal lateral spread rate, high SLA
and S4). The traits were classified so that higher values and fibrous roots (Table 1). Due to the resource limita-
corresponded to increased ES provision. tions and evolutionary history, these theoretical plants
We calculated the mean trait values for each species might not exist in reality. Further, it is also clear that traits
across the different databases for all continuous vari- are not equally important and do not scale linearly with
ables. Instead of the arithmetic mean, we used the geo- ES provision (Lindborg et al., 2017). Nevertheless, as no
metric mean as it is less sensitive to extreme values, and data exist of relative trait contribution to ES generation,
we were interested in getting a representative value for we chose to treat them similarly.
each species that was fairly robust to potential outliers. For each species, the average of all trait values in the
The geometric mean is given by the ‘n’ root of the mul- specific ES trait group were then calculated to receive a
tiplication of all n values, where ‘n’ is the total number species-­specific ES-­value. For the ES ‘meat production’
of observations. Calculations were done in R 3.3.0 (R and ‘pollination’, this was treated differently. For ‘meat
Core Team, 2016) with the EnvStats package (function: production’, species that are poisonous for cattle and
geoMean). To complement missing trait values, informa- therefore cannot be used as fodder (zero as forage value)
tion from alternative databases and the Swedish national were automatically set to an overall zero-­value for the
flora were added as a first measure. For the species still ‘meat production’ ES. Similarly, species not pollinated by
lacking data for some of the traits (foraging value, nectar insects were also set to a zero-­value for ‘pollination’ ES.
and pollen quantity), missing values were at first hand For each of the five ES, the species ES values was
replaced by average values from species within the same then multiplied with the species' occurrence at each
taxonomic genus present in Sweden, and secondly, by grassland site and time step (Rest. sites T1, Rest. sites
species pool average values. For more information re- T2, Ref. sites) resulting in species ES contribution values
garding percentages of missing values in primary data for each site and time step. These species ES contribu-
sources for each of the traits and which alternative data tion values were calculated in two ways: (1) just based
sources that were used, see Table S4. on species presence–­absence data, and (2) each species
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
WALDÉN et al.     | 933

F I G U R E 1 To estimate ecosystem service provision through species and trait data, a five-­step procedure was done for each of the five
ecosystem services (ES) separately. [1] Selecting functional traits coupled to each specific ES, according to available literature. [2] Standardizing
the traits coupled to the specific ES into a 0–­1 scale (based on the community maximum), where higher value equals higher contribution to the
ES (in this figure, we use ES 1 as an example of one of the five ES). [3] Calculating an ES value for each species in the community, by averaging
all the associated trait values for that species. [4] Multiplying the species ES value with either the species presence, or their abundance, at each
grassland site (restored and reference) and time step (T1, T2) (in this figure, we use species abundance at restored sites at T1 as an example).
[5] Using the final site-­specific species ES contribution matrix to calculate an ES community average for each ES and site to analyse the
community differences in t-­test.

contribution to a particular service weighted by its abun- species-­specific ES values (n = 260 × 5) were analysed in
dance. This was done to disentangle how species identity multiple Spearman's rank correlations (function: cor.
versus abundance affects ecosystem services distribu- test). The differences in community ES metrics (i.e. based
tion. A presence–­absence as well as an abundance-­based on either species presence-­absence or abundance data)
species ES contribution matrix was generated for each for the restored sites at T1 compared to T2 were analysed
time step (Rest. sites T1, Rest. sites T2, Ref. sites) result- using paired t-­tests (function: t-­test). Similarly, the dif-
ing in six species ES contribution matrices for every ES ferences in community ES metrics between restored and
(i.e. 30 matrices in total). These were used as the base reference sites at T2 were analysed using Welch t-­tests
for calculating a community-­weighted mean for each ES (function: t-­test). To assess the extent to which changes in
and site (ES community average), to be able to analyse community composition are driving changes in commu-
the community differences between T1 and T2, as well nity ES, we ran a non-­metric multidimensional scaling
as between restored and reference sites. ordination (NMDS; function metaMDS, vegan package)
for each ES including both restored and reference sites.
The species abundance × site matrix underlying each
Statistical analyses NMDS ordination was adapted per ES, using only the
species contributing to a particular ES. Each ordina-
All statistical analyses were performed using R 3.3.0 tion solution was constrained to returning only a two-­
(R Core Team, 2016). Relationships between the axes solution based on the Bray–­ Curtis dissimilarity
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
934 |    BIODIVERSITY ENHANCES ECOSYSTEM SERVICES

metric, which was rotated to ensure that the first NMDS At the community level, plant presence/absence
axis contained the majority of variation in the com- data showed that community ecosystem service provi-
munity data. Post hoc analysis to assess differences sion from restored grassland sites (Table S7) increased
in the positioning of the restored and reference sites with time towards reference levels for all five services
in the ordination space (using the t-­tests as described (Figure 3, top panel). Specifically, the generation of
above) was therefore restricted to the first NMDS axis. ‘meat production’ and ‘pollination’, in the restored hab-
Relationships between the different community ES met- itats (at T2) now resembled the reference semi-­natural
rics, as well as their relation to species richness, were grasslands. Species richness in the restored sites also in-
analysed using multiple Spearman correlations for each creased over time (average increase of 21.9%, t = −4.50,
category (T1, T2 and the relative change over time [∆]) df = 15, p < 0.001; see also Figure S1).
respectively. Correlations between pairs of ES were used Species community evenness increased over time
to investigate potential trade-­offs and synergies, where (Simpson's Index of Diversity 1 − D, Restored T1:
trade-­offs were given by negative correlations and syner- 0.967 ± 0.004 vs. T2: 0.974 ± 0.003, t = −3.85, df = 15,
gies by positive correlations between services. To ana- p-­value = 0.002, see Tables S7 and S8). However, when
lyse community evenness shift between the time steps, species abundance was included in the calculations,
we first calculated Simpson's Index of Diversity (1 − D) we found no increase in ecosystem service provi-
for all sites (package: vegan, function: diversity), and then sion over time in the restored sites (Figure 3, middle
analysed the result in a paired t-­test. panel). We also found community changes in those
species' assemblages contributing to the respective
ecosystem services (Figure 3, bottom panel).
R E SU LT S Positive correlations between all five services were de-
tected, both in the short-­and long-­term after restoration,
The contribution to the generation of selected ES dif- when analysed pair-­wise (Figure 4). For data of propor-
fered between plant species, highlighting that one sin- tional change for each grassland site, see Table S8. All
gle species cannot simultaneously have high values for these individual services also became correlated with spe-
all the traits underpinning multiple ES. This resulted in cies richness over time (compare T1 and T2 in Figure 4).
trade-­offs between ES on a species level (Figure 2). These
trade-­offs occur as groups of traits that enhance the pro-
vision of certain services, do not enhance the provision DISCUSSION
of other services as they are underpinning different eco-
system functions. Trade-­offs were mainly found between Ecosystem service trade-­offs generated at the
‘temperature regulation’ and ‘water retention’ in relation species level
to ‘cultural heritage’ and ‘pollination’. Different species
also contributed to ‘meat production’ and ‘cultural herit- Ecosystem service trade-­offs as an effect of land use and
age’, respectively, leading to trade-­offs between the ES. management (Bennett et al., 2009; Foley et al., 2005)
In contrast, the same species contributed to the services have been frequently discussed in the scientific lit-
‘temperature regulation’, ‘water retention’ and ‘meat erature. These trade-­offs are often related to biodiver-
production’. sity composition as some land uses and management

F I G U R E 2 Species-­level trade-­offs, synergies and non-­significant relationships between ecosystem services provided by plant species
occurring in the semi-­natural grasslands. Individual species ES values are calculated based on their functional traits (see calculation procedure
in Table 1 and Figure 1). Significant Spearman's rank correlations indicated by colour gradient (red is negative, blue is positive), Spearman's ρ
and p-­value (inside parenthesis).
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
WALDÉN et al.     | 935

F I G U R E 3 Ecosystem service (ES) provision (plant community average) for each ES; ‘meat production’, ‘pollination’, ‘temperature
regulation’, ‘water retention’ and ‘cultural heritage’. The ES community average is based on plant species traits connected to the specific
ecosystem functions and services, and the species' presence (top panel) or abundance (middle panel) in restored semi-­natural grasslands at
time 1 (T1, 1–­9 years after restoration) and time 2 (T2, 12–­20 years after restoration), as well as in intact reference semi-­natural grasslands
(Ref) at time 2. The bottom panel represent plant community change by the contributing part of the community following restoration (at T1,
T2 and Ref), represented by the position along the first axis of a non-­metric multi-­d imensional scaling ordination performed on a species
abundance × site matrix (ordinations, Figure S2). Significant differences indicated by asterisks (*p < 0.05–­0.01, **p < 0.01–­0.001, ***p < 0.001).

practices favours some species more than others pollination, are mostly relatively small and thick-­leaved,
(Cardinale et al., 2012; Mace et al., 2012). This study with poorly developed root architecture, making them
found clear trade-­offs among ecosystem services when contribute little to water retention. Some of the trade-­
calculations were made based on the relative contribu- offs found are relatively well-­k nown trade-­offs to occur
tion of each plant species in the grassland community between provisioning services and regulating or cultural
(Figure 2). This suggests that high contribution to some services (Foley et al., 2005). These trade-­offs occur when
services goes along with low contribution to the provi- two services are underpinned by different functions, and
sioning of others. For example, plant species that have dependent on a distinct set of traits given by different
nectar-­r ich flowers and therefore contribute strongly to species (Bennett et al., 2009).
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
936 |    BIODIVERSITY ENHANCES ECOSYSTEM SERVICES

F I G U R E 4 Correlations between five ecosystem services at a plant community level and in relation to species richness in restored semi-­
natural grasslands, at time 1 and 2 (T1: 1–­9 years after restoration, T2: 12–­20 years after restoration) and the relative change over time (Δ).
Provision of the ecosystem services is based on species abundance (see calculation procedure in Figure 1 and Table 1). Positive significant
relationship indicated by blue colour (reporting Spearman's ρ and p-­value in parenthesis).

In other words, trade-­offs and non-­significant rela- with ‘cultural heritage’ and ‘pollination’ on the other.
tionships between service provision show that differ- This supports the findings from previous studies on
ent species, through their functions, underpin different the existence of particular ES bundles, meaning that
services. Furthermore, no single species is able to con- some ES are often positively correlated (as they are un-
tribute to high levels of all the tested services simulta- derpinned by functions performed by the same species)
neously. Similar to Winfree et al. (2015) and Dainese while others are never (Raudsepp-­Hearne et al., 2010;
et al. (2019), we show that individual species can have a Queiroz et al., 2015).
disproportionate large positive effect on the provision To fully understand the role of biodiversity in gener-
of one service, although high levels of species richness ating ecosystem services (Mace et al., 2012), there is a
are needed to generate multiple services (Gamfeldt need to identify not only trade-­offs but also synergies
et al., 2013). This becomes even more evident over among services. In our study, we found positive rela-
time, as high species richness may enhance ecosystem tionships among the services ‘temperature regulation’,
functioning across temporally fluctuating and spa- ‘water retention’ and ‘meat production’. The synergies
tially heterogeneous environments (Isbell et al., 2018). between ‘meat production’, ‘water retention’ and ‘tem-
Trade-­offs between groups of services could also be de- perature regulation’ are mostly an effect of high SLA
tected in our study. For example, ‘water retention’ and values, whereas the stronger synergy between ‘tempera-
‘temperature regulation’ on the one hand, traded-­off ture regulation’ and ‘water retention’ relates to a plant
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
WALDÉN et al.     | 937

being tall, perennial and having a finer root architec- When adding species abundance, no increase in ecosys-
ture. These synergies indicate that those ES are under- tem service provision over time was detected in restored
pinned by functions that are given by the similar groups sites. This suggests that species important for the pro-
of traits. The knowledge that similar traits can underpin vision of multiple services may have so far remained at
two different services are important to fully understand low abundances compared to reference sites. Thus, not
bundles of ES and how to manage them. This highlights only does a species important for a certain service need
the importance of traits selected for the outcome of anal- to be present, but it also needs to be abundant enough
ysis linking traits to services (de Bello et al., 2021; Mace to actually contribute to the expression of that service in
et al., 2012). the community. Shortly after restoration (T1), ecosystem
services' generation was mainly provided by a few abun-
dant species (illustrated in Figure S1) diverse enough in
Ecosystem service generation changes over time their functional character to ensure the generation of
multiple services.
As expected, all the five services selected for this study There was a clear shift in species community assem-
increased with time towards reference levels (Figure 3). blage in restored sites from T1 to T2, with an increase
This was especially clear in the habitats 12–­20 years in community evenness. Yet, generation of ecosystem
after restoration (T2), where the generation of ‘meat services remained stable. This implies that service pro-
production’ and ‘pollination’ now resembled service vision turned from relying on a few abundant species to
generation in reference semi-­ natural grasslands. As relying on more species for upholding the same provi-
species community composition often changes over sion level. By relying on a higher number of species, the
time, particularly in response to restoration, shifts in provision of each ecosystem service is likely to be more
ecosystem service provision can also be expected. In resilient to potential future natural and non-­natural
theory, increasing numbers of co-­existing species within disturbances. When several species are contributing
a community would increase the probability of includ- to the same function, some species can compensate
ing species with high contribution to functions and ser- for the potential loss of other species (functional re-
vices, due to an overall increased functional richness dundancy), especially when the diversity of responses
(Cadotte et al., 2011). This is confirmed by our results, to potential disturbances among those species is high
where species richness in the restored sites increased (response diversity) (Elmqvist et al., 2003). While this
over time, simultaneously underpinning a higher num- remains a reasonable hypothesis anchored in ecologi-
ber of ecosystem services (Figure 3). However, these cal theory, the extension of Price equation partitioning
results should be interpreted carefully, as one of the developed by Fox and Kerr (2012) could be applied in
assumptions of this study was that all traits are equally future research to quantify the effects on a function
important for the provision of a particular ES and that of biodiversity changes (species loss/gain) and contex-
trait values scale linearly with ES provision, which is tual changes between two given communities. This
likely not the case for all ES (Lindborg et al., 2017). would enable a detailed exploration to understand how
Some studies have found idiosyncratic species behav- ecosystem service generation remained stable in spite
iour related to, for example, community composition of individual changes in species presences and abun-
or temperature, partly questioning grouping species dances. Such an approach would unveil in detail the
into functional groups mainly based on their traits dominant trait-­and community-­ based mechanisms
(Lindborg & Eriksson, 2005). Hence, assuming traits driving shifts in ecosystem service provisioning in real-­
to be context-­independent is partly a simplification. world high-­d iversity plant communities pressurized by
However, if functional groups of species are remaining environmental changes. Finally, higher diversity also
relatively constant, at least within similar systems, it is increases the possibility for higher trait diversity that
reasonable to believe that specific groups of species will would more efficiently counteract perturbations, al-
underpin similar functions in many different environ- together reflecting the links between species diversity
ments (Díaz et al., 2007). Further investigation of the and functional redundancy in the community (de Bello
nature and shape of these trait–­ES relationships, and et al., 2021).
if certain traits are more important than others should
be the focus of future studies (e.g. de Bello et al., 2021).
In studies concerning restoration, it is especially im- Biodiversity increases provision of multiple
portant to incorporate abundance of species since that ecosystem services
tends to shift substantially after restoration (Waldén &
Lindborg, 2016). Since the relative abundance of spe- Finding ways of increasing potential synergies and re-
cies are important to the different functions, it might ducing trade-­offs between ecosystem services is impor-
affect ecosystem service provision (Isbell et al., 2017). tant for future ecosystem management. When analysing
To investigate the importance of species abundance, we the overall contribution of the whole plant commu-
also included abundance (frequency) in the calculations. nity to the provision of ecosystem services, we found
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
938 |    BIODIVERSITY ENHANCES ECOSYSTEM SERVICES

strong positive correlations between all five services, ecosystem service provision to biodiversity presents ad-
both short and long time after restoration (Figure 4). ditional arguments for conservation and restoration to
Interestingly, all individual services became highly cor- reach sustainability in a changing world.
related with species richness over time. In comparison
with our study, increasing positive effects of biodiver- AU T HOR C ON T R I BU T ION S
sity have also been detected in BEF experiments over Emelie Waldén came up with some of the design, con-
time (Cardinale et al., 2007; Isbell et al., 2018), stressing ducted the fieldwork, compiled the data, made data anal-
a clear link between traits, ecosystem functioning and ysis and compiled the statistics, and performed most of
ecosystem services (Carlucci et al., 2020). Even though the writing; Cibele Queiroz helped with data compiling,
the communities studied here are characterized by analysis, design and writing; Jan Plue helped with analy-
very high levels of plant diversity, we show that further sis and statistics; and Regina Lindborg came up with the
increase in species richness correlated significantly original idea, design and performed some of the writing.
with increased provision of multiple services. This
is an important finding in relation to most previous AC K NO​W L E​D GE​M E N T S
studies showing a positive relationship between spe- We thank landowners and farmers for letting us col-
cies richness and functional diversity which are mainly lect plant data on their land and two anonymous re-
conducted in low-­d iversity experimental communities viewers for their helpful comments on the article. We
under controlled conditions (Manning et al., 2019). are also grateful to the publicly available plant trait
Our results suggest that this positive relationship thus databases TRY, BiolFlor, LEDA, CLO-­ PLA and
applies even to real-­world high-­d iversity plant commu- Imkerbond Zoersel, and to The Centre for Swedish
nities. Habitat management focusing mainly on biodi- Folk Music and Jazz Research for providing access to
versity could therefore simultaneously support a range the online song and tune catalogue ‘Svenskt Visarkiv’.
of different services including provisioning, regulating We also acknowledge the Swedish Research Council
as well as cultural services. for Environment, Agricultural Sciences and Spatial
Humans depend on nature and its capacity to pro- Planning (FORMAS, grant numbers 2009-­ 1105 and
vide multiple ecosystem services across space and time, 2018-­0 0961) and the Swedish Marianne and Marcus
for their well-­being and survival (Cardinale et al., 2012; Wallenberg Foundation (Grant 2017.0137) for funding
IPBES, 2019). Achieving simultaneous provision of of this research.
multiple services is difficult, as there are fundamental
trade-­offs between functional traits underpinning these C ON F L IC T OF I N T E R E ST STAT E M E N T
services. Here, we show that although these trade-­offs None of the authors have competing interest with this
exist and can be mitigated by managing for increased article.
species richness. As shown by this study, plant diversity
can play an important role in generation of services, even PE E R R E V I E W
in species-­r ich real-­world ecosystems. So far, most stud- The peer review history for this article is available at
ies have focused on biodiversity loss and the following https://www.webof​s cien​c e.com/api/gatew​ay/wos/peer-­
negative impact on functions and services. In contrast, revie​w/10.1111/ele.14220.
we highlight the multiple benefits of active manage-
ment towards increasing species richness in a commu- DATA AVA I L A B I L I T Y STAT E M E N T
nity through restoration (see also Carlucci et al., 2020). We confirm that if the article should be accepted, the
Relying on a few key species might be sufficient when data supporting the results will be archived in the public
focusing on a single service at a specific time and place repository Figshare and the data DOI will be included
(Winfree et al., 2015). That will, however, not be enough at the end of the article. https://doi.org/10.6084/m9.figsh​
to ensure a long-­ term stability of ecosystem services are.21908​964.v4
provision, as the environmental conditions will become
increasingly uncertain under future climate change and ORC I D
other drivers of environmental change. Dade et al. (2019) Jan Plue https://orcid.org/0000-0002-6999-669X
conclude that to fully understand the relation between Regina Lindborg https://orcid.org/0000-0001-​7134-7974
biodiversity and ecosystem services and manage services
over time, assessments should focus on casual drivers and R EF ER ENCE S
process-­based models. Our observed increase in plant Balvanera, P., Siddique, I., Dee, L., Paquette, A., Isbell, F.,
species richness, led to a higher functional redundancy Gonzalez, A. et al. (2014) Linking biodiversity and ecosys-
tem services: current uncertainties and the necessary next
over time, indicating a more resilient service provision
steps. Bioscience, 64(1), 49–­ 57. Available from: https://doi.
with time since restoration. Still, the full implications of org/10.1093/biosc​i /bit003
our findings for ecosystem services resilience should be Bardgett, R.D., Bullock, J., Lavorel, S., Manning, P., Schaffner, U.,
further studied and also include other indicators than just Ostle, N. et al. (2021) Combating global grassland degradation.
functional redundancy of plants. Nevertheless, coupling Nature Reviews Earth & Environment, 2, 720–­735 Available from:
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
WALDÉN et al.     | 939

https://www.nature.com/artic​les/s4301​7- ­021- ­0 0207​-­2 [Accessed Duffy, J.E. (2009) Why biodiversity is important to the functioning of
23rd January 2022]. real-­world ecosystems. Frontiers in Ecology and the Environment,
Bengtsson, J., Bullock, J.M., Egoh, B., Everson, C., Everson, T. & 7(8), 437–­4 44. Available from: https://doi.org/10.1890/070195
O’Connor, T. (2019) Grasslands—­more important for ecosys- Elmqvist, T., Folke, C., Nyström, M., Peterson, G., Bengtsson, J.,
tem services than you might think. Ecosphere. Available from: Walker, B. et al. (2003) Response diversity, ecosystem change,
https://doi.org/10.1002/ecs2.2582 and resilience. Frontiers in Ecology and the Environment,
Bennett, E.M., Peterson, G.D. & Gordon, L.J. (2009) Understanding 1, 488–­ 494. Available from: https://doi.org/10.1890/1540-­
relationships among multiple ecosystem services. Ecology 9295(2003)001[0488:RDECA​R]2.0.CO;2
Letters, 12(12), 1394–­ 1404. Available from: https://doi. Fanin, N., Gundale, M.J., Farrell, M., Ciobanu, M., Baldock, J.A.,
org/10.1111/j.1461-­0248.2009.01387.x Nilsson, M.-­C. et al. (2018) Consistent effects of biodiversity loss
Cadotte, M.W., Carscadden, K. & Mirotchnick, N. (2011) Beyond spe- on multifunctionality across contrasting ecosystems. Nature
cies: functional diversity and the maintenance of ecological pro- Ecology & Evolution, 2(2), 269–­278. Available from: https://doi.
cesses and services. Journal of Applied Ecology, 48(5), 1079–­1087. org/10.1038/s4155​9-­017-­0415-­0
Available from: https://doi.org/10.1111/j.1365-­2664.2011.02048.x Fiedler, S., Perring, M.P. & Tietjen, B. (2018) Integrating trait-­based
Cardinale, B.J., Matulich, K.L., Hooper, D.U., Byrnes, J.E., Duffy, empirical and modeling research to improve ecological resto-
E., Gamfeldt, L. et al. (2011) The functional role of producer di- ration. Ecology and Evolution, 8, 6369–­ 6380. Available from:
versity in ecosystems. American Journal of Botany, 98(3), 572–­ https://doi.org/10.1002/ece3.4043
592. Available from: https://doi.org/10.3732/ajb.1000364 Finn, J.A., Kirwan, L., John Connolly, M., Sebastià, T., Helgadottir,
Cardinale, B.J., Emmett Duffy, J., Gonzalez, A., Hooper, D.U., A., Baadshaug, O.H. et al. (2013) Ecosystem function enhanced
Perrings, C., Venail, P. et al. (2012) Biodiversity loss and its im- by combining four functional types of plant species in intensively
pact on humanity. Nature, 486(7401), 59–­ 67. Available from: managed grassland mixtures: a 3-­year continental-­s cale field ex-
https://doi.org/10.1038/natur​e11148 periment. Journal of Applied Ecology, 50(2), 365–­375. Available
Cardinale, B.J., Wright, J.P., Cadotte, M.W., Carroll, I.T., Hector, from: https://doi.org/10.1111/1365-­2664.12041
A., Srivastava, D.S. et al. (2007) Impacts of plant diversity on Flynn, D.F.B., Gogol-­ Prokurat, M., Nogeire, T., Molinari, N.,
biomass production increase through time because of spe-
­ Richers, B.T., Lin, B.B. et al. (2009) Loss of functional di-
cies complementarity. Proceedings of the National Academy of versity under land use intensification across multiple taxa.
Sciences of the United States of America, 104(46), 18123–­18128 Ecology Letters, 12(1), 22–­ 33. Available from: https://doi.
Available from: http://www.pnas.org/conte​nt/104/46/18123.ab- org/10.1111/j.1461-­0248.2008.01255.x
stract [Accessed 20th September 2019]. Foley, J.A., Defries, R., Asner, G.P., Barford, C., Bonan, G., Carpenter,
Carlucci, M.B., Brancalion, P.H.S., Rodrigues, R.R., Loyola, R. & S.R. et al. (2005) Global consequences of land use. Science,
Cianciaruso, M.V. (2020) Functional traits and ecosystem ser- 309(5734), 570–­574. Available from: https://doi.org/10.1126/scien​
vices in ecological restoration. Restoration Ecology, 28, 1372–­ ce.1111772
1383. Available from: https://doi.org/10.1111/rec.13279 Fox, J.W. & Kerr, B. (2012) Analyzing the effects of species gain and
Chase, J.M., Blowes, S.A., Knight, T.M., Gerstner, K. & May, F. loss on ecosystem function using the extended Price equation
(2020) Ecosystem decay exacerbates biodiversity loss with habi- partition. Oikos, 121(2), 290–­ 298. Available from: https://doi.
tat loss. Nature, 584(7820), 238–­243. Available from: https://doi. org/10.1111/j.1600-­0706.2011.19656.x
org/10.1038/s4158​6 -­020-­2531-­2 Gamfeldt, L. & Roger, F. (2017) Revisiting the biodiversity–­e cosystem
Cresswell, C.J., Cunningham, H.M., Wilcox, A. & Randall, N.P. multifunctionality relationship. Nature Ecology & Evolution, 1(7),
(2018) What specific plant traits support ecosystem services such 0168. Available from: https://doi.org/10.1038/s4155​9-­017-­0168
as pollination, bio-­c ontrol and water quality protection in tem- Gamfeldt, L., Snäll, T., Bagchi, R., Jonsson, M., Gustafsson, L.,
perate climates? A systematic map. Environmental Evidence, 7(1), Kjellander, P. et al. (2013) Higher levels of multiple ecosystem
2. Available from: https://doi.org/10.1186/s1375​0 -­018-­0120-­8 services are found in forests with more tree species. Nature
Dade, M.C., Mitchell, M.G.E., McAlpine, C.A. & Rhodes, J.R. (2019) Communications, 4(January), 1340. Available from: https://doi.
Assessing ecosystem service trade-­offs and synergies: the need org/10.1038/ncomm​s2328
for a more mechanistic approach. Ambio, 48(10), 1116–­ 1128. Grime, J.P. (1998) Benefits of plant diversity to ecosystems: immedi-
Available from: https://doi.org/10.1007/s1328​0 -­018-­1127-­7 ate, filter and founder effects. Journal of Ecology, 86, 902–­910.
Dainese, M., Martin, E.A., Aizen, M.A., Albrecht, M., Bartomeus, I., Available from: https://doi.org/10.2307/2648655
Bommarco, R. et al. (2019) A global synthesis reveals biodiversity-­ Harrison, P.A., Berry, P.M., Simpson, G., Haslett, J.R., Blicharska,
mediated benefits for crop production. Science Advances, 5(10), M., Bucur, M. et al. (2014) Linkages between biodiversity attri-
1–­13. Available from: https://doi.org/10.1126/sciadv.aax0121 butes and ecosystem services: a systematic review. Ecosystem
de Bello, F., Lavorel, S., Gerhold, P., Reier, Ü. & Pärtel, M. (2010) A Services, 9(September), 191–­ 203. Available from: https://doi.
biodiversity monitoring framework for practical conservation of org/10.1016/j.ecoser.2014.05.006
grasslands and shrublands. Biological Conservation, 143(1), 9–­17. Hector, A. & Bagchi, R. (2007) Biodiversity and ecosystem multifunc-
Available from: https://doi.org/10.1016/j.biocon.2009.04.022 tionality. Nature, 448(7150), 188–­190. Available from: https://doi.
de Bello, F., Lavorel, S., Hallett, L.M., Valencia, E., Garnier, E., org/10.1038/natur​e 05947
Roscher, C. et al. (2021) Functional trait effects on ecosys- Hooper, D.U., Carol Adair, E., Cardinale, B.J., Byrnes, J.E.K.,
tem stability: assembling the jigsaw puzzle. Trends in Ecology Hungate, B.A., Matulich, K.L. et al. (2012) A global synthe-
& Evolution, 36(9), 822–­ 836. Available from: https://doi. sis reveals biodiversity loss as a major driver of ecosystem
org/10.1016/j.tree.2021.05.001 change. Nature, 486(7401), 105–­108. Available from: https://doi.
Dee, L.E., De Lara, M., Costello, C. & Gaines, S.D. (2017) To what org/10.1038/natur​e11118
extent can ecosystem services motivate protecting biodiversity? IPBES. (2019) Global assessment report on biodiversity and ecosystem
Ecology Letters, 20(8), 935–­ 946. Available from: https://doi. Services of the Intergovernmental Science-­Policy Platform on biodi-
org/10.1111/ele.12790 versity and ecosystem services. Bonn, Germany: IPBES Secretariat.
Díaz, S., Lavorel, S., de Bello, F., Quétier, F., Grigulis, K., Matthew, T. Available from: https://doi.org/10.5281/zenodo.3831673
et al. (2007) Incorporating plant functional diversity effects in eco- Isbell, F., Calcagno, V., Hector, A., John Connolly, W., Harpole, S.,
system service assessments. Proceedings of the National Academy Reich, P.B. et al. (2011) High plant diversity is needed to maintain
of Sciences of the United States of America, 104(52), 20684–­20689. ecosystem services. Nature, 477(7363), 199–­202. Available from:
Available from: https://doi.org/10.1073/pnas.07047​16104 https://doi.org/10.1038/natur​e10282
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
940 |    BIODIVERSITY ENHANCES ECOSYSTEM SERVICES

Isbell, F., Cowles, J., Dee, L.E., Loreau, M., Reich, P.B., Gonzalez, McDonald, T., Gann, G.D., Jonson, J. & Dixon, K.W. (2016)
A. et al. (2018) Quantifying effects of biodiversity on ecosystem International standards for the practice of ecological resto-
functioning across times and places. Ecology Letters, 21, 763–­ ration –­including principles and key concepts. Washington,
778. Available from: https://doi.org/10.1111/ele.12928 DC. Available from: www.ser.org [Accessed 6th May 2019]:
Isbell, F., Gonzalez, A., Loreau, M., Cowles, J., Díaz, S., Hector, A. National standards for the practice of ecological restoration
et al. (2017) Linking the influence and dependence of people on in Australia.
biodiversity across scales. Nature, 546(7656), 65–­72. Available Meyer, S.T., Ptacnik, R., Hillebrand, H., Bessler, H., Buchmann, N.,
from: https://doi.org/10.1038/natur​e22899 Ebeling, A. et al. (2018) Biodiversity–­ multifunctionality rela-
Jochum, M., Fischer, M., Isbell, F., Roscher, C., van der Plas, F., tionships depend on identity and number of measured functions.
Boch, S. et al. (2020) The results of biodiversity–­e cosystem func- Nature Ecology & Evolution, 2(November), 44–­ 49. Available
tioning experiments are realistic. Nature Ecology & Evolution, from: https://doi.org/10.1038/s4155​9-­017-­0391-­4
4(11), 1485–­ 1494. Available from: https://doi.org/10.1038/s4155​ Naeem, S. (2002) Ecosystem consequences of biodiversity loss: the
9-­020-­1280-­9 evolution of a paradigm. Ecology, 83(6), 1537–­1552. Available
Johnson, K.H., Vogt, K.A., Clark, H.J., Schmitz, O.J. & Vogt, D.J. from: https://doi.org/10.1890/0012-­9658(2002)083[1537:ECOBL​
(1996) Biodiversity and the productivity and stability of ecosys- T]2.0.CO;2
tems. Trends in Ecology & Evolution, 11(9), 372–­377. Available Newbold, T., Hudson, L.N., Hill, S.L.L., Contu, S., Lysenko, I.,
from: https://doi.org/10.1016/0169-­5347(96)10040​-­9 Senior, R.A. et al. (2015) Global effects of land use on local ter-
Jones, H.P., Jones, P.C., Barbier, E.B., Blackburn, R.C., Rey, J.M., restrial biodiversity. Nature, 520(7545), 45–­50. Available from:
Benayas, K.D. et al. (2018) Restoration and repair of Earth's https://doi.org/10.1038/natur​e14324
damaged ecosystems. Proceedings of the Biological Sciences, Queiroz, C., Meacham, M., Richter, K., Norström, A.V., Andersson,
285(1873), 20172577. Available from: https://doi.org/10.1098/ E., Norberg, J. et al. (2015) Mapping bundles of ecosystem
rspb.2017.2577 services reveals distinct types of multifunctionality within a
Kollmann, J., Meyer, S.T., Bateman, R., Conradi, T., Gossner, M.M., Swedish landscape. Ambio, 44(S1), 89–­ 101. Available from:
de Souza, M. et al. (2016) Integrating ecosystem functions into https://doi.org/10.1007/s1328​0 -­014-­0601-­0
restoration ecology-­ recent advances and future directions. R Core Team. (2016) R: a language and environment for statistical
Restoration Ecology, 24(6), 722–­730. Available from: https://doi. computing. Available from: https://www.r-­proje ​ct.org/ [Accessed
org/10.1111/rec.12422 6th May 2016].
Lavorel, S. (2013) Plant functional effects on ecosystem services. Raudsepp-Hearne, C., Peterson, G.D. & Bennett, E.M. (2010)
Journal of Ecology, 101(1), 4–­ 8. Available from: https://doi. Ecosystem service bundles for analyzing tradeoffs in diverse
org/10.1111/1365-­2745.12031 landscapes. Proceedings of the National Academy of Sciences of
Lefcheck, J.S., Byrnes, J.E.K., Isbell, F., Gamfeldt, L., Griffin, J.N., the United States of America, 107, 5242–­5247.
Eisenhauer, N. et al. (2015) Biodiversity enhances ecosystem Ross, S.R.P.-­J., Arnoldi, J.-­F., Loreau, M., White, C.D., Stout, J.C.,
multifunctionality across trophic levels and habitats. Nature Jackson, A.L. et al. (2021) Universal scaling of robustness of eco-
Communications, 6(April), 6936. Available from: https://doi. system services to species loss. Nature Communications, 12(1),
org/10.1038/ncomm​s7936 5167. Available from: https://doi.org/10.1038/s4146​7-­021-­25507​-­5
Lindborg, R. & Eriksson, O. (2005) Functional response to land Rullens, V., Lohrer, A.M., Townsend, M. & Pilditch, C.A. (2019)
use change in grasslands: comparing species and trait data. Ecological mechanisms underpinning ecosystem service bundles
Ecoscience, 12(2), 183–­191. in marine environments -­a case study for shellfish. Frontiers in
Lindborg, R., Gordon, L.J., Malinga, R., Bengtsson, J., Peterson, Marine Science, 6(July), 409.
G., Bommarco, R. et al. (2017) How spatial scale shapes the Solan, M., Cardinale, B.J., Downing, A.L., Engelhardt, K.A.M.,
generation and management of multiple ecosystem services. Ruesink, J.L. & Srivastava, D.S. (2004) Extinction and eco-
Ecosphere, 8(4), e01741. Available from: https://doi.org/10.1002/ system function in the marine benthos. Science, 306(5699),
ecs2.1741 1177–­1180 Available from: http://scien​c e.scien​c emag.org/conte​
Loreau, M. (1998) Biodiversity and ecosystem functioning: a mecha- nt/306/5699/1177.abstract [Accessed 4th April 2019].
nistic model. Proceedings of the National Academy of Sciences of Strijker, D. (2005) Marginal lands in Europe—­c auses of decline. Basic
the United States of America, 95(10), 5632–­5636 Available from: and Applied Ecology, 6(2), 99–­106. Available from: https://doi.
http://www.pnas.org/conte​nt/95/10/5632.abstract [Accessed 3rd org/10.1016/j.baae.2005.01.001
May 2019]. Tilman, D., Reich, P.B. & Knops, J.M.H. (2006) Biodiversity and eco-
Loreau, M., Shahid, N. & Pablo, I. (2002) Biodiversity and ecosys- system stability in a decade-­long grassland experiment. Nature,
tem functioning: synthesis and perspectives. Oxford: Oxford 441(7093), 629–­632. Available from: https://doi.org/10.1038/natur​
University Press. e04742
Mace, G.M., Norris, K. & Fitter, A.H. (2012) Biodiversity and eco- Violle, C., Navas, M.-­ L., Vile, D., Kazakou, E., Fortunel, C.,
system services: a multilayered relationship. Trends in Ecology Hummel, I. et al. (2007) Let the concept of trait be func-
& Evolution, 27, 19–­26. Available from: https://doi.org/10.1016/j. tional! Oikos, 116(5), 882–­ 892. Available from: https://doi.
tree.2011.08.006 org/10.1111/j.0030-­1299.2007.15559.x
Maes, J., Teller, A., Erhard, M., Murphy, P., Paracchini, M.-­L., Cano, Waldén, E. & Lindborg, R. (2016) Long term positive effect of grass-
J.I.B. et al. (2014) Indicators for ecosystem assessments under land restoration on plant diversity -­success or not? PLoS One,
action 5 of the EU biodiversity strategy to 2020. Publications 11(5), e0155836. Available from: https://doi.org/10.1371/journ​
Office of the European Union. Available from: https://doi. al.pone.0155836
org/10.2779/75203 Wardle, D.A., Bonner, K.I. & Nicholson, K.S. (1997) Biodiversity
Manning, P., Loos, J., Barnes, A.D., Batáry, P., Bianchi, F.J.J.A., and plant litter: experimental evidence which does not sup-
Buchmann, N. et al. (2019) Transferring biodiversity-­e cosystem port the view that enhanced species richness improves eco-
function research to the management of ‘real-­world’ ecosystems. system function. Oikos, 79(2), 247. Available from: https://doi.
Advances in Ecological Research, 61(January), 323–­356. Available org/10.2307/3546010
from: https://doi.org/10.1016/bs.aecr.2019.06.009 Watson, J.E.M., Jones, K.R., Fuller, R.A., Di Marco, M., Segan, D.B.,
Manning, P., van der Plas, F., Soliveres, S., Allan, E., Maestre, F.T., Butchart, S.H.M. et al. (2016) Persistent disparities between re-
Mace, G. et al. (2018) Redefining ecosystem multifunctional- cent rates of habitat conversion and protection and implications
ity. Nature Ecology & Evolution, 2(3), 427–­436. Available from: for future global conservation targets. Conservation Letters, 9(6),
https://doi.org/10.1038/s4155​9-­017-­0461-­7 413–­421. Available from: https://doi.org/10.1111/conl.12295
14610248, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ele.14220 by Lanzhou University, Wiley Online Library on [11/02/2024]. 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
WALDÉN et al.     | 941

Wilson, J.B., Peet, R.K., Dengler, J. & Pärtel, M. (2012) Plant


species richness: the world records. Journal of Vegetation S U PP ORT I NG I N F OR M AT ION
Science, 23(4), 796–­ 802. Available from: https://doi. Additional supporting information can be found online
org/10.1111/j.1654-­1103.2012.01400.x in the Supporting Information section at the end of this
Winfree, R., Fox, J.W., Williams, N.M., Reilly, J.R. & Cariveau, article.
D.P. (2015) Abundance of common species, not species rich-
ness, drives delivery of a real-­world ecosystem service. Ecology
Letters, 18(7), 626–­635. Available from: https://doi.org/10.1111/
ele.12424
Zavaleta, E.S., Pasari, J.R., Hulvey, K.B. & David Tilman, G. (2010) How to cite this article: Waldén, E., Queiroz, C.,
Sustaining multiple ecosystem functions in grassland commu- Plue, J. & Lindborg, R. (2023) Biodiversity mitigates
nities requires higher biodiversity. Proceedings of the National trade-­offs among species functional traits
Academy of Sciences of the United States of America, 107(4), underpinning multiple ecosystem services. Ecology
1443–­1446. Available from: https://doi.org/10.1073/pnas.09068​
Letters, 26, 929–941. Available from: https://doi.
29107
Zhao, Y., Liu, Z. & Wu, J. (2020) Grassland ecosystem services: a org/10.1111/ele.14220
systematic review of research advances and future directions.
Landscape Ecology, 35(1), 793–­814. Available from: https://doi.
org/10.1007/s1098​0 -­020-­0 0980​-­3

You might also like