Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

Contents lists available at ScienceDirect

Perspectives in Plant Ecology, Evolution and Systematics

journal homepage: www.elsevier.com/locate/ppees

Facilitation mediates species presence beyond their environmental optimum T

a,e b c
Michael J. O’Brien , Luis Fernando Tavares de Menezes , Kari Anne Bråthen ,
⁎
Gianalberto Losapiod, Francisco I. Pugnairea,
a
Estación Experimental de Zonas Áridas, Consejo Superior de Investigaciones Científicas (EEZA-CSIC), Carretera de Sacramento s/n, E-04120 La Cañada, Almería, Spain
b
Universidade Federal do Espírito Santo, Centro Universitário Norte do Espírito Santo, BR 101 Norte, Km 60 - Bairro Litorâneo, São Mateus, 29.932-540, Espírito Santo,
Brazil
c
Department of Arctic and Marine Biology, UiT-The Arctic University of Norway, NO-9037 Tromsø, Norway
d
ETH Zurich, Department of Environmental Systems Science, Schmelzbergstrasse 9, 8092, Zurich, Switzerland
e
Área de Biodiversidad y Conservación, Universidad Rey Juan Carlos, c/Tulipán s/n., E-28933 Móstoles, Spain

A R T I C LE I N FO A B S T R A C T

Keywords: Species distributions are driven by abiotic conditions that filter species with specific traits and physiological
Species distributions tolerances and match them with their suitable environment. Plant–plant interactions can constrict (through
Climate change competition) or loosen (through facilitation) the strength of these environmental filters, which in turn inhibit or
Community assembly enhance establishment and recruitment of plant species at a finer spatial scale. Although competition is often the
Environmental filters
focus of community assembly processes that further impede the entry of plant species into a site, facilitation is
Nurse-plants
Plant-climate interactions
also important for potentially loosening environmental filters (especially climatic filters such as temperature and
precipitation), ultimately enhancing species occurrence beyond their physiological optimum. We used multiple
data sets from the arid site of Rambla del Saltador Valley to test the hypothesis that facilitation by a nurse-plant
promotes the presence of herbaceous, beneficiary species beyond their environmental optimum relative to open
sites. Furthermore, we propose that the median elevation and community composition of herbaceous species
expands with the age of the nurse-plant, and we tested this hypothesis by examining 105 beneficiary species
under 50 nurse-plant shrubs varying in age from 6 to 48 years old. We found nurse-plants both facilitate her-
baceous species occurrence beyond their median elevation and support more diverse and a distinctly different
composition of species in contrast to open sites. Specifically, herbaceous species that originate from a median
elevation more than 600 to 700 m above the site only existed beneath nurse-plants, and below this median
elevation, half the species only occurred below the nurse-plant. Moreover, the richness and elevation provenance
of the herbaceous species increased with increasing nurse-plant age. Our results highlight the importance of
facilitation for alleviating physiological strain (in support of the strain hypothesis) and mediating regional
species distributions, which has implications for understanding species movements and community assembly at
larger-scales under hotter and drier climates.

1. Introduction Marshall, 2013; Kraft et al., 2015; Schöb et al., 2013) while, in contrast,
facilitation can increase the realized niche of plant species at local and
Understanding the processes that underlie patterns of species dis- regional scales by ameliorating environmental filters — such as climatic
tribution is a cornerstone of ecology (MacArthur, 1972; Whittaker and conditions across environmental gradients — thereby promoting spe-
Niering, 1975), which is increasingly relevant for predicting commu- cies presence beyond their environmental optimum (Armas et al., 2011;
nity assembly in a changing climate (Alexander et al., 2016, 2015; Gross et al., 2010; Liancourt et al., 2017; Schöb et al., 2014a; Soliveres
Soliveres et al., 2015). Species diversity at any location is governed by et al., 2015; Valiente-Banuet et al., 2006; Wisz et al., 2013). Assessing
large-scale biogeographic and evolutionary processes as well as by local the influence of nurse-plants on regional-scale species distribution can
biotic and abiotic mechanisms (Brooker et al., 2008; Cavieres et al., help our understanding of species dynamics and community assembly
2014; Harrison and Cornell, 2008; Ricklefs, 2008, 2004). Within the under a changing environment (Bruno et al., 2003; Cavieres et al.,
suite of plant–plant interactions, competitive exclusion can inhibit 2016; Liancourt et al., 2017; O’Brien et al., 2017b; Pugnaire et al.,
species occurrence (Alexander et al., 2015; Chesson, 2000; Hart and 1996; Schöb et al., 2012).

⁎
Corresponding authors.
E-mail addresses: mikey.j.obrien@gmail.com (M.J. O’Brien), fip@eeza.csic.es (F.I. Pugnaire).

https://doi.org/10.1016/j.ppees.2019.03.004
Received 28 June 2018; Received in revised form 17 January 2019; Accepted 24 March 2019
Available online 01 April 2019
1433-8319/ © 2019 Elsevier GmbH. All rights reserved.
M.J. O’Brien, et al. Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

Facilitation is a widespread process particularly important for spe- allows species occurrence far from their environmental optimum (i.e.
cies coexistence in low resource or stressful climatic environments median elevation). Using the model nurse-plant, Retama sphaerocarpa
(Callaway, 2007; Filazzola and Lortie, 2014; Fugère et al., 2012; He (Retama; Pugnaire et al., 1996), we tested the hypothesis that the di-
et al., 2013). In plant communities, facilitation is often, but not solely, versity and breadth of median elevation of the beneficiary species in-
shown as the effect of a nurse-plant on beneficiary species whereby the creases with Retama age. We posit that older nurse-plants support
nurse promotes the establishment, growth or reproduction of stress- species far from their median elevation due to a greater complexity of
sensitive species under their canopy (Bertness and Callaway, 1994). The microsites and accumulation of resources beneath them. We expect that
relative importance of facilitation for the persistence of a species in- species originating from higher elevations and cooler and wetter cli-
creases as the environmental conditions at a site diverge from the mates can occur in lower and more arid elevations due to facilitation,
physiological optimum of the species — i.e. the more strain (termed the while the relative importance of facilitation will be weaker on species
strain hypothesis) on a species at a site (Gross et al., 2010; Liancourt moving from more arid elevations.
et al., 2017). In arid systems, this hypothesis would imply that facil-
itation is asymmetric with species from higher elevations (cooler and 2. Materials and methods
wetter sites) benefitting more from facilitation than species from lower
elevations (hotter and drier sites) due to the greater physiological tol- 2.1. Field site and data collection
erance to arid conditions of lower elevation species and more physio-
logical strain on higher elevation species (Liancourt et al., 2017). In All species occurrence data were collected from the Rambla del
other words, nurse species can expand niche space within sub-optimal Saltador valley in the Almeria province, Southeast Spain (37°07’43” N,
environments thereby allowing species to exist in locations beyond 2°22’13” W; 630 m elevation). The site is arid with a dry season from
their environmental optimum (Wiens, 2011). This process can in turn May to September and a wetter season from December to March
increase local community diversity and productivity, which mediates (Puigdefábregas et al., 1996). The average rainfall is 256 mm per year
geographic distributions at larger scales (Bulleri et al., 2016; Cavieres with irregular intra- and inter-annual distribution. The valley is alluvial
et al., 2014; Ettinger and HilleRisLambers, 2017; Kraft et al., 2015; detritus with low water holding capacity, organic matter and nutrient
Schöb et al., 2012). concentrations. It is dominated by randomly and sparsely distributed
Common mechanisms underpinning facilitation are amelioration of Retama sphaerocarpa (Retama) individuals from the Fabaceae family, a
microclimate conditions and improvement of understory soil proper- leafless leguminous shrub with a deep root system which performs
ties, and as nurse-plants grow older, the complexity and heterogeneity hydraulic lift (Prieto et al., 2011). Other vascular plants are con-
of soil and climatic conditions will increase (Lozano et al., 2017; centrated around Retama with sparse vegetation between shrubs
Pugnaire et al., 2011, 1996). This correlation between the complexity of dominated by drought tolerant plants (Pugnaire et al., 1996), especially
the understorey niche and nurse-plant age suggests that older nurse- in low rainfall years where few plants occur in open sites (O’Brien et al.,
plants will support more beneficiary species by providing niche space to 2017b). Retama shrubs improve their understorey environment by in-
a larger breadth of physiological optimum. Pugnaire et al., (1996) creasing soil water content and soil organic matter and nutrients
found older and larger nurse-plants host more than twice the number of (Pugnaire et al., 1996). Shrub canopies also buffer solar radiation
beneficiary species than younger nurse-plants. This was mainly attrib- reaching the soil up to ˜40%, leading to a decrease in mean daily
uted to the increase of organic matter, N and P availability and water temperatures of up to 3 °C compared to gaps and 7 °C in max tem-
availability with increasing nurse-plant age (Pugnaire et al., 2011, perature (Moro et al., 1997b). Combined, microclimate conditions and
1996). Although such increases in resource availability with nurse-plant litter accumulation under the canopy produce a number of strong
age may enhance competition in the understory, the facilitation effects gradients –often opposing each other (Moro et al., 1997a), that increase
of the nurse often overshadow negative interactions among beneficiary niche availability (Michalet et al., 2015; O’Brien et al., 2017b) and
species (Pugnaire et al., 1996; Schöb et al., 2013). allow for the establishment of a great diversity of plant species. The
Climatic conditions in many ecosystems are becoming hotter and positive effects of Retama on understorey productivity and diversity has
drier (O’Brien et al., 2017a), and for plant species to avoid extinction been widely shown in previous studies (Michalet et al., 2015; Moro
(either local, regional or global) under novel climatic regimes, they will et al., 1997b; O’Brien et al., 2017b; Prieto et al., 2010; Pugnaire et al.,
need to either change their distributions to track their current climatic 1996; Pugnaire and Luque, 2001), although the facilitative effects of
conditions or persist under novel climatic conditions in their current Retama decrease with increasing elevation (Pugnaire and Luque, 2001).
distributions (Alexander et al., 2015; Ettinger and HilleRisLambers, Two data sets were used for this analysis. The first dataset was from
2017). For example, in alpine systems, climate change is increasing Pugnaire and Lázaro (2000) which includes the herbaceous beneficiary
temperatures across the elevation gradient, and plant species will need species under 50 Retama shrubs that ranged from 6 to 48 years of age
to either move upward in elevation to avoid warming or persist under (measured from destructive harvests and ring counts) and with crown
warmer conditions at their current elevation (Alexander et al., 2015; areas ranging from 0.4–44 m2 (estimated allometrically from length and
Valiente-Banuet et al., 2006). In arid systems such as the Mediterra- width). The second dataset was compiled from a number of experiments
nean, temperatures decrease but precipitation increases with increasing that were carried out at the same field site over multiple years (Armas
elevation (Liancourt et al., 2017; Schöb et al., 2014b), so that hotter, et al., 2011; Hortal et al., 2015; O’Brien et al., 2017b; Schöb et al.,
drier conditions are moving up in elevation under climate change 2013). These data were of herbaceous plants collected in open spaces
(Scherrer and Körner, 2010). Nurse-plants may buffer species from between Retama in sampled areas ranging from 0.2–11.3 m2 (total
warmer and drier conditions thereby providing refugia from the sample area of 12.6 m2) and provided an open site contrast to the
changing climate (O’Brien et al., 2017b). The presence of beneficiary herbaceous community under Retama.
species at lower and more arid elevations beyond their preferred en- Elevation ranges of beneficiary species were compiled from Blanca
vironment would indicate the buffering ability of nurse-plants in novel et al., (2009). We used the median elevation of the recorded range of
climatic conditions, and as nurse-plants age and grow larger, they may each species (i.e. median point of the lower and upper limits) as the unit
be able to buffer a greater quantity and diversity of beneficiary species. to define their optimal elevation. Using median elevation is a robust
Particularly, the more beneficiary species deviate from their physiolo- measure of central tendency of species distribution but limits between-
gical optimum the greater the importance of nurse-plants for mediating species comparison as it ignores the shape (i.e. assumes normal dis-
higher diversity. tribution) and width of the distribution such that two species with a
We test the concept that nurse-plants alleviate environmental con- median elevation of 500 m are assumed the same even though one
ditions (i.e. reduce physiological strain) at an arid site (Fig. 1) which could exist from 400 to 600 m and the other from 200 to 800 m.

25
M.J. O’Brien, et al. Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

Fig. 1. Microclimate under nurse-plants increase with age

providing establishment for species from different elevations.
The distribution of a species is determined by a matching of
species characteristics to specific environmental conditions
(i.e. environmental filtering). We examine whether nurse-
plants mediate species from outside their optimal environment
and whether facilitation increases in importance the further
from that optimal environment a species deviates (greater
physiological strain). Here we show an example that species
from higher elevations (colours represent increasing elevation
and decreasing aridity from brown to white) can persist at
lower elevations due to environmental modification by nurse-
plants and that older nurse-plants have more microsites be-
neath them that facilitate a greater range of species (i.e. only
local species from the same site under younger Retama and a
range of high elevation to local species under older Retama).
Arrows match elevation colours to Retama size.

However, our goal was to assess the level of deviation between our link function (see ANOVA Table S1 of Wald statistics in the Appendix).
study site elevation and the optimal elevation for each species rather We included a random effect for Retama plant (a random term with 50
than specifically contrasting the differences among species. For prac- levels, the number of individual shrubs assessed). We included a cov-
tical reasons, we assumed that the median elevation ± 200 m from the ariate for Retama canopy area to control for the effects of plant size.
elevation of the site were native to the site (Scherrer and Körner, 2010). To estimate the significance of the observed species richness across
Based on these assumptions, species were assigned to three elevation the age gradient, we used a null model approach. Random communities
categories: 1) less than 430 masl (low elevation compared to our study were assigned according to a probabilistic null model that builds
site), 2) 430–830 masl (home elevation of our study site) and 3) greater communities from the community matrix considering the probability of
than 830 masl (high elevation compared to our study site). We tested species presence-absence among Retama shrubs (Bascompte et al.,
the importance of this categorization by altering the cut-off of eleva- 2003). In particular, the community matrix (M = Retama x Species)
tions for these three levels and found that our results were qualitatively indicates the presence or absence of a beneficiary species j beneath a
robust regardless of the width of the category examined (see Figure S1 Retama i. For each Retama–species pair, the probability that a species j
and S2 in Appendix). occurs beneath a Retama i is drawn from a binomial distribution with
density

2.2. Statistical analysis 1 ⎛ aSpecies a

d= ⎜ + Retama ⎞⎟
2 ⎝ Retama Species ⎠
To assess the importance of Retama for species richness of plants
originating from different median elevations, we calculated the ratio of where aSpecies is the number of Retama a beneficiary species j is oc-
the number of species present only under Retama from a median ele- curring beneath, aRetama is the number of beneficiary species occurring
vation versus the total number of species from that median elevation beneath a Retama i, ‘Retama’ is the total number of Retama and open
present under Retama and in open sites. Therefore, if all species from a plot, ‘Species’ is the total number of beneficiary species. Thus, the oc-
median elevation were only found under Retama, the ratio would be currence probability of a beneficiary species beneath Retama depends
one, and if all species were only found in open sites, the ratio would be on both the distribution of each beneficiary species among Retama
zero. We calculated this ratio at 100 m increments from greater than shrubs and on the capability of each Retama of hosting beneficiary
1000 m difference downward (species with a median elevation above species. To test whether the observed species richness was significantly
1630 m) to greater than 500 m difference upward (species with a different from the probabilistic null model, we calculated species
median elevation of less than 130 m). We used a simple loess smoother richness from 1000 probabilistic communities and used linear mixed-
to show the trend across the change in elevation. effects model of richness as a function of area sampled and age (see
To test for differences in composition between Retama and open site, ANOVA Table S2 of Wald statistics in Appendix). This analysis was
we calculated the dissimilarities in the herbaceous community com- performed separately for the community of beneficiary species from
positions using the Jaccard index based on species presence–absence low, home and high elevation. We also assessed the standardized effect
per plot. We performed constrained analysis of proximities on the size with a z-score, which was calculated as the difference between the
Jaccard distance matrix to test the effect of the constraining terms site observed richness and mean simulated richness under a Retama divided
(a fixed factor with 2 levels; Retama and open) and age (a continuous by the standard deviation of the simulated richness. We analyzed this as
variable; open were considered zero). We tested the significance of the a function of Retama age in a linear model separately for the community
constraining terms with a permutation test. If the inertia in the per- of beneficiary species from low, home and high elevation (see ANOVA
muted models was lower than in the constrained model, then the as- Table S3 of Wald statistics in Appendix). The mixed effects models were
sociation was considered statistically significant. These analyses were performed with the asreml-R package (ASReml 3, VSN International,
done with R statistical software (version 3.3.2; http://r-project.org). UK) in the R statistical software (version 3.3.2; http://r-project.org).
The vegdist function in the vegan package (Oksanen et al., 2015) was
used to calculate Jaccard dissimilarities. The constrained analysis of 3. Results
proximities was performed with the capscale function (Legendre and
Anderson, 1999), and the permutation test was done with the anova After examining species distributions based on the most up-to-date
function in the vegan package. taxonomy and floristic data (Blanca et al., 2009), a total of 105 species
Species richness was analysed as a function of Retama age (a con- were identified in our Retama dataset (see Table S4 for full species list
tinuous variable), origin elevation of the species (a fixed factor with 3 in Appendix), and one species, Polycarpon tetraphyllum, was removed
levels; low, home and high elevation) and their interaction using a from analysis due to the uncertainty of the subspecies. Of the 104
generalized-linear mixed model with a Poisson distribution and a log species analysed, 60 species (˜58% of the total) had their median

26
M.J. O’Brien, et al. Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

Fig. 2. Elevation range and composition of species under Retama and open sites. The site used in this study is at 630 masl. Expansion of home elevation was
calculated as the difference in median home elevation reported for the species and the elevation of the site. (a) The number of species found under all Retama (black)
and in open areas (grey) that originated from elevations greater than 1000 m above the site (-1000 m difference) to 500 m below the site (500 m difference). There
were 104 species in total found under Retama shrubs and 61 species in total in open areas. (b) The proportion of the number of species found under Retama relative to
the total species richness found from that median elevation both under Retama and in open sites. The black line represents a loess smoother with 95% CI. (c) The
composition was statistically different between open and Retama communities as well as from older to younger Retama. Numbers represent Retama age, and the line
represents the directional effect of age.

elevation within 200 m of the elevation of the site (home origin; elevation species (slope of age on log scale = 0.02, 95% CI: 0.0004 –
Fig. 2a). Of the remaining species, Retama provided habitat for 11 0.03) while the richness of low elevation species was similar across all
species (˜11%) with a lower median elevation, and 33 species (32%) Retama ages (slope of age on log scale = 0.01, 95% CI: -0.01 – 0.03).
species with a higher median elevation (Fig. 2a). In open site data, 61 Therefore, home elevation species richness increased by ˜15 species
species were identified (see Table S4 for full species list) with 35 species from 6-year-old Retama to 27-year-old Retama while low elevation
(57%) from the home origin while 9 species (15%) were from a lower species richness increased by ˜1 species and high elevation species
elevation and 17 species (28%) were from a higher elevation. richness increased by ˜3 species across the same Retama age range. This
The range of species found in open plots was much narrower with resulted in a near doubling of species richness between younger and
species originating from 200 to 1200 m in the open versus 200–1850 m older Retama. Our analysis compared to a probabilistic null model
under Retama. Therefore, the range of median elevations of species at confirmed these results as the difference between the observed and null
the study site was more than 1.5 times larger with facilitation by community increased with increasing Retama age (see Figure S4 in
Retama than without. In support of our hypothesis, moving up in ele- Appendix). The standardized z-scores also significantly increased with
vation was less common — Silene littorea was the lowest species found Retama age.
with a distribution between 0 and 400 m — although there is a lower
elevation limit of 0 m as species are not moving up from below sea-
level. Alternatively, species moved down in elevation as much as 4. Discussion
1220 m (e.g. Alyssum granatense found at an elevation range of
1600–2100 masl), and the community under Retama had twice as many Our combined analyses of beneficiary species under and outside
species from higher elevations than open sites. Furthermore, 100% of canopies of nurse-plants shows that nurse-plants support species from a
the species originating from a median elevation above 1300 m were larger median elevation breadth, and the relative importance of facil-
only found under Retama (Fig. 2b) as no species from the high eleva- itation increases as the median elevation of the beneficiary species in-
tions occurred in open sites. Below this elevation, 50% of the species on creases. These results provide direct evidence in support of reduced
average from all other median elevations were found solely under Re- physiological strain hypothesis (Gross et al., 2010; Liancourt et al.,
tama. This resulted in Retama hosting a distinctly different community 2017). However, nurse-plants also supported distinctly different species
than that found in open sites (P < 0.001), and beneficiary community than the open sites even for beneficiary species that originated from a
composition under older Retama was significantly different from similar median elevation as the site, suggesting facilitation also func-
younger Retama communities (P < 0.001; Fig. 2c). The results in tions to support species unable to survive in open areas. These processes
figure two were robust if the largest shrubs (> 12 m2) were removed led to a distinct community composition under Retama relative to open
from the analysis to control for differences in sampling area between sites, confirming that facilitation is an important process affecting
open and Retama plots (Figure S3 in Appendix). community assembly and diversity (Brooker et al., 2008; Tirado and
Overall species richness increased with increasing Retama age re- Pugnaire, 2005). Specifically, these results highlight that the im-
gardless of median elevation origin (Fig. 3a). Species richness under the portance of facilitation increases with increasing differences between
youngest Retama shrubs (6 years old) was highest for beneficiary spe- beneficiary-species average habitat and nurse-plant habitat. In addition,
cies originating from the home site (12.0 species, 95% CI: 10.1–14.3), our results from multiple analyses indicate that older Retama plants
lowest for species from low elevations (2.8 species, 95% CI: 2.1–3.6) support nearly twice as many species and from a larger breadth of
and intermediate from the high elevations (6.4 species, 95% CI: median elevation than younger Retama plants.
5.2–7.8). After controlling for Retama canopy area, species richness Mechanisms of facilitation for reducing physiological strain are due
significantly increased with Retama age for the home and high elevation to effects of facilitator shrubs on climatic conditions under their canopy
groupings but not for the species from the low elevation (Fig. 3b-d). The (Pugnaire et al., 1996) with fundamental consequences for relaxing
increase in species richness was greatest for home elevation species environmental filters associated with elevation gradients in arid eco-
(slope of age on log scale = 0.03, 95% CI: 0.01 – 0.04) followed by high systems (Moro et al., 1997b). Therefore, nurse-plants in arid environ-
ments provide refugia from high temperatures and low availability of

27
M.J. O’Brien, et al. Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

Fig. 3. Beneficiary species richness as a function of Retama age and origin elevation. Species richness (95% CI) significantly increased with Retama age, after
controlling for the effects of area, for beneficiary species originating at elevations from and above the site but not from below the site (a). Average species richness
was lowest for beneficiary species originating from low elevation (b), highest for beneficiary species from the home elevation (c) and intermediate for beneficiary
species from high elevation (d). Therefore, older Retama had more species than younger Retama and expanded the elevation range of more species than younger
Retama.

resources, particularly water and nutrients (Pugnaire et al., 2011), habitats (including microhabitats under facilitator species) may have
which allowed the presence of beneficiary species far from their op- higher predictive power at smaller spatial resolutions. Indeed, our study
timal environmental conditions, especially species from higher eleva- indicates not only that the structure of plant communities is driven by
tions (wetter and cooler sites). These results have implications for the local biotic interactions but also that local plant–plant interactions can
distribution of species due to changing climates because it suggests that shape the large-scale distribution of species. These results imply that
plant–plant interactions can supersede climatic filters that limit species macroecological patterns may result from microecological processes
establishment and survival, and in turn may allow species to either (Wisz et al., 2013). Even though the data from open locations also
persist in altered climates or more effectively track shifting climatic supported some species originating from higher and lower elevations,
conditions. Models forecasting the presence of species (i.e. distribu- the overall distribution of the median elevation of species was smaller,
tions) under climate change should therefore consider the effects of and no species from above 1300 m were found in open sites. This
nurse-plant refugia. smaller range was found despite the fact that the open data came from
Our results show that plant–plant interactions at the local-scale in- multiple years and also included the 2009–2010 year, which had sig-
fluence meso-scale diversity patterns, which directly supports the re- nificantly higher than average rainfall (O’Brien et al., 2017b). These
sults of Cavieres et al., (2014) that showed similar processes shape di- multiple years of sampling and high rainfall years likely inflated the
versity in alpine ecosystems worldwide. Species distributions are often species richness found in open sites. Our analysis of species composition
defined by bioclimatic envelopes (Elith and Leathwick, 2009; Sexton further supported the importance of Retama presence and age for har-
et al., 2009), since climate variables are primary determinants of the bouring unique beneficiary species unable to persist outside the Retama
species richness and composition of vascular plant communities (Marini canopy. Therefore, facilitation relaxed climatic filters on plant diversity
et al., 2008; Moser et al., 2005). However, the presence of favourable that restricted establishment and recruitment of species from other

28
M.J. O’Brien, et al. Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

environments while also supporting species originating locally that are Acknowledgements
unable to exist outside the nurse-plant canopy (Armas et al., 2011;
O’Brien et al., 2017b; Schöb et al., 2013). This work was funded by the Spanish Research Agency (AEI)
In support of our prediction, more species from higher elevations (CGL2017-84515-R). MOB was supported by the Swiss National Science
than from lower elevations benefitted from facilitation. A primary Foundation through an Advanced Postdoc Mobility Fellowship
component of facilitation in these arid sites is reduced temperatures (P300PA_167758). This work contributes to the University of Zurich
from increased shading and hydraulic lift from deeply rooted nurse- Research Priority Program on Global Change and Biodiversity (URPP-
plants (Butterfield et al., 2010; López-Pintor et al., 2006; Moro et al., GCB).
1997b), which would create environmental conditions more suitable to
plants originating from cooler and wetter sites at higher elevations than Appendix A. Supplementary data
from warmer and drier sites at lower elevations. This result supports the
strain hypothesis (Gross et al., 2010; Liancourt et al., 2017). The ele- Supplementary material related to this article can be found, in the
vation in this arid system is associated with temperature and pre- online version, at doi:https://doi.org/10.1016/j.ppees.2019.03.004.
cipitation patterns that influence plant growth and lead to changes in
species composition and physiognomy of plant communities over large References
environmental gradients (Austin, 1987; Körner, 2007; Pendry and
Proctor, 1997; Proctor et al., 1988), but Retama sustained species that Alexander, J.M., Diez, J.M., Levine, J.M., 2015. Novel competitors shape species’ re-
originated from sites as much as 1200 m above the study site. There- sponses to climate change. Nature 525, 515–518. https://doi.org/10.1038/
nature14952.
fore, Retama substantially relaxed environmental constraints, which Alexander, J.M., Diez, J.M., Hart, S.P., Levine, J.M., 2016. When climate reshuffles
may ultimately result in higher species diversity. Dispersal is also a competitors: a call for experimental macroecology. Trends Ecol. Evol. 31, 831–841.
likely secondary component of the greater number of species from high https://doi.org/10.1016/j.tree.2016.08.003.
Armas, C., Rodríguez-Echeverría, S., Pugnaire, F.I., 2011. A field test of the stress-gradient
elevation at the site because it is easier to disperse downslope via wind hypothesis along an aridity gradient. J. Veg. Sci. 22, 818–827. https://doi.org/10.
and water than upslope (Thompson and Katul, 2009; Venable et al., 1111/j.1654-1103.2011.01301.x.
2008). Therefore, introduction rates of species from higher elevations Austin, M.P., 1987. Models for the analysis of species’ response to environmental gra-
dients. Vegetatio 69, 35–45. https://doi.org/10.1007/BF00038685.
would be greater than that of species from lower elevations. There was Bascompte, J., Jordano, P., Melian, C.J., Olesen, J.M., 2003. The nested assembly of
a constraint on the median elevation limit of lower elevation species plant-animal mutualistic networks. Proc. Natl. Acad. Sci. 100, 9383–9387. https://
(i.e. species were limited to 630 m below the site due to sea-level), but doi.org/10.1073/pnas.1633576100.
Bertness, M.D., Callaway, R., 1994. Positive interactions in communities. Trends Ecol.
the results show the proportion of species richness that depended on
Evol. 9, 187–191. https://doi.org/10.1016/0169-5347(94)90087-6.
facilitation was decreasing with decreasing median elevation. Blanca, G., Cabezudo, B., Cueto, M., Fernández López, C., Morales Torres, C., 2009. Flora
Not surprisingly, most species were from the same environmental Vascular de Andalucía Oriental. Consejería de Medio Ambiente, Junta de Andalucía,
belt as the field site (Thermomediterranean semiarid; Lázaro et al., Sevilla.
Brooker, R.W., Maestre, F.T., Callaway, R.M., Lortie, C.L., Cavieres, L.A., Kunstler, G.,
2001). However, the herbaceous communities were comprised of many Liancourt, P., Tielbörger, K., Travis, J.M.J., Anthelme, F., Armas, C., Coll, L., Corcket,
species found in wetter environments. For example, Armas et al., E., Delzon, S., Forey, E., Kikvidze, Z., Olofsson, J., Pugnaire, F., Quiroz, C.L., Saccone,
(2011) reported 10 species from this site that are distributed across a P., Schiffers, K., Seifan, M., Touzard, B., Michalet, R., 2008. Facilitation in plant
communities: the past, the present, and the future. J. Ecol. 96, 18–34. https://doi.
gradient from 250 to 2000 mm of annual rainfall (4-fold change in org/10.1111/j.1365-2745.2007.01295.x.
mean annual precipitation). Therefore, it is important to note that the Bruno, J.F., Stachowicz, J.J., Bertness, M.D., 2003. Inclusion of facilitation into ecological
elevation gradient is an indicator of both a temperature and pre- theory. Trends Ecol. Evol. 18, 119–125. https://doi.org/10.1016/S0169-5347(02)
00045-9.
cipitation gradient, and facilitation at the site is also contributing to the Bulleri, F., Bruno, J.F., Silliman, B.R., Stachowicz, J.J., 2016. Facilitation and the niche:
distribution of species across latitudinal and longitudinal precipitation implications for coexistence, range shifts and ecosystem functioning. Funct. Ecol. 30,
gradients. 70–78. https://doi.org/10.1111/1365-2435.12528.
Butterfield, B.J., Betancourt, J.L., Turner, R.M., Briggs, J.M., 2010. Facilitation drives 65
Temperatures are increasing in many terrestrial systems around the years of vegetation change in the Sonoran desert. Ecology 91, 1132–1139. https://
globe, and plants will either need to track the current climate associated doi.org/10.1890/09-0145.1.
with their home elevation range or persist in novel climates (Alexander Callaway, R.M., 2007. Interaction between competition and facilitation. Positive
Interactions and Interdependence in Plant Communities. pp. 179–254.
et al., 2016). In addition to warming, precipitation is also changing
Cavieres, L.A., Brooker, R.W., Butterfield, B.J., Cook, B.J., Kikvidze, Z., Lortie, C.J.,
worldwide, and many places are already facing reductions in pre- Michalet, R., Pugnaire, F.I., Schöb, C., Xiao, S., Anthelme, F., Bjork, R.G., Dickinson,
cipitation and more variable rainfall patterns (O’Brien et al., 2017a). K.J.M., Cranston, B.H., Gavilán, R., Gutiérrez-Girón, A., Kanka, R., Maalouf, J.P.,
Our results suggest that plants are already moving in the attempt to find Mark, A.F., Noroozi, J., Parajuli, R., Phoenix, G.K., Reid, A.M., Ridenour, W.M.,
Rixen, C., Wipf, S., Zhao, L., Escudero, A., Zaitchik, B.F., Lingua, E., Aschehoug, E.T.,
suitable microhabitats to escape these climatic stresses. Facilitation Callaway, R.M., 2014. Facilitative plant interactions and climate simultaneously
may support biodiversity by providing refugia for plants to persist drive alpine plant diversity. Ecol. Lett. 17, 193–202. https://doi.org/10.1111/ele.
under hotter and drier conditions at their native sites (O’Brien et al., 12217.
Cavieres, L.A., Hernández-Fuentes, C., Sierra-Almeida, A., Kikvidze, Z., 2016. Facilitation
2017b) as well as mediate the movement of species tracking climatic among plants as an insurance policy for diversity in Alpine communities. Funct. Ecol.
conditions. Recent work from Ettinger and HilleRisLambers (2017) 30, 52–59. https://doi.org/10.1111/1365-2435.12545.
showed that interactions between juvenile and adult trees were facil- Chesson, P., 2000. Mechanisms of maintenance of species diversity. Annu. Rev. Ecol. Syst.
31, 343–366.
itative at the upper elevation limits, which improved establishment of Elith, J., Leathwick, J.R., 2009. Species distribution models: ecological explanation and
juveniles. Therefore, it is important to consider facilitation as a factor prediction across space and time. Annu. Rev. Ecol. Evol. Syst. 40, 677–697. https://
promoting community assembly as climate change drives interactions doi.org/10.1146/annurev.ecolsys.110308.120159.
Ettinger, A., HilleRisLambers, J., 2017. Competition and facilitation may lead to asym-
among novel species neighbours (Alexander et al., 2015).
metric range shift dynamics with climate change. Glob. Change Biol. 23, 3921–3933.
https://doi.org/10.1111/gcb.13649.
Filazzola, A., Lortie, C.J., 2014. A systematic review and conceptual framework for the
mechanistic pathways of nurse plants. Glob. Ecol. Biogeogr. 23, 1335–1345. https://
Author contributions
doi.org/10.1111/geb.12202.
Fugère, V., Andino, P., Espinosa, R., Anthelme, F., Jacobsen, D., Dangles, O., 2012.
FIP conceived the experiment, provided community data and con- Testing the stress-gradient hypothesis with aquatic detritivorous invertebrates: in-
tributed to data compilation. MOB analyzed the data and wrote the sights for biodiversity-ecosystem functioning research. J. Anim. Ecol. 81, 1259–1267.
https://doi.org/10.1111/j.1365-2656.2012.01994.x.
manuscript. LFT compiled the data. KAB contributed to data compiling Gross, N., Liancourt, P., Choler, P., Suding, K.N., Lavorel, S., 2010. Strain and vegetation
and provided an initial version of the manuscript. GL performed the effects on local limiting resources explain the outcomes of biotic interactions.
null model testing. All authors contributed to revisions. Perspect. Plant Ecol. Evol. Syst. 12, 9–19. https://doi.org/10.1016/j.ppees.2009.09.

29
M.J. O’Brien, et al. Perspectives in Plant Ecology, Evolution and Systematics 38 (2019) 24–30

001. on Gunung Silam, a small ultrabasic mountain in Sabah, Malaysia. I. environment,

Harrison, S., Cornell, H., 2008. Toward a better understanding of the regional causes of forest structure and floristics. J. Ecol. 76, 320–340. https://doi.org/10.2307/
local community richness. Ecol. Lett. 11, 969–979. https://doi.org/10.1111/j.1461- 2260596.
0248.2008.01210.x. Pugnaire, F.I., Lázaro, R., 2000. Seed bank and understorey species composition in a semi-
Hart, S.P., Marshall, D.J., 2013. Environmental stress, facilitation, competition, and co- arid environment: the effect of shrub age and rainfall. Ann. Bot. 86, 807–813. https://
existence. Ecology 94, 2719–2731. doi.org/10.1006/anbo.2000.1240.
He, Q., Bertness, M.D., Altieri, A.H., 2013. Global shifts towards positive species inter- Pugnaire, F.I., Luque, M.T., 2001. Changes in plant interactions along a gradient of en-
actions with increasing environmental stress. Ecol. Lett. 16, 695–706. https://doi. vironmental stress. Oikos 93, 42–49.
org/10.1111/ele.12080. Pugnaire, F.I., Haase, P., Puigdefábregas, J., Cueto, M., Clark, S.C., Incoll, L.D., 1996.
Hortal, S., Bastida, F., Moreno, J.L., Armas, C., García, C., Pugnaire, F.I., 2015. Benefactor Facilitation and succession under the canopy of a leguminous shrub, Retama
and allelopathic shrub species have different effects on the soil microbial community sphaerocarpa, in a semi-arid environment in south-east Spain. Oikos 76, 455–464.
along an environmental severity gradient. Soil Biol. Biochem. 88, 48–57. https://doi. https://doi.org/10.2307/3546339.
org/10.1016/j.soilbio.2015.05.009. Pugnaire, F.I., Armas, C., Maestre, F.T., 2011. Positive plant interactions in the Iberian
Körner, C., 2007. The use of “altitude” in ecological research. Trends Ecol. Evol. 22, Southeast: mechanisms, environmental gradients, and ecosystem function. J. Arid
569–574. https://doi.org/10.1016/j.tree.2007.09.006. Environ. 75, 1310–1320. https://doi.org/10.1016/j.jaridenv.2011.01.016.
Kraft, N.J.B., Adler, P.B., Godoy, O., James, E.C., Fuller, S., Levine, J.M., 2015. Puigdefábregas, J., Alonso, J.M., Delgado, L., Domingo, F., Cueto, M., Gutiérrez, L.,
Community assembly, coexistence and the environmental filtering metaphor. Funct. Lázaro, R., Nicolau, J.M., Sánchez, G., Solé, A., Vidal, S., Aguilera, C., Brenner, A.,
Ecol. 29, 592–599. https://doi.org/10.1111/1365-2435.12345. Clark, S., Incoll, L., 1996. The Rambla Honda filed site: interactions of soil and ve-
Lázaro, R., Rodrigo, F.S.S., Gutiérrez, L., Domingo, F., Puigdefábregas, J., 2001. Analysis getation along a catena in semi-arid Southeast Spain. Mediterranean Desertification
of a 30-year rainfall record (1967-1997) in semi-arid SE spain for implications on and Land Use. pp. 137–168.
vegetation. J. Arid Environ. 48, 373–395. https://doi.org/10.1006/jare.2000.0755. Ricklefs, R.E., 2004. A comprehensive framework for global patterns in biodiversity. Ecol.
Legendre, P., Anderson, M.J., 1999. Distace-based redundancy analysis: testing multi- Lett. 7, 1–15. https://doi.org/10.1046/j.1461-0248.2003.00554.x.
species responses in multifactorial experiments. Ecol. Monogr. 69, 1–24. https://doi. Ricklefs, R.E., 2008. Disintegration of the ecological community. Am. Nat. 172, 741–750.
org/10.1890/0012-9615(1999)069[0001:DBRATM]2.0.CO;2. https://doi.org/10.1086/593002.
Liancourt, P., Le Bagousse-Pinguet, Y., Rixen, C., Dolezal, J., 2017. SGH: stress or strain Scherrer, D., Körner, C., 2010. Infra-red thermometry of alpine landscapes challenges
gradient hypothesis? Insights from an elevation gradient on the roof of the world. climatic warming projections. Glob. Change Biol. 16, 2602–2613. https://doi.org/10.
Ann. Bot. 120, 29–38. https://doi.org/10.1093/aob/mcx037. 1111/j.1365-2486.2009.02122.x.
López-Pintor, A., Gómez Sal, A., Rey Benayas, J.M., 2006. Shrubs as a source of spatial Schöb, C., Butterfield, B.J., Pugnaire, F.I., 2012. Foundation species influence trait-based
heterogeneity—the case of Retama sphaerocarpa in Mediterranean pastures of central community assembly. New Phytol. 196, 824–834. https://doi.org/10.1111/j.1469-
Spain. Acta Oecologica 29, 247–255. https://doi.org/10.1016/j.actao.2005.11.001. 8137.2012.04306.x.
Lozano, Y.M., Armas, C., Hortal, S., Casanoves, F., Pugnaire, F.I., 2017. Disentangling Schöb, C., Armas, C., Pugnaire, F.I., 2013. Direct and indirect interactions co-determine
above- and below-ground facilitation drivers in arid environments: the role of soil species composition in nurse plant systems. Oikos 122, 1371–1379. https://doi.org/
microorganisms, soil properties and microhabitat. New Phytol. 216, 1236–1246. 10.1111/j.1600-0706.2013.00390.x.
https://doi.org/10.1111/nph.14499. Schöb, C., Callaway, R.M., Anthelme, F., Brooker, R.W., Cavieres, L.A., Kikvidze, Z.,
MacArthur, R.H., 1972. Geographical Ecology: Patterns in the Distribution of Species. Lortie, C.J., Michalet, R., Pugnaire, F.I., Xiao, S., Cranston, B.H., García, M.C., Hupp,
Princeton University Press. N.R., Llambí, L.D., Lingua, E., Reid, A.M., Zhao, L., Butterfield, B.J., 2014a. The
Marini, L., Scotton, M., Klimek, S., Pecile, A., 2008. Patterns of plant species richness in context dependence of beneficiary feedback effects on benefactors in plant facilita-
Alpine hay meadows: local vs. landscape controls. Basic Appl. Ecol. 9, 365–372. tion. J. Physiol. 204, 386–396. https://doi.org/10.1111/nph.12908.
https://doi.org/10.1016/j.baae.2007.06.011. Schöb, C., Prieto, I., Armas, C., Pugnaire, F.I., 2014b. Consequences of facilitation: one
Michalet, R., Brooker, R.W., Lortie, C.J., Maalouf, J.P., Pugnaire, F.I., 2015. Disentangling plant’s benefit is another plant’s cost. Funct. Ecol. 28, 500–508. https://doi.org/10.
direct and indirect effects of a legume shrub on its understorey community. Oikos 1111/1365-2435.12185.
124, 1251–1262. https://doi.org/10.1111/oik.01819. Sexton, J.P., Mcintyre, P.J., Angert, A.L., Rice, K.J., 2009. Evolution and ecology of
Moro, M., Pugnaire, F.I., Haase, P., Puigdefábregas, J., 1997a. Mechanisms of interaction species range limits. Annu. Rev. Ecol. Syst. 40, 415–436. https://doi.org/10.1146/
between a leguminous shrub and its understorey in a semi- arid environment. annurev.ecolsys.110308.120317.
Ecography (Cop.). 20, 175–184. Soliveres, S., Smit, C., Maestre, F.T., 2015. Moving forward on facilitation research: re-
Moro, M., Puignaire, F.I., Haase, P., Puigdefábregas, J., 1997b. Effect of the canopy of sponse to changing environments and effects on the diversity, functioning and evo-
Retama sphaerocarpa on its understorey in a semiarid environment. Funct. Ecol. 11, lution of plant communities. Biol. Rev. Camb. Philos. Soc. 90, 297–313. https://doi.
425–431. https://doi.org/10.1046/j.1365-2435.1997.00106.x. org/10.1111/brv.12110.
Moser, D., Dullinger, S., Englisch, T., Niklfeld, H., Plutzar, C., Sauberer, N., Zechmeister, Thompson, S., Katul, G., 2009. Secondary seed dispersal and its role in landscape orga-
H.G., Grabherr, G., 2005. Environmental determinants of vascular plant species nization. Geophys. Res. Lett. 36, L02402. https://doi.org/10.1029/2008GL036044.
richness in the Austrian Alps. J. Biogeogr. 32, 1117–1127. https://doi.org/10.1111/j. Tirado, R., Pugnaire, F.I., 2005. Community structure and positive interactions in con-
1365-2699.2005.01265.x. straining environments. Oikos 111, 437–444. https://doi.org/10.1111/j.1600-0706.
O’Brien, M.J., Engelbrecht, B., Joswig, J., Pereyra, G., Schuldt, B., Jansen, S., Kattge, J., 2005.14094.x.
Landhäusser, S., Levick, S., Preisler, Y., Väänänen, P., Macinnis-Ng, C., 2017a. A Valiente-Banuet, A., Rumebe, A.V., Verdu, M., Callaway, R.M., 2006. Modern Quaternary
synthesis of tree functional traits related to drought-induced mortality in forests plant lineages promote diversity through facilitation of ancient Tertiary lineages.
across climatic zones. J. Appl. Ecol. 54, 1669–1686. https://doi.org/10.1111/1365- Proc. Natl. Acad. Sci. 103, 16812–16817. https://doi.org/10.1073/pnas.
2664.12874. 0604933103.
O’Brien, M.J., Pugnaire, F.I., Armas, C., Rodríguez-Echeverría, S., Schöb, C., 2017b. The Venable, D.L., Flores-Martinez, A., Muller-Landau, H.C., Barron-Gafford, G., Becerra, J.X.,
shift from plant-plant facilitation to competition under severe water deficit is spa- 2008. Seed dispersal of desert annuals. Ecology 89, 2218–2227. https://doi.org/10.
tially explicit. Ecol. Evol. 7, 2441–2448. https://doi.org/10.1002/ece3.2875. 1890/07-0386.1.
Oksanen, J., Blanchet, F., Kindt, R., Legendre, P., Minchin, P., O’Hara, R., Simpson, G., Whittaker, R.H., Niering, W.A., 1975. Vegetation of the Santa Catalina Mountains,
Solymos, P., Stevens, M., Wagner, H., 2015. vegan: Community Ecology Package. Arizona. V. biomass, production, and diversity along the elevation gradient. Ecology
Pendry, C.A., Proctor, J., 1997. Altitudinal zonation of rain forest on Bukit Belalong, 56, 771–790. https://doi.org/10.2307/1936291.
Brunei: soils, forest structure and floristics. J. Trop. Ecol. 13, 221–241. Wiens, J.J., 2011. The niche, biogeography and species interactions. Philos. Trans. R. Soc.
Prieto, I., Kikvidze, Z., Pugnaire, F.I., 2010. Hydraulic lift: soil processes and transpiration B Biol. Sci. 366, 2336–2350. https://doi.org/10.1098/rstb.2011.0059.
in the Mediterranean leguminous shrub Retama sphaerocarpa (L.) Boiss. Plant Soil Wisz, M.S., Pottier, J., Kissling, W.D., Pellissier, L., Lenoir, J., Damgaard, C.F., Dormann,
329, 447–456. https://doi.org/10.1007/s11104-009-0170-3. C.F., Forchhammer, M.C., Grytnes, J.A., Guisan, A., Heikkinen, R.K., Høye, T.T.,
Prieto, I., Padilla, F.M., Armas, C., Pugnaire, F.I., 2011. The role of hydraulic lift on Kühn, I., Luoto, M., Maiorano, L., Nilsson, M.C., Normand, S., Öckinger, E., Schmidt,
seedling establishment under a nurse plant species in a semi-arid environment. N.M., Termansen, M., Timmermann, A., Wardle, D.A., Aastrup, P., Svenning, J.C.,
Perspect. Plant Ecol. Evol. Syst. 13, 181–187. https://doi.org/10.1016/j.ppees.2011. 2013. The role of biotic interactions in shaping distributions and realised assemblages
05.002. of species: implications for species distribution modelling. Biol. Rev. 88, 15–30.
Proctor, J., Lee, Y.F., Langley, A.M., Munro, W.R.C., Nelson, T., 1988. Ecological studies https://doi.org/10.1111/j.1469-185X.2012.00235.x.

30