GLOBAL CHANGE ECOLOGY - ORI GI NAL RESEARCH

Facilitative interactions do not wane with warming at high

elevations in the Andes
Lohengrin A. Cavieres

Angela Sierra-Almeida
Received: 24 August 2011 / Accepted: 20 March 2012
Springer-Verlag 2012
Abstract Positive interactions between species are
known to play an important role in the structure and
dynamics of alpine plant communities. The balance
between negative and positive interactions is known to
shift along spatial and temporal gradients, with positive
effects prevailing over negative ones as the environmental
stress increases. Thus, this balance is likely to be affected
by climate change. We hypothesized that increases in
temperature (a global warming scenario) should decrease
the importance of positive interactions for the survival and
growth of alpine plant species. To test this hypothesis, we
selected individuals of the native grass species Hordeum
comosum growing within the nurse cushion species
Azorella madreporica at 3,600 m.a.s.l. in Los Andes
(Chile), and performed nurse removal and seedling survival
experiments under natural and warmer conditions. For
warmer conditions, we used open-top chambers, which
increased the temperature by 4 C. After two growing
seasons, we compared the effect of nurse removal on the
survival, biomass, and photochemical efciency of
H. comosum individuals under warmer and natural condi-
tions. Nurse removal signicantly decreased the survival,
biomass, and photochemical efciency of H. comosum,
demonstrating the facilitative effects of nurse cushions.
Seedling survival was also enhanced by cushions, even
under warmer conditions. However, warmer conditions
only partially mitigated the negative effects of nurse
removal, suggesting that facilitative effects of cushions do
not wane under warmer conditions. Thus, facilitative
interactions are vital to the performance and survival of
alpine species, and these positive interactions will continue
to be important in the warmer conditions of the future in
high-alpine habitats.
Keywords Facilitation Positive interactions Climate
change Global warming Nurse effect Cushion plants
Neighbor removal Photochemical efciency Alpine
Introduction
Positive interactions are dened as nontrophic interactions
between species that increase the average individual tness
of at least one of the species involved (Callaway 1995;
Bruno et al. 2003). The most common example of a posi-
tive plantplant interaction is facilitation, where the tness
of one plant species benets from the improved
(micro)environmental conditions created directly or indi-
rectly by neighboring plants, which outweighs the costs of
living close to other individuals (Callaway 2007; Brooker
et al. 2008). According to the stress gradient hypothesis
(SGH), positive interactions are expected to be more fre-
quent or important than negative ones under stressful
conditions that constrain plant performance and survival
(Bertness and Callaway 1994; Brooker and Callaghan
1998; Maestre et al. 2009). Conversely, under low envi-
ronmental severity, the balance between positive and
Communicated by Allan Green.
Electronic supplementary material The online version of this
article (doi:10.1007/s00442-012-2316-x) contains supplementary
material, which is available to authorized users.
L. A. Cavieres (&) A. Sierra-Almeida
ECOBIOSIS, Departamento de Botanica, Facultad de Ciencias
Naturales y Oceanogracas, Universidad de Concepcion,
Concepcion, Chile
e-mail: lcaviere@udec.cl
L. A. Cavieres A. Sierra-Almeida
Instituto de Ecolog a y Biodiversidad (IEB), Santiago, Chile
1 3
Oecologia
DOI 10.1007/s00442-012-2316-x
negative interactions can shift towards negative interac-
tions. Thus, facilitative interactions are expected to be a
frequent and important process driving the compositions
and structures of plant communities in stressful environ-
ments such as arctic and alpine habitats (Bertness and
Callaway 1994; Brooker and Callaghan 1998), and several
studies have demonstrated that facilitation is common
and important in these habitats (e.g., Choler et al. 2001;
Callaway et al. 2002; Cavieres et al. 2002, 2006; Dullinger
et al. 2007).
Global warming is predicted to have large effects on
alpine ecosystems because the rate of warming in these
ecosystems is projected to be two to three times higher than
that recorded during the twentieth century (Guisan and
Theurillat 2000; ACIA 2004; Nogues-Bravo et al. 2007).
The observed upward migration of low-elevation plant
species in some alpine areas has been considered evidence
of the sensitivity of alpine habitats to the warmer condi-
tions that these habitats have been experiencing during the
last few decades (Grabherr et al. 1994; Bahn and Korner
2003; Cannone et al. 2007; Hallinger et al. 2010). If global
warming makes alpine environments less stressful, the
balance between negative and facilitative interactions
among species can be expected to shift, which could alter
the compositions and structures of plant communities.
Thus, understanding the balance between positive and
negative species interactions in warmer conditions could
help to assess the potential effects of climate change on
alpine vegetation dynamics.
Despite the vast area covered by arctic and alpine tun-
dra, and the rapid climatic change that they are experi-
encing, few studies have examined the potential for
changes in the sign of plantplant interactions under higher
temperatures in these habitats (Shetsova et al. 1997;
Hobbie et al. 1999; Dormann et al. 2004; Klanderud 2005;
Klanderud and Totland 2005; Rixen and Mulder 2009). All
previous studies have been conducted in arctic or subarctic
sites of the northern hemisphere, and have yielded con-
trasting results. For example, Klanderud (2005) and
Klanderud and Totland (2005) showed that the presence of
neighbors has negative effects on Carex vaginata and
Thalictrum alpinum, and that warmer conditions make
these interactions even more negative. Hobbie et al. (1999)
did not nd variations in the types of interactions among
plant species with and without experimental warming, while
Rixen and Mulder (2009) and Shetsova et al. (1997) reported
shifts fromfacilitative to negative interactions with warming.
Hence, it is not yet clear how global warming will affect the
balance between positive and negative interactions in cold
climates, particularly in alpine sites where facilitative inter-
actions are very frequent (Callaway et al. 2002).
Cushion-forming plants are one of the most conspicuous
species found in the most exposed alpine habitats, and
several studies have reported plant species growing inside
of cushion plants, suggesting a facilitative role for cushions
(e.g., Pysek and Lyska 1991; Nunez et al. 1999; Arroyo
et al. 2003; Cavieres et al. 2002, 2006; le Roux and
McGeoch 2008). Cushion plants are known for their ability
to transform their immediate microenvironment, because
their low stature and compact form ameliorate extreme
abiotic conditions, allowing the persistence of other species
within them (Cavieres et al. 2006, 2007). However, if
global warming makes alpine habitats less extreme,
according to the SGH, the microclimatic attenuation pro-
vided by cushions should be less relevant for the estab-
lishment and growth of other plant species. In other words,
facilitative interactions are expected to be less crucial
under warmer conditions at alpine sites.
In the present study, we tested this hypothesis by
assessing the effects of simulated global warming on the
relevance of the presence of the nurse cushion species
Azorella madreporica for the photosynthetic performance,
survival, growth, and seedling establishment of Hordeum
comosum in the high alpine zone of the Andes of central
Chile. Recent studies forecast increases of 34 C in the
mean temperature during the next 50 years for this part of
the Andes (CONAMA 2006).
Materials and methods
Study site
This study was carried out at the summit of Mt. Francis-
cano (3600 m a.s.l.; 3320
0
S, 7016
0
W), central Chilean
Andes (33S), roughly 200 m below the upper elevational
limit of plant growth. In this area, the alpine zone extends
from just above the treeline of Kageneckia angustifolia D.
Don (Rosaceae) at 2,200 m to the upper distribution limit
of plants at an elevation of 3,800 m (Cavieres et al. 2000).
The climate in the study area is alpine, with a strong
inuence of the Mediterranean-type climate that predomi-
nates at lower elevations, generating a dry growing season.
At an elevation of 3,600 m the annual precipitation is ca.
900 mm, with much of this precipitation falling as snow
during the winter months of the southern hemisphere
(Santibanez and Uribe 1990). Average air temperatures
follow seasonal variations, with mean temperatures of ca.
-6 C and 6.8 C for winter (JuneAugust) and summer
(DecemberMarch), respectively (Cavieres et al. 2007).
Vegetation cover at the study site is low (\20 %), and it
is dominated by the cushion plant A. madreporica Clos
(Apiaceae). This species forms at and tightly knit cush-
ions with diameters that range from 10 to [100 cm (Faj-
ardo et al. 2008). Other less abundant species, among
which rosette-forming perennial herbs are the most
Oecologia
1 3
common, are: Adesmia capitellata, A. glomerula, Nassau-
via lagascae, Oxalis compacta, Barneoudia major, and
Viola atropurpurea (Cavieres et al. 2000). Given the high
elevation of this site, there are no native herbivores, and
only occasional herbivory by livestock can be observed at
lower elevations on the small peat-bog areas (vegas) that
can be found across the study area.
In previous studies conducted on this site, we have
found that several plant species grow associated with
A. madreporica cushions, suggesting facilitative interac-
tions among them (Badano et al. 2006; Fajardo et al. 2008;
Quiroz et al. 2009). Seedling survival experiments and
measurements of the microclimatic conditions within
cushions and on open areas demonstrated that A. madre-
porica facilitated the establishment of herbaceous species
by increasing the frequency of days and the number of
hours with temperatures above 0 C. For example, Cavi-
eres et al. (2007) reported that, at this site, soil tempera-
tures in bare ground dropped to 0 C or below on 32 % of
the days during the growing season, while the temperature
within A. madreporica cushions fell below zero degrees on
only 2 days. Similar results were reported by Badano et al.
(2006), who reported that the nitrogen and phosphorus
contents in the soil beneath A. madreporica cushions are
higher compared to those in open areas.
Studied species
Hordeum comosum J Presl (Poaceae) is a low-growing
perennial grass of height ca. 1020 cm that is widely dis-
tributed along the southern Andes from central Patagonia
in Argentina to central Chile, where it is particularly
abundant above an elevation of 3,000 m (Nicora 1978). At
the study site, this species usually grows associated with
A. madreporica cushions (Fajardo et al. 2008; Quiroz et al.
2009), where its seedlings have a higher survival rate
compared to open areas (Cavieres et al. 2007). Neverthe-
less, there are individuals growing outside cushions at this
elevation, although at a lower abundance than inside
cushions (Badano et al. 2006).
Experimental design
At the beginning of the growing season (29 December
2007), 20 A. madreporica cushions ca. 1 m in diameter,
each hosting at least one owering H. comosum individual
ca. 10 cm in diameter, were selected. On ten of these
cushions, an open-top chamber (OTC) was placed so as to
completely include the cushion inside the OTC. The other
ten selected cushions were marked with a metallic stick.
The OTCs were hexagonal chambers of transparent
Plexiglass

, with walls 50 cm in height and 120 cm in

diameter and open at the top. The walls of the OTCs were
punched with 25 holes each 1.5 cm in diameter, to allow
some wind to pass through and hence avoid excessive
increases in the air temperature. These OTCs are passive
warming systems that have been widely used in warming
experiments in alpine and arctic tundras (Henry and Molau
1997; Marion et al. 1997).
For each cushion, either with or without an OTC, one
H. comosum individual was selected and marked (n = 20).
Selected H. comosum individuals were of a similar size and
reproductive status (no reproductive buds were present).
Five individuals with OTCs and ve without OTCs were
then randomly assigned to a nurse plant removal experi-
ment, where the aerial portion of their host cushion plant
was removed for 10 cm around the target individual.
Removal was perfomed with a small knife to carefully
remove the aerial part (i.e., aboveground) of the cushion
(see Photo 1 in the Electronic supplementary material).
Using this procedure, we aimed to remove the thermal
refuge effect reported for this cushion species (Cavieres
et al. 2007). Thus, our experimental design was made up of
ve H. comosum individuals growing within cushions
under normal temperature conditions (control), ve indi-
viduals growing under normal conditions but without the
protection of a cushion (nurse removed), ve individuals
growing within cushions under warmer conditions (nurse ?
OTC), and ve individuals growing without cushion pro-
tection but under warmer conditions (nurse removed ?
OTC). This low number of individuals was selected for
conservation purposes, given the low vegetation cover and
the high fragility of the system. The survival of these indi-
viduals was recorded every month until the end of the rst
growing season (March 2008). Two months after the start of
the experiment we quantied the quantum (photochemical)
efciency of the target plants (see below). At the end of the
following growing season (March 2009), we registered the
number of surviving individuals on each experimental con-
dition, and again we quantied their photochemical ef-
ciency. After this, all of the surviving individuals were
carefully removed (both above- and belowground) and
transported to the laboratory, where they were dried for 48 h
at 70 C. For each individual, an entire tiller (i.e., including
its roots) containing an empty spike was removed and
weighed in an analytical balance (Boeco, Hamburg, Ger-
many). Tiller biomass was used as an indicator of the veg-
etative vigor.
Air temperatures were recorded inside and outside the
OTC, both with and without cushions, during the two
growing seasons. HOBO mini-loggers (Onset Corp.,
Bourne, MA, USA) were placed 15 cm above the soil (or
cushion) surface and protected with plastic shelters from
direct exposure to the sun (n = 3 for each condition). Air
sensors were placed at this distance from the soil surface
because the thermal conditions around the plants were
Oecologia
1 3
found to be the same at distances of B15 cm (Sierra-
Almeida et al. 2010) and all individual plants were\15 cm
height, so air sensors were able to capture the thermal
conditions experienced by the experimental individuals.
The mini-loggers were programmed to record the temper-
ature every hour during the entire growing season. For the
rst growing season, air temperature was registered from
17 December 2007 to 29 March 2008, while records started
on 31 December 2008 and ended on 29 March 2009 for the
second growing season.
Air temperature data were used to estimate monthly
growing degree days (GDDs) (McMaster and Wilhelm
1997). GDDs were used as a measure of the accumulated
amount of heat (in C) above a base temperature to rep-
resent a cumulative index of the energy available to
growing plants, according to the formula
GDDs

maximum daily temperature

minimum daily temperature

=2

base temperature:
The daily GDDs were summed per month and then an
averaged monthly GDD was calculated for each condition.
We used 0 C as a conservative base growing temperature
(the temperature above which plants can perform metabolic
functions), because alpine plants generally vary in their
absolute base growing temperature, and this value
encompasses this variability (Korner 2003).
Quantication of the quantum efciency
of photosystem II as an indicator of the individual stress
The uorescence of chlorophyll a associated with photo-
system II (PSII) has developed into a sensitive tool for
probing the state of the photosynthetic apparatus in vivo. It
has become a useful technique for quantifying the stress on
the photosynthetic performance through the photochemical
efciency of PSII under eld conditions (Lambers et al.
2008). Decreases in the photochemical efciency of PSII
indicate restrictions on or damage to the photosynthetic
apparatus that compromise carbon gain, so they are con-
sidered one of the rst manifestations of stress at the
individual level (Maxwell and Johnson 2000). On 1 March
2008, we quantied the quantum efciency of PSII (Fv/Fm
and UPSII) as an indicator of the stress experienced by
H. comosum individuals growing under the different
experimental conditions studied. Fv/Fm indicates the
maximal quantum efciency of PSII, while UPSII indicates
the effective quantum efciency of PSII under ambient
light conditions (Schreiber et al. 1995). Fully developed
leaves on each H. comosum individual were dark adapted
with leaf clips provided by the manufacturer of the uo-
rometer (to obtain open PSII centers) for 30 min before
measurements in order to ensure maximum photochemical
efciency. Fluorescence signals were generated by a pulse-
amplitude modulated uorometer (FMS 2, Hansatech
Instruments Ltd., Kings Lynn, UK). The minimum uo-
rescence (F0) with all PSII reaction centers in the open
state was determined by applying weak modulated light
(0.4 lmol m
-2
s
-1
). The maximum uorescence (Fm)
with all PSII reaction centers in the closed state was induced
by a 0.8 s saturating pulse of white light (9,000 lmol
m
-2
s
-1
). After 15 s, the actinic light (180 lmol m
-2
s
-1
)
was turned on and the same saturating pulse described
previously was applied every 60 s until steady-state photo-
synthesis was reached, in order to obtain Fs and Fm
0
. Finally,
F0
0
was measured after turning the actinic light off. Mea-
surements were done at noon, under clear-sky conditions, so
all experimental individuals were exposed to similar solar
radiation conditions. As mentioned, these measurements were
repeated using the same protocol at the end of the following
growing season (31 March 2009).
Seedling survival
During December 2008, we randomly selected 12 A. mad-
reporica cushions and 12 bare ground areas separated by at
least 2 m from the nearest cushion to perform seedling
survival experiments. Six of the selected A. madreporica
cushions and six of the bare ground areas were randomly
assigned to a warming treatment, where an OTC was
placed around each cushion and bare ground area. These
cushions (with and without OTCs) were different to those
used in the removal experiment.
Seeds of H. comosum collected during March 2008 were
germinated in growth chambers under controlled temper-
ature conditions (20/10 C day/night) during October 2008.
Emerging seedlings were planted in small plastic bags
(100 cm
3
) with commercial soil, and were maintained in a
growth chamber at 10/5 C (day/night) for one month. We
kept the seedlings in their bags during transportation to the
site, and seedlings were acclimatized to the site over two
nights before planting. Ten one-month-old seedlings were
planted in six replicates of each treatment. Seedlings were
planted on 28 December 2008 in two rows separated by
5 cm (to minimize between-seedling competition). Seed-
ling survival was monitored fortnightly from January to
March 2009.
Statistical analyses
Differences in air temperatures (i.e., mean, max, min, and
GDDs) among experimental conditions were assessed by
paired Student t tests. Quantum efciencies of photosystem
II and the biomasses of tillers of H. comosum individuals at
each experimental treatment were compared with two-way
ANOVAs where the presence/absence of an OTC and a
Oecologia
1 3
nurse were considered xed factors. Data were checked for
normality before analyses. Survival curves of seedlings at
each experimental treatment were estimated by means of
the KaplanMeier method, and statistical differences were
assessed with the CoxMantel test (Fox 1993). For all of
the analyses, we used the software R v.2.3 (R Development
Core Team 2005).
Results
Air temperature
Both with and without cushion plants, the OTCs increased the
mean temperature of the air during both growing seasons
(Table 1). Whilst outside the cushions the OTCs increased the
air temperature by ca. 4 K, within the cushions the OTCs
warmed the air by ca. 2 K (Table 1). During both growing
seasons, the OTCs signicantly increased the maximum
temperatures but also signicantly affected the minimum
temperatures by slightly decreasing them by \1.0 C
(Table 1). Nevertheless, estimations of the accumulated
amount of heat indicated that during the two growing seasons,
the OTCs signicantly increased the GDDs both inside and
outside the cushions (Table 1), with the cushions showing
slightly higher GDDs than the open areas (Table 1).
Survival and nal biomass
After two growing seasons, all of the H. comosum indi-
viduals growing within cushions under control conditions
had survived. However, when the nurse cushion was
removed, only three out of the ve experimental individ-
uals survived (Fig. 1a). Survival with an OTC was 100 %,
with or without the removal of the nurse.
The results of the ANOVA indicated that only the
removal of the nurse plants had a signicant effect on tiller
biomass (F
1,14
= 9.98; p = 0.007); neither OTC presence/
absence (F
1,14
= 0.47; N.S.) nor the interaction factor
(F
1,14
= 0.32; N.S.) had signicant effects. Removal of the
nurse, under both natural and OTC conditions, signicantly
decreased tiller biomass, although this decrease was lower
with an OTC (Fig. 1b).
Quantum efciency of photosystem II
During the rst growing season, only nurse removal had a
signicant effect on the maximal quantum efciency of
photosystem II (Fv/Fm) (Table 2), with lower Fv/Fm val-
ues observed after nurse removal under natural conditions,
but not under warmer conditions (Fig. 2). In contrast, both
the presence of an OTC and nurse removal signicantly
affected the effective quantum efciency of PSII (UPSII)
(Table 2). Removal of the nurse decreased UPSII both with
and without an OTC (Fig. 2), whilst the presence of an
OTC signicantly increased the UPSII both in plants
growing with nurse and in those with the nurse removed
(Fig. 2). Interestingly, these patterns were completely
consistent with those observed at the end of the second
growing season (Fig. 2; Table 2), although only three
individuals survived after the removal of the nurse without
an OTC.
Table 1 Air temperature (15 cm height) inside and outside the OTCs placed within and outside Azorella madreporica cushions during two
growing seasons in the central Chilean Andes at 3,600 m elevation
Outside cushion Inside cushion
Control Warming Control Warming
Growing season 20072008
Length of the growing season (days) 101
Mean air temperature (C) 5.6 0.3
a
9.4 0.5
b
5.4 0.3
a
7.7 0.5
c
Minimum air temperature (C) -2.8 0.5
a
-3.0 0.5
a
-3.0 0.4
a
-3.7 0.4
b
Maximum air temperature (C) 15.8 0.7
a
28.3 1.1
b
15.7 0.7
a
25.2 0.8
c
Mean monthly GDD (C day
-1
) 165.8 51
a
322.1 50.2
b
181 56.9
a
260.2 48
b
Growing season 20082009
Length of the growing season (days) 87
Mean air temperature (C) 7.0 1.6
a
10.8 1.7
b
6.7 0.3
a
8.1 0.5
c
Minimum air temperature (C) -1.3 0.4
a
-3.1 0.4
b
-1 0.4
a
-4.8 0.4
c
Maximum air temperature (C) 17.8 0.6
a
30.8 0.9
b
16.6 0.7
c
25.4 0.7
d
Mean monthly GDD (C day
-1
) 242.6 23.9
a
407.8 53.3
b
288.3 22.3
a
302 28.3
c
Values correspond to mean 2SE
Different lower-case letters indicate signicant differences according to the paired Student t test (p \0.05) between treatments
Oecologia
1 3
Seedling survival
Under natural conditions, the survival of seedlings of
H. comosum planted within cushions was signicantly
higher than that of seedlings planted on the bare ground
(CoxMantel test = 8.93, p \0.0001; Fig. 3). Likewise,
under warmer conditions, seedlings growing in association
with cushion plants showed signicantly higher survival
than seedlings on bare ground (CoxMantel test = 4.97,
p \0.0001; Fig. 3). While survival on the bare ground was
the same regardless of OTC presence (CoxMantel
test = 1.58, p = 0.11, Fig. 3), survival within cushions
with an OTC was lower compared to that without an OTC
(CoxMantel test = 3.47, p = 0.0005, Fig. 3).
Discussion
Previous studies in alpine habitats have shown that positive
associations between cushion plants and other plant species
are particularly frequent at higher elevations (e.g., Cavieres
et al. 2002; Arroyo et al. 2003; le Roux and McGeoch
2008; Antonsson et al. 2009). Experimental studies con-
ducted on some of these sites demonstrated that these
positive associations are due to microclimatic mitigations
provided by cushion plants, which facilitate the establish-
ment and growth of other plants (Cavieres et al. 2006,
2007, 2008). In agreement with previous studies, our
results clearly show that the removal of A. madreporica
cushions had negative effects on the survival, biomass
accumulation, and photochemical performance of the
perennial native grass H. comosum. In addition, planted
seedlings of this species show low survival without the
presence of the nurse, conrming the facilitative interac-
tion among these species in this high-elevation environ-
ment. Successful seedling establishment is a key process in
the population dynamics of alpine plant species (Forbis
2003; Venn and Morgan 2009), so cushion plants play a
C
o
n
tro
l
N
u
rse
re
m
o
ve
d
O
T
C
N
u
rse
re
m
o
ve
d
+
O
T
C
S
u
r
v
i
v
a
l

(
%
)
0
20
40
60
80
100
Treatment
C
o
n
tro
l
N
u
rse
re
m
o
ve
d
O
T
C
N
u
rse
re
m
o
ve
d
+
O
T
C
B
i
o
m
a
s
s

(
g
)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
a a
ab
b
A
B
Fig. 1 Survival rate (a) and tiller biomass (b) for Hordeum comosum
individuals during two growing seasons growing within Azorella
madreporica cushions under natural temperature conditions (Con-
trol), with the cushion removed under natural temperature conditions
(Nurse removed), within cushions under warmer temperatures (OTC),
and with the cushion removed under warmer temperatures (Nurse
removed ? OTC). See Methods for details. Error bars indicate
2SE
Table 2 ANOVA tables for the effects of nurse removal and warmer
temperatures (presence of OTC) on the maximal and effective
quantum efciencies of photosystem II (Fv/Fm and UPSII, respec-
tively) in Hordeum comosum plants during two growing seasons
(20072008 and 20082009) in the High Andes of central Chile
Effect SS df MS F p
Season 20072008
Fv/Fm
Nurse removal 0.008 1 0.008 5.50 0.032
Presence of OTC 0.007 1 0.007 4.38 0.053
Interaction 0.005 1 0.005 3.09 0.097
Error 0.024 16 0.002
UPSII
Nurse removal 0.027 1 0.008 6.43 0.022
Presence of OTC 0.047 1 0.007 11.12 0.004
Interaction 0.002 1 0.005 0.54 0.474
Error 0.004 16 0.002
Season 20072008
Fv/Fm
Nurse removal 0.012 1 0.012 22.49 \0.001
Presence of OTC 0.015 1 0.015 27.42 \0.001
Interaction 0.007 1 0.007 13.19 0.003
Error 0.007 14 0.0001
UPSII
Nurse removal 0.008 1 0.008 8.60 0.001
Presence of OTC 0.006 1 0.006 6. 82 0.020
Interaction 0.0002 1 0.0002 0.24 0.633
Error 0.013 14 0.0009
Oecologia
1 3
crucial role in the dynamics and diversity of these eco-
systems (Cavieres and Badano 2009).
Based on the expectation derived from the SGH,
some authors have predicted a general shift in species
interactions for arctic and alpine tundras from positive to
negative as these environments warm (e.g., Klanderud
2005; Klanderud and Totland 2005). Here we found that,
under warmer conditions, the presence of the nurse cushion
mostly continued to show positive effects. In other words,
although OTCs provided thermal conditions similar to
those provided by cushions (i.e., GDDs), alleviating one
stress factor, the presence of the nurse continues to be
important to the performance of alpine species.
Although survival did not decrease after nurse removal
under warmer conditions, the biomass of these individuals
was lower compared to the biomass of those growing
within cushions, whether under normal or warmer condi-
tions. This suggests that while warmer conditions might
facilitate survival, the presence of the nurse is important for
growth. It has been previously reported (e.g., Nunez et al.
1999; Cavieres et al. 2006, 2008; Yang et al. 2010;
Anthelme et al. 2012) that soils beneath cushions contain a
higher concentration of nutrients than those of bare ground.
Cushions can trap their own litter and intercept wind-blown
litter from other sources, which, in combination with the
higher availability of water and the milder temperatures
observed within cushions (Cavieres et al. 2007), could
enhance the activities of nitrogen-xing organisms,
resulting in the locally increased concentrations of soil
F
v
/

F
m
0.55
0.60
0.65
0.70
0.75
0.80
0.85
C
o
n
tro
l
N
u
rse
re
m
o
ve
d
O
T
C
N
u
rse
re
m
o
ve
d
+
O
T
C
P
S

I
I
0.55
0.60
0.65
0.70
0.75
0.80
0.85
C
o
n
tro
l
N
u
rse
re
m
o
ve
d
O
T
C
N
u
rse
re
m
o
ve
d
+
O
T
C
a
a a
a
b
a
a
b
a
a
b
a
a
a
b
a

Fig. 2 Maximal (Fv/Fm) and effective (UPSII) quantum efciencies

of photosystem II, as assessed by the chlorophyll uorescence of
Hordeum comosum individuals growing within Azorella madreporica
cushions under natural temperature conditions (Control), with the
cushion removed under natural temperature conditions (Nurse
removed), within cushions under warmer temperatures (OTC), and
with the cushion removed under warmer temperatures (Nurse
removed ? OTC). Measurements were taken two months after
removal (left panels) and at the end of the following growing season
(right panels). Error bars indicate 2SE, and different letters indicate
signicant differences according to Tukeys HSD a posteriori test
Time (Days)
0 20 40 60 80 100 120
S
e
e
d
l
i
n
g

s
u
r
v
i
v
a
l

(
%
)
0
20
40
60
80
100
Nurse + OTC
Nurse
Bare ground + OTC
Bare ground
b
c
c
a
Fig. 3 KaplanMeyer estimated survival of experimentally planted
seedlings of Hordeum comsum within cushions of Azorella monantha
under normal temperature conditions (Nurse), within cushions under
warmer temperatures (Nurse ? OTC), on bare ground under normal
temperature conditions (Bare ground), and on bare ground under
warmer conditions (Bare ground ? OTC)
Oecologia
1 3
nutrients that have been observed beneath cushions at this
site (Badano et al. 2006). Casanova-Katny et al. (2011)
showed that for several species (including H. comosum)
sampled at our study site, the mycorrhization level of
individuals growing in association with A. madreporica
cushions was six times higher than in individuals of the
same species growing on bare ground. In addition, Venn
et al. (2009) have shown the importance of neighbors in
decreasing the frequency and intensity of frost heave
events in the soil, which represent one the most important
sources of seedling mortality in high-alpine habitats. Thus,
considering the key role of cushions in providing more
benign conditions for plant nutrition and substrate stability,
the presence of these nurses is and will be important, even
under warmer conditions.
The effective quantum efciency of PSII in the presence
of ambient light (UPSII) decreased with the removal of the
nurse under both control and warmer conditions, indicating
the positive effect of cushions on the photosynthetic per-
formance of the target species. Whilst H. comosum indi-
viduals growing under warmer conditions without a nurse
had similar UPSII values to those growing within a nurse
under control conditions, the highest UPSII values were
obtained in those individuals growing within a nurse under
warmer conditions. These results suggest that warmer
conditions, whether provided by the presence of the nurse
or by an OTC, are important for high photochemical ef-
ciency. The lower Fv/Fm values observed only after nurse
removal under control but not warmer conditions also
suggest the importance of warmer conditions for complete
recovery of open PSII reaction centers. Although alpine
plants are adapted to photosynthesize at low temperatures
(Friend and Woodward 1990; Korner 2003), during the
frequent overcast conditions that occur at high elevations,
photosynthesis is co-limited by low temperature and low
irradiance, and under these conditions warmer tempera-
tures might signicantly improve the carbon gain (Korner
2003). The combination of freezing temperatures during
the night and a high intensity of light during the morning is
characteristic of alpine habitats (Korner 2003), and the
combination of low temperatures and high solar irradiance
has been associated with low-temperature photoinhibition
(Streb et al. 1997; Germino and Smith 2000, 2001; Cui
et al. 2003). Warmer conditions such as those experienced
by H. comosum plants inside OTCs might facilitate the
recovery of the photosynthetic apparatus after photoinhi-
bition caused by the combination of freezing temperatures
during the night and high irradiance in the morning (Cui
et al. 2003). Nevertheless, the higher UPSII values
obtained under warmer conditions within the nurse suggest
synergistic effects between both factors (i.e., warmer
temperatures and presence of the nurse), which could be
related to the higher soil nutrient availability and myco-
rrhization provided by the nurse cushion plants.
Although seedling survival within cushions decreased
under warmer conditions, it was still signicantly higher
than that observed in warmer open areas. This indicates
that although the facilitative effect of cushions on seedling
survival decreased under warmer conditions (as predicted
by the stress gradient hypothesis), it did not disappear,
suggesting that the importance of facilitative interactions
provided by cushion plants might be a more important
phenomenon in warmer alpine environments than previ-
ously thought.
Few studies have tested for temporal or spatial varia-
tions in the sign of the interactions among plant species
under a global warming scenario in tundra habitats. For
example, Dormann et al. (2004) reported that the herba-
ceous Luzula confusa and the prostrate shrub Salix polaris
showed competition in an arctic tundra, and this increased
after experimental warming in the eld (see also Klanderud
2005; Klanderud and Totland 2005 for similar results).
However, Shetsova et al. (1997) reported that the positive
effects of elevated temperature on the growth and repro-
duction of Vaccinium vitis-idaea and Empetrum nigrum in
subarctic Finland were amplied by the presence of
neighborsin this instance, a reduced level of environ-
mental severity (warming) led to an increased impact of
facilitative interactions. In a North American arctic tundra
system where articially advanced snowmelt (a phenome-
non expected under global warming) created a more severe
environment through increased early-season frost events,
neighbors increasingly facilitated the survival, growth, and
reproduction of E. nigrum (Wipf et al. 2006). It is notable
that the increased role of facilitation with decreasing
severity found by Shetsova et al. (1997) occurs within an
environment that is apparently less severe than, for
example, the open tundra systems studied by Dormann
et al. (2004), and in which the opposite pattern was
observed. Such apparently contradictory ndings indicate
that the perception of environmental severity is to some
extent species specic (Rixen and Mulder 2009). These
ndings and our results support the argument that the
relationship between environmental severity and the rela-
tive role of interactions is more complicated than set out in
the original SGH (see Michalet et al. 2006); that is, alle-
viation of stress does not decrease the relevance of positive
interactions. Improving our knowledge of the variation in
interactions under different environmental conditions
among different species, growth forms, or strategies (e.g.,
competitive vs. stress-tolerant strategies) and how these
differ from trends at the community level will aid the
development of a general theory that helps to explain such
apparently contradictory results (Brooker 2006).
Oecologia
1 3
In contrast to expectations derived from the SGH, we
found that the positive impacts of cushion nurse plants did
not weaken under warmer conditions in a high-alpine site.
Recent studies have shown that climate change will reduce
biodiversity in tundra regions (e.g., Klein et al. 2004;
Walker et al. 2006). However, if nurse species can mitigate
the effects of climate change, they might play an increas-
ingly pivotal role in the maintenance of diversity, sug-
gesting a potential role for facilitation as a mediator of
global changes in alpine habitats.
Acknowledgments We thank the staff of La Parva and Valle
Nevado Ski Resort for their help with the access to our study sites.
We also thank Victor and Angelica Rojas from Valparaiso Lodge, our
second home. This study was supported by FONDECYT 1060710 and
1090389. Additional support was provided by projects P05-002 F
ICM of the Millennium Scientic Initiative of the Ministry of
Economy, Public Works, and Tourism of the Chilean Government,
and CONICYT PFB-23, supporting the Center for Advanced Studies
in Ecology and Research on Biodiversity (IEB).
References
ACIA (2004) Arctic climate impact assessment. Cambridge Univer-
sity Press, Cambridge
Anthelme F, Buendia B, Mazoyer C, Dangles O (2012) Unexpected
mechanisms sustain the stress gradient hypothesis in a tropical
alpine environment. J Veg Sci 23:6272
Antonsson H, Bjork RG, Molau U (2009) Nurse plant effect of the
cushion plant Silene acaulis (L.) Jacq. in an alpine environment
in the subarctic Scandes, Sweden. Plant Ecol Divers 2:1725
Arroyo MTK, Cavieres LA, Penaloza A, Arroyo-Kalin MA (2003)
Positive associations between the cushion plant Azorella
monantha (Apiaceae) and alpine plant species in the Chilean
Patagonian Andes. Plant Ecol 169:121129
Badano EI, Jones CG, Cavieres LA, Wright JP (2006) Assessing
impacts of ecosystem engineers on community organization: a
general approach illustrated by effects of a high-Andean cushion
plant. Oikos 115:369385
Bahn M, Korner C (2003) Recent increases in summit ora caused by
warming in the Alps. Ecol Stud 167:437441
Bertness MD, Callaway RM (1994) Positive interactions in commu-
nities. Trends Ecol Evol 9:191193
Brooker RW (2006) Plantplant interactions and environmental
change. New Phytol 171:271289
Brooker RW, Callaghan TV (1998) The balance between positive and
negative interactions and its relationship to environmental
gradient: a model. Oikos 81:196207
Brooker RW, Maestre FT, Callaway RM, Lortie CL, Cavieres LA,
Kunstler G, Liancourt P, Tielborger K, Travis JMJ, Anthelme F,
Armas C, Coll L, Corcket E, Delzon S, Forey E, Kikvidze Z,
Olofsson J, Pugnaire FI, Quiroz CL, Saccone P, Schiffers K,
Seifan M, Touzard B, Michalet R (2008) Facilitation in plant
communities: the past, the present and the future. J Ecol
96:1834
Bruno JF, Stachowicz JJ, Bertness MD (2003) Inclusion of facilitation
into ecological theory. Trends Ecol Evol 18:119125
Callaway RM (1995) Positive interactions among plants. Bot Rev
61:306349
Callaway RM (2007) Positive interactions and interdependence in
plant communities. Springer, Berlin
Callaway RM, Brooker RW, Choler P, Kikvidze Z, Lortie CJ,
Michalet R, Paolini L, Pugnarie FI, Newingham B, Aschehoug
ET, Armas C, Kikodze D, Cook BJ (2002) Positive interactions
among alpine plants increase with stress. Nature 417:844848
Cannone N, Sgorbati S, Guglielmin M (2007) Unexpected impact of
climate change on alpine vegetation. Front Ecol Environ
5:360364
Casanova-Katny MA, Torres-Mellado A, Goetz P, Cavieres LA
(2011) The best for the guest: high Andean nurse cushions of
Azorella madreporica enhance arbuscular mycorrhizal status in
associated plant species. Mycorrhiza 21:613622
Cavieres LA, Badano EI (2009) Do facilitative interactions increase
species richness at the entire community level? J Ecol
97:11811191
Cavieres LA, Penaloza A, Arroyo MTK (2000) Altitudinal vegetation
belts in the high Andes of central Chile (33S). Revista Chilena
de Historia Natural 73:331344
Cavieres LA, Arroyo MTK, Penaloza A, Molina-Montenegro M,
Torres C (2002) Nurse effect of Bolax gummnifera cushion
plants in the alpine vegetation of the Chilean Patagonian Andes.
J Veg Sci 13:547554
Cavieres LA, Badano EI, Sierra-Almeida A, Gomez-Gonzalez S,
Molina-Montenegro M (2006) Positive interactions between
alpine plant species and the nurse cushion plant Laretia acaulis
do not increase with elevation in the Andes of central Chile. New
Phytol 169:5969
Cavieres LA, Badano EI, Sierra-Almeida A, Molina-Montenegro M
(2007) Microclimatic modications of cushion plants and their
consequences for seedlings survival of native and non-native
plants in the high-Andes of central Chile. Arct Antarct Alp Res
39:229236
Cavieres LA, Quiroz CL, Molina-Montenegro MA (2008) Facilitation
of the non-native Taraxacum ofcinale by native nurse cushion
species in the high-Andes of central Chile: are there differences
between nurses? Funct Ecol 22:148156
Choler P, Michalet R, Callaway RM (2001) Facilitation and
competition on gradients in alpine plant communities. Ecology
82:32953308
CONAMA (2006) Estudio de la variabilidad climatica en Chile para
el siglo XXI. Facultad de Ciencias F sicas y Matematicas,
Universidad de Chile, Departamento de Geof sica, Santiago
Cui X, Tang Y, Gu S, Nishimura S, Shi S, Zhao X (2003)
Photosynthetic depression in relation to plant architecture in
two alpine herbaceous species. Environ Exp Bot 50:125135
Dormann C, Van der Wal R, Woodin SJ (2004) Neighbour identity
modies effects of elevated temperature on plant performance in
the high Arctic. Glob Change Biol 10:15871598
Dullinger S, Kleinbauer I, Pauli H, Gottfried M, Brooker R, Nagy L,
Theurillat JP, Holten JI, Abdaladze O, Benito JL, Borel JL,
Coldea G, Ghosn D, Kanka R, Merzouki A, Klettner C, Moiseev
P, Molau U, Reiter K, Rossi G, Stanisci A, Tomaselli M,
Unterlugauer P, Vittoz P, Grabherr G (2007) Weak and variable
relationships between environmental severity and small-scale co-
occurrence in alpine plant communities. J Ecol 95:12841295
Fajardo A, Quiroz CL, Cavieres LA (2008) Spatial patterns in cushion-
dominated plant communities of the high Andes of central Chile:
how frequent are positive associations? J Veg Sci 19:8796
Forbis TA (2003) Seedling demography in an alpine ecosystem. Am J
Bot 90:11971206
Fox GA (1993) Failure-time analysis: emergence, owering survi-
vorship, and other waiting time. In: Scheiner SM, Gurevitch J
(eds) Design and analysis of ecological experiments. Chapman
and Hall, New York, pp 253266
Friend AD, Woodward FI (1990) Evolutionary and ecophysiological
responses of mountain plants to the growing season environ-
ment. Adv Ecol Res 20:59124
Oecologia
1 3
Germino MJ, Smith WK (2000) High resistance to low-temperature
photoinhibition in two alpine, snowbank species. Physiol Plant
110:8995
Germino MJ, Smith WK (2001) Relative importance of microhabitat,
plant form and photosynthetic physiology to carbon gain in two
alpine herbs. Funct Ecol 15:243251
Grabherr G, Gottfried M, Pauli H (1994) Climate effects on mountain
plants. Nature 369:448
Guisan A, Theurillat JP (2000) Assessing alpine vulnerability to
climate change, a modelling perspective. Integr Assess
1:307320
Hallinger M, Manthey M, Wilmking M (2010) Establishing a missing
link: warm summers and winter snow cover promote shrubs
expansion into alpine tundra in Scandinavia. NewPhytol 186:890
899
Henry GHR, Molau U (1997) Tundra plants and climate change: the
International Tundra Experiment (ITEX). Glob Change Biol
3:19
Hobbie SE, Shevtsova A, Chapin FS (1999) Plant responses to species
removal and experiment warming in Alaskan tussock tundra.
Oikos 84:417434
Klanderud K (2005) Climate change effects on species interactions in
an alpine plant community. J Ecol 93:127137
Klanderud K, Totland O (2005) The relative importance of neigh-
bours and abiotic environmental conditions for population
dynamic parameters of two alpine plant species. J Ecol 93:493
501
Klein JA, Harte J, Zhao ZQ (2004) Experimental warming causes
large and rapid species loss, dampened by simulated gazing, on
the Tibetan plateau. Ecol Lett 7:11701179
Korner C (2003) Alpine plant life (2nd ed). Springer, Berlin
Lambers H, Chapin FS, Pons TL (2008) Plant physiological ecology,
2nd edn. Springer, New York
le Roux PC, McGeoch MA (2008) Spatial variation in plant
interactions across a severity gradient in the sub-Antarctic.
Oecologia 155:831844
Maestre FT, Callaway RM, Valladares F, Lortie C (2009) Rening
the stress-gradient hypothesis for competition and facilitation in
plant communities. J Ecol 97:199205
Marion GM, Henry GHR, Freckman DW, Johnstone J, Jones G, Jones
MH, Levesque E, Molau U, Molgaard P, Parsons AN, Svoboda J,
Virginia RA (1997) Open-top designs for manipulating eld
temperature in high-latitude ecosystems. Glob Change Biol
3:2032
Maxwell K, Johnson GN (2000) Chlorophyll uorescencea
practical guide. J Exp Bot 51:659668
McMaster GS, Wilhelm WW (1997) Growing degree-days: one
equation, two interpretations. Agric For Meteorol 87:291300
Michalet R, Brooker RW, Cavieres LA, Kikvidze Z, Lortie CJ,
Pugnaire FI, Valiente-Banuet A, Callaway RM (2006) Do biotic
interactions shape both sides of the humped-back model of
species richness in plant communities? Ecol Lett 9:767773
Nicora EG (1978) Graminae. In: Correa MN (ed) Flora Patagonica,
Tomo VIII, Parte III. Coleccion Cient ca del Instituto Nacional
de Tecnolog a Agropecuaria (INTA), Buenos Aires
Nogues-Bravo D, Araujo MD, Errea MP, Mart nez-Rica JP (2007)
Exposure of global mountain systems to climate warming during
the 21st century. Glob Environ Change 17:420428
Nunez C, Aizen M, Ezcurra C (1999) Species associations and nurse
plant effects in patches of high-Andean vegetation. J Veg Sci
10:357364
Pysek P, Lyska J (1991) Colonization of Sibbaldia tetrandra cushions
on alpine scree in the Palmiro Alai mountains, Central Asia. Arct
Alp Res 23:263272
Quiroz CL, Badano EI, Cavieres LA (2009) Floristic changes induced
by cushion species Azorella madreporica at two contrasting
elevations. Revista Chilena de Historia Natural 82:171184
Rixen C, Mulder CPH (2009) Species removal and experimental
warming in a subarctic tundra plant community. Oecologia
161:173186
Santibanez F, Uribe JM (1990) Atlas agroclimatico de Chile.
Universidad de Chile, Facultad de Ciencias Agrarias y Fores-
tales, Santiago
Schreiber U, Bilger W, Neubauer C (1995) Chlorophyll uorescence
as a non-intrusive indicator for a rapid assessment of in vivo
photosynthesis. In: Schulze E-D, Caldwell MM (eds) Ecophys-
iology of photosynthesis. Springer, Berlin, pp 4970
Shetsova A, Haukioja E, Ojala A (1997) Growth response of subarctic
dwarf shrubs, Empetrum nigrum and Vaccinium vitis-idaea, to
manipulated environmental conditions and species removal.
Oikos 78:440458
Sierra-Almeida A, Cavieres LA, Bravo LA (2010) Freezing resistance
of high-elevation plant species is not related to their height or
growth-form in the Central Chilean Andes. Environ Exp Bot
69:273278
Streb P, Feierabend J, Bligny R (1997) Resistance to photoinhibition
of photosystem II and catalase and antioxidative protection in
high mountain plants. Plant Cell Environ 20:10301040
Venn S, Morgan JW (2009) Patterns in alpine seedling emergence and
establishment across a stress gradient of mountain summits in
south-eastern Australia. Plant Ecol Divers 2:516
Venn S, Morgan JW, Green PT (2009) Do facilitative interactions
with neighboring plants assist the growth of seedlings at high
altitudes in Alpine Australia? Arct Antarct Alp Res 41:381387
Walker MD, Wahren CH, Hollister RD (2006) Plant community
responses to experimental warming across the tundra biome.
Proc Natl Acad Sci USA 103:13421346
Wipf S, Rixen C, Mulder CPH (2006) Advanced snowmelt causes
shift toward positive neighbour interactions in a subarctic tundra
community. Glob Change Biol 12:111
Yang Y, Niu Y, Cavieres LA, Sun H (2010) Positive associations
between the cushion plant Arenaria polytrichoides (Caryophyll-
aceae) and other alpine plant species increases with altitude in
the Sino-Himalayas. J Veg Sci 21:10481057
Oecologia
1 3