Content uploaded by Aitor Ameztegui
Author content
All content in this area was uploaded by Aitor Ameztegui on Jan 08, 2019
Content may be subject to copyright.
Unraveling the role of light and biotic interactions on seedling
performance of four Pyrenean species along environmental gradients
Aitor Ameztegui
a,
⇑
, Lluís Coll
a,b
a
Forest Sciences Center of Catalonia (CTFC), Ctra. Sant Llorenç de Morunys km.2, E-25280 Solsona, Spain
b
CREAF, Centre for Ecological Research and Forestry Applications, Autonomous University of Barcelona, Bellaterra E-08193, Catalonia, Spain
article info
Article history:
Received 11 January 2013
Received in revised form 12 March 2013
Accepted 8 April 2013
Available online 6 May 2013
Keywords:
Seedlings
Plant–plant interactions
Elevational gradient
Climate change
Pyrenees
Facilitation
abstract
The predicted upward displacement of forest species due to climate warming is expected to be modu-
lated by a medley of abiotic and biotic factors acting at microsite level. Species-specific differences in
plant responses to this set of environmental factors can thus have strong implications in the future
dynamics of forest ecosystems. To gain a better understanding of the main fine-scale factors and pro-
cesses driving present and future species performance in the montane and subalpine belt of the Eastern
Pyrenees (NE Spain), we established a set of experimental mixed plantations along elevational and envi-
ronmental gradients using the four tree species dominating these areas (Pinus sylvestris,Pinus uncinata,
Abies alba and Betula pendula). Once the plantations had been established, the performance and growth
of 72 seedlings of each species was monitored and linear and non-linear models were fitted to identify
the main factors controlling their survival and growth.
We found most of the mortality to occur during the third growing season, following a harsh winter and
a drought period during summer. Mortality patterns were highly species- and site-specific. At the subal-
pine belt, shrubs were found to have a facilitative effect on winter survival of P. sylvestris (mortal-
ity < 10%) but not on the other species. At the montane belt, A. alba mortality during the summer
increased in areas with high light exposure and herbaceous cover (mortality > 30%). All species except
P. uncinata showed lower height growth at high elevation, with differences between sites matching dif-
ferences in growing season duration (20%).
Our results underline the strong impact that short periods of extreme climate can have in the perfor-
mance of plants developing in mountainous areas far from their optimal elevational range. However, they
also underline a potentially critical role played by biotic and abiotic microsite factors in mediating species
responses to these climatic events.
Ó2013 Elsevier B.V. All rights reserved.
1. Introduction
The predicted increase in temperatures caused by global warm-
ing (IPCC, 2007) is expected to have large effects on mountain eco-
systems, where the elevational ranges of trees are mainly
controlled by temperature (Grabherr et al., 1994; Walther et al.,
2002; Peñuelas and Boada, 2003; Lenoir et al., 2008). Accordingly,
most simulations based on ‘climate envelopes’ predict upward or
poleward displacement of species under future warming scenarios
(Guisan et al., 1998; Dullinger et al., 2004), with some species even
becoming extinct if the rate of change exceeds their pace of biolog-
ical response (Thomas et al., 2004; Thuiller et al., 2005). However,
realized ecological niches are multidimensional, and species distri-
bution is not only explained by macro-climate but also by species-
specific responses to a medley of abiotic and biotic factors that of-
ten operate at finer temporal and spatial scales (Vetaas, 2002; Wal-
ther et al., 2002; Dullinger et al., 2004; Holtmeier and Broll, 2005).
The effects of these factors are often not adequately captured by
the ‘climate envelope’ models (Ackerly et al., 2010; Scherrer and
Körner, 2011), sometimes leading to unrealistic predictions of spe-
cies distributional changes (Randin et al., 2009; Willis and Bhag-
wat, 2009; Martínez et al., 2012).
In many mountain ecosystems, abiotic stresses are considered
the major mechanism setting the upper limit of species’ elevational
ranges. To successfully migrate upwards, tree species must be able
to grow and survive outside their current elevational range, thus
facing climatic conditions that are at the limits of their physiolog-
ical tolerance (Lenoir et al., 2009, 2010), particularly at their youn-
ger stages (Germino et al., 2002; Gómez-Aparicio et al., 2008a).
Therefore, in these areas, short periods of extreme climatic condi-
tions (e.g. extreme cold or freezing events) play a critical role in
shaping future species composition (Schneider, 2004; Lindner
et al., 2010). However, some recent studies have shown that mi-
0378-1127/$ - see front matter Ó2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.foreco.2013.04.011
⇑
Corresponding author. Tel.: +34 973481752.
E-mail address: aitor.ameztegui@ctfc.cat (A. Ameztegui).
Forest Ecology and Management 303 (2013) 25–34
Contents lists available at SciVerse ScienceDirect
Forest Ecology and Management
journal homepage: www.elsevier.com/locate/foreco
cro-topographical factors can buffer or even override the effects of
harsh climate on plant survival (Ackerly et al., 2010; Scherrer and
Körner, 2011). Positive plant–plant interactions could also play a
major role in these stressful environments, as postulated by the
‘stress-gradient hypothesis’ (SGH; (Bertness and Callaway, 1994;
Maestre et al., 2009). Shrubs for example are known to facilitate
the survival of young plants of species growing at their upper ele-
vational limit by reducing wind abrasion, mitigating the minimum
temperatures to which they are exposed, protecting them from
radiation, or reducing snowdrift (Batllori et al., 2009; Barbeito
et al., 2012).
Conversely, competitive effects are considered the major mech-
anism setting the lower limit of species’ elevational ranges (Lenoir
et al., 2010). However, this might not occur when environmental
severity increases at low elevation. In such cases, species’ sensitiv-
ity to the main stress factor (e.g. drought in Mediterranean moun-
tains) would significantly define its lower range margin (Macias
et al., 2006; Candel-Pérez et al., 2012; Linares and Camarero,
2012). Other processes not directly related to climate, such as hab-
itat modification due to changes in human activities have led to se-
vere canopy closure in many European forests (Poyatos et al., 2003;
Ameztegui et al., 2010), which can also induce significant shifts in
species distribution, even favoring downslope displacement of
some shade-tolerant species (Lenoir et al., 2010; Bodin et al., 2013).
In the Pyrenees, the rise in temperatures associated to climate
change has prompted predictions of species displacements (Resco
et al., 2006) with a progressive upward encroachment of montane
species into the subalpine belt (Ameztegui and Coll, 2011). How-
ever, these areas are characterized by a high variability of abiotic
and biotic factors acting at local scale, whose role in these appar-
ently climate-driven processes remains unclear. With the ultimate
aim of advancing understanding of the relative impact of these fac-
tors in the future dynamics of the Pyrenean mountainous forests,
we set up a 4-year field experiment in which seedlings of the 4
most widespread tree species in the montane and subalpine belt
of the Eastern Pyrenees (Pinus sylvestris L., Pinus uncinata Ram,
Abies alba Mill. and Betula pendula Roth.) were planted along gradi-
ents of elevation and microsite conditions (light availability and
herbaceous and shrub cover). Specifically, we aimed to answer
the following questions: (i) how does the performance (survival
and growth) of these species vary along environmental gradients
including variation in climate, light availability and biotic interac-
tions?; (ii) what role do short extreme climatic events play in seed-
ling survival and growth?; (iii) are the intensity and sign of biotic
interactions (competitive vs. facilitative) species-specific?; and (iv)
can plant-plant interactions favor or limit species range expansion
by modulating the effects of climate change through facilitation
and competition?
We hypothesized that climate (minimum temperatures at high
elevation and drought at low elevation) would play a determinant
role in seedling mortality, but that biotic interactions could par-
tially buffer this effect, especially for those species established far
from their optimal elevational range.
2. Methods
2.1. Study area
The experiment was conducted at two different elevations in
the northern slopes of the Serra del Cadí, a Pyrenean mountain
range in the Cadí-Moixerò Natural Park (42°17
0
N; 1°42
0
E). The
‘‘low-elevation’’ plots were located in the montane belt, slightly
below the P. sylvestris–P. uncinata transition zone (around
1500 m a.s.l.). The forest overstory in this site was dominated by
P. sylvestris (more than 75% of total basal area) with some P. unci-
nata and B. pendula individuals, whereas the main species in the
understory were common box (Buxus sempervirens L.) and common
juniper (Juniperus communis L.). The ‘‘high-elevation’’ plots were lo-
cated in the subalpine belt (around 2000 m a.s.l.), close to the opti-
mum elevational distribution for P. uncinata. In this site, the
overstory was overwhelmingly dominated by P. uncinata, although
some disperse individuals of A. alba and B. pendula could be found.
The shrub layer was almost exclusively composed by J. communis.
These two areas (located in the same valley but set 12 km apart)
present contrasting climates associated to the abrupt terrain
involving marked elevational zonation of the vegetation (Ninot
et al., 2007).
2.2. Species selection and characteristics
This study focused on the 4 most widespread tree species in the
area: Scots pine (P. sylvestris), a shade-intolerant species that in the
Pyrenees can be found between 1000 and 1800 m a.s.l., thus dom-
inating the montane belt of the Pyrenees; mountain pine (P. unci-
nata), a shade-intolerant conifer that reaches its southern
distributional limit in the Iberian Peninsula and constitutes most
of the treelines in the Pyrenees as it is restricted to the subalpine
belt (between 1600 and 2300 m a.s.l.); silver fir (A. alba), the most
shade-tolerant conifer species in the Pyrenees, usually restricted to
humid sites on north-facing, shady slopes between 1200 and
2000 m a.s.l. where the risk of water stress in summer is lower
(Macias et al., 2006); and silver birch (B. pendula), a shade-intoler-
ant pioneer species that usually colonizes disturbed areas between
1000 and 1800 m a.s.l. but only rarely reaches the canopy (Ruiz de
la Torre, 2006). These species differ in their ecological require-
ments, and they can be ordered from most (rank = 5) to least
(rank = 1) shade tolerant following the ranking system developed
by Niinemets and Valladares (2006):A. alba (4.6 ± 0.06;
mean ± SE), B. pendula (2.03 ± 0.09), P. sylvestris (1.67 ± 0.33), and
P. uncinata (1.2). Moreover, the drought tolerance ranking order
according to the same authors would be: P. sylvestris
(4.34 ± 0.47); P. uncinata (3.88), B. pendula (1.85 ± 0.21) and A. alba
(1.81 ± 0.28). Despite their different ecological requirements, these
four species are able to coexist in a strip between 1600 and
2000 m a.s.l. constituting the local montane-subalpine mixed
forest.
2.3. Experimental design
At each one of the two sites, 72 two-year-old seedlings of each
of the 4 studied species were distributed into 12 plots and planted
in early summer 2008. Plot size varied between 40 and 50 m
2
and
each plot included 24 seedlings (6 per species) planted at least one
meter apart to avoid interaction between them. Half of the plots
were located in the forest understory and the other half in natu-
rally-occurring gaps (Fig. 1). Gap surface was between 150 and
350 m
2
(196.0 ± 33.4 m
2
; mean ± SE). Seedlings were randomly
distributed across each plot, and were carefully planted to mini-
mize alteration of the micro-environment. As domestic herds of
cattle and horses are led into these forests during the summer,
the seedlings were protected from browsing with an individual
protector (90 cm height and 33 cm of diameter) with a
20 20 mm mesh net (Nortène, Lille, France) to exclude the influ-
ence of animal damage. All the seedlings were 2 years old at the
time of planting and had been grown in a local nursery (Forestal
Catalana, Pobla de Lillet, Spain) from seeds collected in neighboring
forests. For all species, seedling source, nursery and plantation area
were all inside the same region of provenance (Alía et al., 2005).
26 A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34
2.4. Characterization of the environment and explanatory variables
To better characterize climatic differences between the two
sites, two meteorological stations were installed (one per site),
and air temperature (at 1 m height), below-ground soil tempera-
ture (at 10 cm depth) and precipitation were measured continu-
ously using ECH2O sensors (Decagon Devices, Pullman, WA, USA;
see Table 1). Data from the in situ meteorological stations regis-
tered high climatic variability over the 4 years of the study. In Au-
gust and early September 2011, there was a rather warm and dry
period (with 60% less precipitation than the average for the last
10 years, Fig. 2) that exposed the vegetation to a significant
drought stress that was visually appreciable even in the adult
stand. The winters of 2009 and 2010 were particularly cold, with
minimum temperatures reaching 15.8 °C in the high-elevation
sites. In early winter 2010, the cold period coincided with low pre-
cipitations, resulting in the shallowest snow layer seen in recent
years (Fig. 2). Throughout the duration of the study, the seedlings
established at high elevation (subalpine sites) were exposed to
lower mean temperatures, higher precipitations, higher Thorn-
thwaite index and a 20% shorter growing period than the seedlings
established at lower elevation (Table 1).
To characterize the microsite conditions of the planted seed-
lings, light availability, percentage of herbaceous cover and dis-
tance to nearest shrub were measured for each of the seedlings
(Table 1). Light availability was measured using two Li-190SA
quantum sensors (Li-COR, Lincoln, NE, USA). The sensors were used
in paired mode, i.e. one of the sensors was placed at the top of each
seedling and the other in an adjacent open area, following standard
procedure (see Messier and Puttonen (1995) and Parent and Mess-
ier (1996) for a complete description of the method). This approach
Table 1
Main abiotic and biotic characteristics of the studied stands.
Variable Montane sites Subalpine sites
Latitude (N)/longitude (E) 42°19
0
/1°43
0
42°18
0
/1°42
0
Elevation (m a.s.l.) 1550 1955
Aspect/slope (degrees) NE/39 NE/53
Bedrock Limestone Limestone
Mean annual/summer temperature (°C) 7.4/14.8 4.9/11.7
Total annual/summer precipitation (mm) 992/271 1118/327
Thornthwaite index
a
70.3 120.7
Length of the growing season
b
(days) 194 147
Mean summer maximum temperature (°C) 21.0 17.3
Mean winter minimum temperature (°C) 3.4 4.6
Dominant species
c
Ps, Pu, Bp Pu, Aa, Bp
Light availability (%PPFD) 19.3 ± 9.8
[2.8–55.1]
23.9 ± 13.3
[5.4–58.6]
Herbaceous cover (%) 39.1 ± 23.1
[0–95]
51.1 ± 30.6
[0–100]
Distance to nearest shrub (cm) 56.5 ± 45.8
[10–350]
118.9 ± 76.3
[20–430]
Values for light availability, herbaceous cover, and distance to nearest shrub are
means ± SD (n= 286). Values in brackets are minimum and maximum observed
values. The number of decimal positions indicates precision of the variable when
measured.
a
Calculated as the ratio between precipitation and potential evapotranspiration
(Thornthwaite et al., 1957).
b
Calculated as in Körner and Paulsen (2004).
c
Listed for each site in decreasing order of dominance: Pu: Pinus uncinata; Aa:
Abies alba; Ps: Pinus sylvestris; Bp: Betula pendula.
Fig. 1. Representation of the experimental design and the measured abiotic and biotic variables. At each of the two sites (montane and subalpine belt), 6 plots were located in
the forest understory and 6 in naturally-occurring gaps. At each plot, 24 seedlings (6 per species) were planted, and environmental variables at the microsite scale (light
availability, herbaceous cover, and distance to nearest shrub) were measured.
A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34 27
makes it possible to calculate light availability as a percentage of
transmitted photosynthetic photon flux density (%PPFD), and con-
sequently ranged from 0 (complete obstruction) to 100 (open sky).
The measurements showed mean light availability to be slightly
higher in the subalpine plantations, but the range was similar at
both elevational sites, with a maximum of about 60% (Table 1). Per-
centage of herbaceous cover surrounding the seedlings was visu-
ally estimated to the nearest 5% using an 80 80 cm grid
centered on each seedling, and was found to be higher overall at
the subalpine belt (although extremely variable, with values rang-
ing from 0% to 90–95% at both sites). Finally, distance from each
seedling to the nearest shrub was measured with a tape meter.
Fig. 2. Evolution of precipitation, monthly average of the maximum (TMAX) and minimum (TMIN) temperatures and snow depth in the study area over the 4-year study
period (2008–2011). Dashed lines and shaded areas indicate mean ± SD during the last 10 years, whereas solid lines indicate measured values. Data are from the Prat d’Aguiló
meteorological station (2138 m a.s.l.), located less than 1 km away from the high-elevation site. Vertical arrows indicate the extreme climatic events that occurred during
late-autumn 2010 and summer 2011 (see text for further details).
28 A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34
We defined a shrub as any woody plant with several stems arising
from the base, so other elements such as neighbor seedlings, adult
trees, logs or rocks were not considered. The montane sites showed
higher shrub density, resulting in a lower mean distance of seed-
lings to the nearest shrub (Table 1).
2.5. Seedling monitoring
Seedling mortality was regularly monitored throughout the
duration of the experiment. With the aim of disentangling the
main climatic variables driving mortality for each species at each
elevation, we divided the observed mortality into either summer
mortality (occurring during the growing season) or winter mortal-
ity (occurring during the winter). The plots were frequently visited,
and we only assigned mortality to a given period (summer or win-
ter) when plants were found to be dead in the beginning of a period
but had been recorded as healthy (absence of any symptom of dis-
ease) at the end of the previous one. Despite the protective net, se-
ven seedlings were damaged by animals, mainly by trampling, and
were subsequently excluded from the analysis. Furthermore, 13
seedlings died due to small disturbances (e.g. stones falling, small
landslides, etc.) and were also excluded, giving a total of 552 seed-
lings analyzed.
Seedling size was monitored by measuring total height and
diameter at the root collar at the end of every growing season.
Using this dataset, we determined the following response vari-
ables: (1) survival rate along the 39 months of the study; (2) winter
and summer mortality; (3) height at the end of the study period;
and (4) diameter at the end of the study period.
2.6. Data analyses
For each species and site (montane vs. subalpine), survival func-
tion curves were developed based on Kaplan–Meier estimates, and
the Mantel–Cox log-rank test was used to determine significant
differences between sites. To test the effect of categorical or con-
tinuous covariates on our censored survival data, we used a Mixed
Effects Cox model (Therneau and Grambsch, 2000), which is a
modification of the commonly used Cox’s Proportional Hazards
(coxPH) model (Cox, 1972) that allows for inclusion of random
covariates. For each combination of response variable (summer
and winter mortality) and species, we fitted a separate model in
which the effects of the three explanatory variables at microsite le-
vel (herbaceous cover, distance to shrubs, and light availability)
and elevation (site) were introduced as fixed factors, whereas plot
was introduced as a random factor. The equation fitted by the
model was:
kðtÞ¼k
0
ðtÞe
XbþZb
bG0;XðhÞ
where k
0
(t) is an unspecified baseline hazard function, Xand Zare
the design matrices for the fixed and random effects, respectively,
bis the vector of fixed-effects coefficients, and b is the vector of ran-
dom effects coefficients. When comparing two groups, the hazard
ratio (e
b
) is the quotient of the hazard functions for each of the
groups. For a continuous variable, the hazard ratio indicates the
change in the risk of mortality if the parameter in question rises
by one unit. The random effects distribution Gis modeled as Gauss-
ian with mean zero and a variance matrix
R
, which in turn depends
on a vector of parameters h. To test the significance of each variable,
we performed a likelihood ratio test (LRT) to compare deviances of a
pair of nested models: a null model (in which the variable was ab-
sent) and an alternative model including it. Interactions between
variables were only included if the LRT of the model indicated sig-
nificant difference with both the simpler models. Goodness-of-fit of
the models was assessed through the concordance statistic (C),
Fig. 3. Survival curves for seedlings of the four species of trees planted at the two experimental sites over the course of the study period, based on Kaplan–Meier estimates.
Black and grey lines represent seedlings in the montane and subalpine belts, respectively. Solid lines represent Kaplan–Meier estimates whereas dashed lines are 95%
confidence intervals. P-values indicate significance of the log-rank test between sites for each species. Shaded areas in the x-axis correspond to the vegetative period (from
May to October). Note that the y-axis starts at 0.5 for greater clarity and easier comparison among species.
A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34 29
which is analogous to Kendall’s tau between the prediction and the
outcome but can be used with censored data.
Several model formulations (linear, exponential, power and
Michaelis–Menten; see Appendices for details on the model equa-
tions) were used to analyze the effect of the different explanatory
variables (light availability, herbaceous cover and distance to near-
est shrub) on seedling size (height and diameter) at the end of the
study period (39 months). Initial seedling size was introduced as a
covariate with an exponent
a
to allow for non-linear relationships
between final and initial size. The maximum likelihood parameter
values were estimated using simulated annealing (Goffe et al.,
1994) and the asymptotic 2-unit support intervals were used to as-
sess the strength of evidence for individual maximum likelihood
parameter estimates. The R
2
of the regression of observed vs. pre-
dicted values provided a measure of the goodness-of-fit of each
model, and alternative models were compared using
D
AICc, the
difference in corrected Akaike information criterion (Burnham
and Anderson, 2002). Following the likelihood approach, we used
comparison of alternate models to test our hypotheses. For each
explanatory variable, the best of the 4 formulations (linear, expo-
nential, power and Michaelis–Menten) was compared (in terms
of
D
AICc) to a null model in which there was no effect of the inde-
pendent variable, and we considered that an effect of this variable
was supported when
D
AICc > 2. When there was substantial sup-
port for more than one independent variable, we also tested a
bivariate model in which both variables were included, and the
strength of evidence for this model was compared with univariate
models. Finally, when the best model had been identified, we also
tested for an effect of elevation by comparing the strength of evi-
dence of a model in which parameters were estimated separately
for low- and high-elevation sites against another model in which
all the data were pooled together, and we retained the first model
only if it was substantially supported by the data (
D
AICc > 2).
All analyses were performed using R 2.14.1 software (R Devel-
opment Core Team, 2011) and the ‘likelihood’ v. 1.5 (Murphy,
2012), ‘survival’ v. 2.37–2 (Therneau, 2011) and ‘coxme’ v.2.2–3
(Therneau, 2012) packages for R.
3. Results
3.1. Seedling survival
Survival rate after three growing seasons was significantly dif-
ferent among species (P< 0.001) and ranged from more than 80%
for P. uncinata to less than 65% for P. sylvestris. For the three conifer
species (A. alba,P. sylvestris and P. uncinata), most of the mortality
at both elevations (more than 60% of total dead seedlings) was ob-
served during the third year. Between-site differences in survival
rates were only significant for P. sylvestris (P< 0.001) and A. alba
(P= 0.044; Fig. 3). These species presented opposite patterns, with
P. sylvestris showing higher mortality in subalpine sites and A. alba
showing higher mortality in montane sites. The seasonal patterns
of mortality were also species-specific. While most of the P. unci-
nata and A. alba mortality occurred during the growing season,
for the other two species it occurred during the winter (Fig. 3).
P. sylvestris was the only species that showed an effect of the
explanatory variables on winter mortality. The winter mortality
of P. sylvestris was found to be positively associated with elevation
(hazard ratio [HR] = 4.13; 95% CI: 2.15–7.95; P< 0.001) and with
seedling distance to the nearest shrub (HR = 1.004; 95% CI:
1.002–1.007; P< 0.001; Fig. 4). At the subalpine belt, mortality rate
for seedlings planted at less than 0.5 m from a shrub was 0.11,
whereas mortality rate for the rest of the seedlings ranged between
0.28 and 0.44. The positive effect of the shrubs on winter survival
was not as marked at the montane belt, where mortality rates ran-
ged from 0 to 0.13, but we did not find a significant effect of the
interaction between distance to nearest shrub and elevation
(P= 0.112). For the other three species (P. uncinata,B. pendula
and A. alba), we could not detect an effect of any of the explanatory
variables (including elevation) on winter mortality (Appendix A).
A. alba was the only species that showed a significant effect of
the analyzed abiotic and biotic factors on summer mortality. For
Fig. 4. Predicted variation in the log hazard ratio for winter mortality as a function
of distance to nearest shrub for P. sylvestris seedlings planted at montane (black)
and subalpine (grey) sites in the Eastern Pyrenees. Solid lines represent predicted
models whereas shaded areas correspond to 95% confidence intervals.
Fig. 5. Predicted variation in the log hazard ratio for summer mortality as a
function of (a) light availability and (b) percentage of herbaceous cover, for A. alba
seedlings planted at montane (black) and subalpine (grey) sites in the Eastern
Pyrenees. In (a), no significant effect of site was detected, and pooled data for both
sites are represented in dark grey. Solid lines represent predicted models whereas
shaded areas correspond to 95% confidence intervals.
30 A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34
A. alba, both herbaceous cover and light availability were positively
associated with mortality. The analyses also revealed a significant
correlation between both variables (Pearson correlation coeffi-
cient = 0.21 for the montane belt; 0.42 for the subalpine belt) and
the bivariate model including them was not significantly better
than the univariate models. Herbaceous cover showed the most
significant effect on mortality (P< 0.001), and we found a signifi-
cant interaction between this variable and site (P< 0.001): the ef-
fect of herbaceous cover on mortality was only significant at the
montane belt, where the observed mortality rate ranged from
0.08 for low classes of herbaceous cover (<20%) to 0.33 for seed-
lings planted in sites with more than 80% herbaceous cover
(HR = 1.120; 95% CI: 1.047–1.198; P< 0.001; Fig. 5). Light availabil-
ity was also positively associated with a higher mortality rate
(HR = 1.056; 95% CI: 1.017–1.097; P= 0.036), with a maximum of
0.25 for seedlings with more than 40% PPFD available, but without
significant effect of elevation in this trend (P= 0.882; Fig. 5). An ef-
fect of herbaceous cover in the mortality of B. pendula was also de-
tected, although the model including this variable was only
marginally significant compared to the null model (HR = 1.020;
95% CI: 0.998–1.043; P= 0.055; Appendix B).
3.2. Seedling growth
There were significant among-species differences in absolute
height growth, which ranged from 15 mm y
1
for A. alba to more
than 70 mm y
1
for B. pendula (Table 3). These differences were
weaker when considering relative height growth, but B. pendula
still had higher growth than both pines, which in turn grew faster
in height than A. alba. All species showed higher relative height
growth at low-elevation sites, but the differences were not signif-
icant for P. uncinata (P = 0.73). In contrast, all species showed high-
er relative radial growth at the subalpine site, but the differences
were only significant for the two pine species (Table 3). However,
measured radial growth over the 3 years was rather low (annual
relative radial growth <0.1 mm for all species), and we discarded
it from further analyses as these values were below the measure-
ment accuracy.
Light availability was the main environmental factor controlling
height of the three conifer species (P. uncinata,P. sylvestris and A.
alba). For A. alba and P. sylvestris, the effect of light availability on
growth was modulated by elevation, with seedling growth show-
ing lower response to enhanced light (or almost null response in
the case of A. alba) at low elevation than at high elevation
(Fig. 6). For A. alba, the observed response was linear, whereas
for P. sylvestris it followed a power equation (Table 2;Fig. 6). These
differences in response with elevation were not observed for P.
uncinata, for which relative height growth increased with light
availability but was saturated at.% PPFD > 40%, following a Michae-
lis–Menten equation (Table 2;Fig. 6). For these species, the expo-
nent of the initial height (introduced as a covariate in the model)
ranged between 0.4 and 0.6, thus indicating a non-linear relation
between this variable and final height. In contrast, the relation be-
tween initial and final height growth was linear for B. pendula
Fig. 6. Predicted variation in height as a function of initial height and light availability for seedlings of the 4 studied species planted at montane (black) and subalpine (grey)
sites in the eastern Pyrenees. When no effect of elevation was predicted, data were pooled together, and symbols are represented in dark grey. Solid lines represent predicted
models whereas shaded areas correspond to 2-unit support intervals. Horizontal lines indicate lack of effect of the predictive variable for that species, and are shown for
comparative purposes.
A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34 31
(
a
= 1.01), which was the only species for which there was no
detectable effect of microsite factors or elevation on height
(Appendix C).
4. Discussion
Mortality events in the established plots took place at different
periods of the year in the two studied elevations. At the subalpine
belt, a majority of mortality occurred during winter, whereas at the
montane belt a majority of mortality occurred during the growing
season. Although mortality was relatively low for all species during
the first two years after plantation, it sharply increased during the
third year when the area was exposed to a particularly cold late-
autumn and early-winter followed by a drought episode during
summer. The potentially strong impact of short events of extreme
climatic conditions on juvenile tree mortality has already been ob-
served elsewhere (Schneider, 2004; Saccone et al., 2009; Lindner
et al., 2010), and in our sites we observed marked species-specific
responses to these events. The harsh winter particularly affected P.
sylvestris plants growing in the subalpine belt, where this species
ranked as the least adapted to the climatic conditions, whereas
as expected, summer drought mainly affected A. alba, which is
the least drought-tolerant of the four studied species. Interestingly,
the negative effect of drought in A. alba survival was only observed
in the montane sites.
Positive plant-plant interactions were found to be species-spe-
cific and to vary with elevation, being particularly important for
the survival of species establishing at the extremes of their eleva-
tional range. This was the case for P. sylvestris seedlings, a species
typical of the montane belt and consequently the least adapted to
the harsh climate of the high-elevation plots. Our results agreed
with the stress-gradient hypothesis (Bertness and Callaway,
1994) and are consistent with previous research showing facilita-
tion to be more important for species planted at experimental sites
located at higher elevations than their distributional mean (Batllori
et al., 2009). Our results also stress the major role that facilitation
could play in modulating the effects of extreme climatic events
such the extremely cold late-autumn and winter of 2010 in our
experimental sites (Brooker et al., 2008; Saccone et al., 2009). We
did not find a facilitative effect of shrubs on B. pendula, A. alba or
P. uncinata plants. For the first two species, this could be explained
by the fact that they are less drought-tolerant than pines (in partic-
ular A. alba), and competition for water from the neighboring
shrubs may probably overcome their positive nurse effect. As
pointed out by Maestre et al. (2009), in extremely severe environ-
mental conditions, resource uptake by facilitators can overcome
their positive effect if the stress is resource-based and the benefi-
ciary species are not stress-tolerant. The lack of facilitative effects
of P. uncinata was not surprising, given that this species is the best
adapted to climate at the subalpine belt and presented relatively
low overall mortality over the course of the experiment.
At the montane belt, most of the mortality occurred during
summer and more specifically during the third year’s growing sea-
son when the area was exposed to a major drought episode. This
event mainly affected A. alba, a species well-known to be highly
sensitive to water deficit, being negatively affected by high tem-
perature conditions and the related drought stress (Rolland et al.,
1999; Pagès et al., 2003; Peguero-Pina et al., 2007; Toromani
et al., 2011). The positive effect of light closure on A. alba survival
might indicate a facilitative effect of tree cover on A. alba seedlings
by reducing the Vapor Pressure Deficit to which the seedlings were
exposed and thus indirectly limiting the development of competi-
tive herbaceous neighbors (Pagès et al., 2003; Saccone et al., 2009).
This indirect facilitative effect of canopy cover seemed to be partic-
ularly important in our study sites given the marked negative rela-
tionship found between herbaceous cover and A. alba survival, that
was probably associated to competition for water. The herbaceous
cover also increased mortality of B. pendula in the montane belt,
although this effect was only marginally significant. In general, B.
pendula is considered to tolerate drought much better than A. alba
but shows a lower ability to compete for water than pines (Prevo-
sto and Balandier, 2007).
In the montane site, we expected to find a positive effect of
shrubs on plant survival (at least for the most drought-sensitive
species) after the marked drought period that occurred during
the third growing season. In the drier areas of Mediterranean
mountains, seedlings frequently benefit from habitat amelioration
by shrubs which reduce the radiation and temperature to which
they are exposed and thereby improve their water status (Castro
et al., 2002, 2004; Gómez-Aparicio et al., 2004, 2008b) However,
we did not detect any positive effect of shrubs on plant survival,
indicating that in this relatively mesic area, the net effect of the
and negative interactions occurring above- and below-ground be-
tween the shrubs and the seedlings was neutral.
Interestingly, we did not observe a higher mortality of P. unci-
nata in the montane belt, where this species was planted below
its current elevational range, and neither biotic nor abiotic factors
exerted any influence on the rate of P. uncinata mortality. Although
the lower limit of species’ elevational ranges in Mediterranean
mountains is often considered to be set by drought-induced stress
(Macias et al., 2006; Candel-Pérez et al., 2012; Linares and Camare-
ro, 2012), our results suggest that this might not be the case for P.
uncinata. It is possible that the current low elevational limit of P.
uncinata is not climatically-driven but is instead set by other fac-
tors, such as competitive interactions with low-elevation species,
mainly P. sylvestris (Callaway et al., 2002; Ameztegui and Coll,
2011).
In contrast to the abovementioned species-specific and tempo-
ral patterns, all the species studied grew faster in height in the
montane belt than in the subalpine belt. Growth is known to be
mainly limited by duration of the growing season in high-elevation
forests (Grace and Norton, 1990; Grace et al., 2002; Hoch and Kör-
ner, 2003; Cailleret and Davi, 2010). Here, the average reduction in
growth (20%) found in P. sylvestris, A. alba and B. pendula seedlings
matched the average difference in length of the growing season ob-
Table 2
Summary of the models predicting seedling height as a function of microsite
conditions and initial height for seedlings of four species of trees planted at two
experimental sites in the Eastern Pyrenees. For each species and explanatory variable,
the AIC of the best-fit model is provided for all seedlings pooled together (AIC
P
) and
separated into two groups (montane vs. subalpine, AIC
S
).
Height Model AIC
P
AIC
S
R
2
Pinus uncinata
Light MM 2113.58 2123.75 0.33
Herb. cover Lin 2136.17 2141.73 0.21
Null Null 2207.46 2201.72 0.12
Abies alba
Light Lin 1669.49 1673.84 0.41
Herb. cover Lin 1676.75 1679.01 0.30
Null Null 1688.31 1688.16 0.25
Pinus sylvestris
Light Pow 2067.94 2055.57 0.36
Herb. cover Lin 2076.28 2063.71 0.22
Null Null 2085.66 2078.33 0.18
Betula pendula
Null Null 1771.06 1775.33 0.22
Factors are ranked from highest to lowest support according to AIC
P
. For each
species, only models with stronger empirical support than the null model (i.e. an
AIC
P
at least two units lower) are provided. Models are: Lin = linear; Exp = expo-
nential, Pow = power, MM = Michaelis–Menten; see Appendix C for details on the
equations. Lower AIC
S
values than AIC
P
values indicate stronger empirical support
for separated data than for pooled data, i.e. an elevational effect.
32 A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34
served between both sites during the experiment (Table 3). In con-
trast, P. uncinata (the species currently dominating the subalpine
belt) presented the same height growth–light relationship at both
elevations, following the typical saturating curve for environments
where light is the most limiting factor (Ameztegui and Coll, 2011).
However, the height growth of all species increased with higher
light availability, thus indicating that the microsite requirements
for seedling survival were different from those required for growth,
as observed in the Alps by Barbeito et al. (2012).
In summary, this study showed that short periods of extreme
climate can have a strong impact on the mortality of species grow-
ing far from their mean elevational range (e.g. P. sylvestris in the
subalpine belt; A. alba at the montane belt). We found that positive
plant–plant interactions can play a critical role in mediating the ef-
fects of these unfavorable climate conditions on the performance
of these species when growing above their current limits. In con-
trast, in the lower limit of species’ elevational ranges, competition
plays a more important role. Overall, we conclude that species-spe-
cific differences in performance under different environmental
conditions and the role of plant–plant interactions should be
explicitly considered when making predictions of climate
change-induced species displacement and when designing or
implementing management plans to contend with the impacts of
climate change.
Acknowledgements
This research was primarily supported by the Spanish Ministry
of Science and Innovation via the projects Consolider-Ingenio Mon-
tes (CSD2008-00040), DINAMIX (AGL2009-13270-C02) and RESIL-
FOR (AGL2012-40039-C02-01). The Spanish Ministry of Science
provided LC with support through a Ramon y Cajal contract
(RYC-2009-04985), and the Spanish Ministry of Education pro-
vided AA with support through a predoctoral grant (FPU Program
– AP2007-01663). The authors are particularly grateful to A. Barg-
ués, S. Martín, L. Ivorra and M. Pallarés for their invaluable work
during field sampling and laboratory processing. F. Cano helped
us to find the most suitable forests for this study, and the Socarrel
team kindly offered their facilities during the fieldwork stage. We
also thank the ‘Parc Natural del Cadí-Moixeró’ and the ‘Montellà i
Martinet’ municipality for kindly giving permission to access the
park and sample the data. Three anonymous reviewers provided
helpful comments and suggestions.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.foreco.2013.
04.011.
References
Ackerly, D.D., Loarie, S.R., Cornwell, W.K., Weiss, S.B., Hamilton, H., Branciforte, R.,
Kraft, N.J.B., 2010. The geography of climate change: implications for
conservation biogeography. Diversity and Distributions 16, 476–487.
Alía, R., Alba, N., Agúndez, D., Iglesias, S., 2005. Manual para la comercialización y
producción de semillas y plantas forestales. Materiales de base y de
reproducción. Serie Forestal. DGB., Madrid. 384 pp.
Ameztegui, A., Coll, L., 2011. Tree dynamics and co-existence in the montane-sub-
alpine ecotone: the role of different light-induced strategies. Journal of
Vegetation Science 22, 1049–1061.
Ameztegui, A., Brotons, L., Coll, L., 2010. Land-use changes as major drivers of
mountain pine (Pinus uncinata Ram.) expansion in the Pyrenees. Global Ecology
and Biogeography 19, 632–641.
Barbeito, I., Dawes, M.A., Rixen, C., Senn, J., Bebi, P., 2012. Factors driving mortality
and growth at treeline: a 30-year experiment of 92 000 conifers. Ecology 93,
389–401.
Batllori, E., Camarero, J.J., Ninot, J.M., Gutiérrez, E., 2009. Seedling recruitment,
survival and facilitation in alpine Pinus uncinata tree line ecotones. Implications
and potential responses to climate warming. Global Ecology and Biogeography
18, 460–472.
Bertness, M.D., Callaway, R., 1994. Positive interactions in communities. Trends in
Ecology & Evolution 9, 191–193.
Bodin, J., Badeau, V., Bruno, E., Cluzeau, C., Moisselin, J.-M., Walther, G.-R., Dupouey,
J.-L., 2013. Shifts of forest species along an elevational gradient in Southeast
France: climate change or stand maturation? Journal of Vegetation Science 24,
269–283.
Brooker, R.W., Maestre, F.T., Callaway, R.M., Lortie, C.J., Cavieres, L.a., Kunstler, G.,
Liancourt, P., Tielbörger, K., Travis, J.M.J., Anthelme, F., Armas, C., Coll, L.,
Corcket, E., Delzon, S., Forey, E., Kikvidze, Z., Olofsson, J., Pugnaire, F., Quiroz,
C.L., Saccone, P., Schiffers, K., Seifan, M., Touzard, B., Michalet, R., 2008.
Facilitation in plant communities: the past, the present, and the future. Journal
of Ecology 96, 18–34.
Burnham, K.P., Anderson, D., 2002. Model Selection and Multi-Model Inference.
Springer.
Cailleret, M., Davi, H., 2010. Effects of climate on diameter growth of co-occurring
Fagus sylvatica and Abies alba along an altitudinal gradient. Trees 25, 265–276.
Callaway, R.M., Brooker, R.W., Choler, P., Kikvidze, Z., Lortie, C.J., Michalet, R., Paolini,
L., Pugnaire, F., Newingham, B., Aschehoug, E., Armas, C., Kikodze, D., Cook, B.J.,
2002. Positive interactions among alpine plants increase with stress. Nature
417, 844–848.
Candel-Pérez, D., Linares, J.C., Viñegla, B., Lucas-Borja, M.E., 2012. Assessing
climate–growth relationships under contrasting stands of co-occurring Iberian
pines along an altitudinal gradient. Forest Ecology and Management 274, 48–
57.
Castro, J., Zamora, R., Hódar, J.A., Gómez, J.M., 2002. Use of shrubs as nurse plants: a
new technique for reforestation in Mediterranean mountains. Restoration
Ecology 10, 297–305.
Castro, J., Zamora, R., Hódar, J.A., Gómez, J.M., Gómez-Aparicio, L., 2004. Benefits of
using shrubs as nurse plants for reforestation in Mediterranean mountains: a 4-
year study. Restoration Ecology 12, 352–358.
Cox, D.R., 1972. Regression models and life tables (with discussion). Royal Statistical
Society, Series B 34, 187–220.
Dullinger, S., Dirnböck, T., Grabherr, G., 2004. Modelling climate change-driven
treeline shifts: relative effects of temperature increase, dispersal and
invasibility. Journal of Ecology 92, 241–252.
Germino, M.J., Smith, W.K., Resor, A.C., 2002. Conifer seedling distribution and
survival in an alpine-treeline ecotone. Plant Ecology 162, 157–168.
Goffe, W.L., Ferrier, G.D., Rogers, J., 1994. Global optimization of statistical functions
with simulated annealing. Journal of Econometrics 60, 65–99.
Gómez-Aparicio, L., Zamora, R., Gómez, J.M., Hódar, J.A., Castro, J., Baraza, E., 2004.
Applying plant positive interactions to reforestation in Mediterranean
mountains: a meta-analysis of the use of shrubs as nurse plants. Ecological
Applications 14, 1128–1138.
Gómez-Aparicio, L., Pérez-Ramos, I.M., Mendoza, I., Matías, L., Quero, J.L., Castro, J.,
Zamora, R., Marañón, T., 2008a. Oak seedling survival and growth along
resource gradients in Mediterranean forests: implications for regeneration in
current and future environmental scenarios. Oikos 117, 1683–1699.
Table 3
Absolute and relative height growth at the end of the study period for seedlings of four species of trees planted at two experimental sites (montane and subalpine) in the Eastern
Pyrenees. Absolute growth is expressed as average annual growth, and relative growth is annual growth relative to the initial state.
Species Abs. height growth (mm y
1
) Rel. height growth
Montane belt Subalpine belt Montane belt Subalpine belt
Betula pendula 76.23 ± 4.44 a 69.73 ± 5.45 a 0.381 ±0.024 a 0.305 ±0.027 a
Pinus sylvestris 40.76 ± 1.91 b 34.98 ± 2.14 b 0.289 ±0.018 b 0.232 ±0.016 b
Pinus uncinata 30.48 ± 1.71 c 30.28 ± 2.04 b 0.283 ± 0.021 b⁄0.267 ± 0.021 b
Abies alba 16.48 ± 1.10 d 13.98 ± 1.18 c 0.207 ±0.016 c 0.160 ±0.016 c
Values are means ± SE. Values with the same letter indicate lack of significance between species. Values in bold indicate significant differences in relative growth between
sites for a given species (P< 0.05).
A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34 33
Gómez-Aparicio, L., Zamora, R., Castro, J., Hódar, J.A., 2008b. Facilitation of tree
saplings by nurse plants: microhabitat amelioration or protection against
herbivores ? Journal of Vegetation Science 19, 161–172.
Grabherr, G., Gottfried, M., Pauli, H., 1994. Climate effects on mountain plants.
Nature 369, 448.
Grace, J., Norton, D.A., 1990. Climate and growth of Pinus sylvestris at its upper
altitudinal limit in Scotland: evidence from tree growth-rings. Journal of
Ecology 78, 601–610.
Grace, J., Berninger, F., Nagy, L., 2002. Impacts of climate change on the tree line.
Annals of Botany, 537–544.
Guisan, A., Theurillat, J.P., Kienast, F., 1998. Predicting the potential distribution of
plant species in an alpine environment. Journal of Vegetation Science 9, 65–74.
Hoch, G., Körner, C., 2003. The carbon charging of pines at the climatic treeline: a
global comparison. Oecologia 135, 10–21.
Holtmeier, F.K., Broll, G., 2005. Sensitivity and response of northern hemisphere
altitudinal and polar treelines to environmental change at landscape and local
scales. Global Ecology and Biogeography 14, 395–410.
IPCC, 2007. Climate Change 2007: the physical science basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change. Cambridge University Press, UK and New York, NY,
USA.
Körner, C., Paulsen, J., 2004. A world-wide study of high altitude treeline
temperatures. Journal of Biogeography 31, 713–732.
Lenoir, J., Gégout, J.-C., Marquet, P.a., de Ruffray, Brisse, H.,, P., Brisse, H., 2008. A
significant upward shift in plant species optimum elevation during the 20th
century. Science 320, 1768–1771.
Lenoir, J., Gégout, J.-C., Pierrat, J.-C., Bontemps, J.-D., Dhôte, J.-F., 2009. Differences
between tree species seedling and adult altitudinal distribution in mountain
forests during the recent warm period (1986–2006). Ecography 32, 765–777.
Lenoir, J., Gégout, J.-C., Guisan, A., Vittoz, P., Wohlgemuth, T., Zimmermann, N.E.,
Dullinger, S., Pauli, H., Willner, W., Svenning, J.-C., 2010. Going against the flow:
potential mechanisms for unexpected downslope range shifts in a warming
climate. Ecography 2005, 295–303.
Linares, J.C., Camarero, J.J., 2012. Growth patterns and sensitivity to climate predict
silver fir decline in the Spanish Pyrenees. European Journal of Forest Research
131, 1001–1012.
Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J.,
Seidl, R., Delzon, S., Corona, P., Kolström, M., Lexer, M.J., Marchetti, M., 2010.
Climate change impacts, adaptive capacity, and vulnerability of European forest
ecosystems. Forest Ecology and Management 259, 698–709.
Macias, M., Andreu, L., Bosch, O., Camarero, J.J., Gutiérrez, E., 2006. Increasing aridity
is enhancing silver fir (Abies alba mill.) water stress in its south-western
distribution limit. Climatic Change 79, 289–313.
Maestre, F.T., Callaway, R.M., Valladares, F., Lortie, C.J., 2009. Refining the stress-
gradient hypothesis for competition and facilitation in plant communities.
Journal of Ecology 97, 199–205.
Martínez, I., González-Taboada, F., Wiegand, T., Camarero, J.J., Gutiérrez, E., 2012.
Dispersal limitation and spatial scale affect model based projections of Pinus
uncinata response to climate change in the Pyrenees. Global Change Biology 18,
1714–1724.
Messier, C., Puttonen, P., 1995. Spatial and temporal variation in the light
environment of developing Scots pine stands: the basis for a quick and
efficient method of characterizing light. Canadian Journal of Forest Research 25,
343–354.
Murphy, L., 2012. Likelihood: methods for maximum likelihood estimation. R
package version 1.5. Available online at <http://cran.r-project.org/web/
packages/likelihood/index.html>.
Niinemets, Ü., Valladares, F., 2006. Tolerance to shade, drought and waterlogging of
temperate northern hemisphere trees and shrubs. Ecological Monographs 76,
521–547.
Ninot, J.M., Carrillo, E., Font, X., Carreras, J., Ferre, A., Masalles, R.M., Soriano, I., Vigo,
J., 2007. Altitude zonation in the Pyrenees. A geobotanic interpretation.
Phytocoenologia 37, 371–398.
Pagès, J.-P., Pache, G., Joud, D., Magnan, N., Michalet, R., 2003. Direct and indirect
effects of shade on four forest tree seedlings in the French Alps. Ecology 84,
2741–2750.
Parent, S., Messier, C., 1996. A simple and efficient method to estimate microsite
light availability under a forest canopy. Canadian Journal of Forest Research 26,
151–154.
Peguero-Pina, J.J., Camarero, J.J., Abadia, A., Martin, E., Gonzalez-Cascon, R., Morales,
F., Gil-Pelegrín, E., 2007. Physiological performance of silver-fir (Abies alba
Mill.) populations under contrasting climates near the south-western
distribution limit of the species. Flora 202, 226–236.
Peñuelas, J., Boada, M., 2003. A global change-induced biome shift in the Montseny
mountains (NE Spain). Global Change Biology 9, 131–140.
Poyatos, R., Latron, J., Llorens, P., 2003. Land use and land cover change
after agricultural abandonment – the case of a Mediterranean Mountain
Area (Catalan Pre-Pyrenees). Mountain Research and Development 23,
362–368.
Prevosto, B., Balandier, P., 2007. Influence of nurse birch and Scots pine seedlings on
early aerial development of European beech seedlings in an open-field
plantation of Central France. Forestry 80, 253–264.
R Development Core Team, 2011. R: A language and environment for statistical
computing. In. R Foundation for Statistical Computing, Vienna, Austria.
Randin, C.F., Engler, R., Normand, S., Zappa, M., Zimmermann, N.E., Pearman, P.B.,
Vittoz, P., Thuiller, W., Guisan, A., 2009. Climate change and plant distribution:
local models predict high-elevation persistence. Global Change Biology 15,
1557–1569.
Resco de Dios, V., Fischer, C.R., Colinas, C., 2006. Climate Change Effects on
Mediterranean Forests and Preventive Measures. New Forests 33, 29–40.
Rolland, C., Michalet, R., Desplanque, C., Petetin, A., Aimé, S., 1999. Ecological
requirements of Abies alba in the French Alps derived from dendro-ecological
analysis. Journal of Vegetation Science, 297–306.
Ruiz de la Torre, J., 2006. Flora Mayor. Dirección General para la Biodiversidad.
Ministerio de Medio Ambiente, Madrid..
Saccone, P., Delzon, S., Pagès, J.-P., Brun, J.-J., Michalet, R., 2009. The role of biotic
interactions in altering tree seedling responses to an extreme climatic event.
Journal of Vegetation Science 20, 403–414.
Scherrer, D., Körner, C., 2011. Topographically controlled thermal-habitat
differentiation buffers alpine plant diversity against climate warming. Journal
of Biogeography 38, 406–416.
Schneider, S.H., 2004. Abrupt non-linear climate change, irreversibility and surprise.
Global Environmental Change 14, 245–258.
Therneau, T., Grambsch, P., 2000. Modeling Survival Data: Extending the Cox Model.
Springer, New York.
Therneau, T., 2011. A Package for Survival Analysis in S. R package version 2.37-2.
Available online at <http://CRAN.R-project.org/package=survival>..
Therneau, T., 2012. coxme: Mixed Effects Cox Models. R package version 2.2-3.
Available online at <http://cran.r-project.org/web/packages/coxme/
index.html>.
Thomas, C.D., Cameron, A., Green, R.E., Bakkenes, M., Beaumont, L.J., Collingham,
Y.C., Erasmus, B.F.N., de Siqueira, M.F., Grainger, A., Hannah, L., Hughes, L.,
Huntley, B., van Jaarsveld, A.S., Midgley, G.F., Miles, L., Ortega-Huerta, M.A.,
Townsend Peterson, A., Phillips, O.L., Williams, S.E., 2004. Extinction risk from
climate change. Nature 427, 145–148.
Thornthwaite, C.W., Mather, J.R., Carter, D.B., 1957. Instructions and tables for
computing potential evapotranspiration and the water balance. Publications in
Climatology 10, 183–311.
Thuiller, W., Lavorel, S., Araujo, M.B., Sykes, M.T., Prentice, I.C., 2005. Climate change
threats to plant diversity in Europe. PNAS 102, 8245–8250.
Toromani, E., Sanxhaku, M., Pasho, E., 2011. Growth responses to climate and
drought in silver fir (Abies alba) along an altitudinal gradient in southern
Kosovo. Canadian Journal of Forest Research 41, 1795–1807.
Vetaas, O.R., 2002. Realized and potential climate niches: a comparison of four
Rhododendron tree species. Journal of Biogeography 29, 545–554.
Walther, G.-R., Post, E., Convey, P., Menzel, A., Parmesank, C., Beebee, T.J.C., Frome
ntin, J.-M., Hoegh-Guldberg, O., Bairlein, F., 2002. Ecological responses to recent
climate change. Nature 416, 389–395.
Willis, K.J., Bhagwat, S.A., 2009. Biodiversity and climate change. Science 326, 806–
807.
34 A. Ameztegui, L. Coll / Forest Ecology and Management 303 (2013) 25–34
