Environmental filtering drives community specific leaf area in Spanish forests and predicts relevant changes under future climatic conditions

Forest Ecology and Management 405 (2017) 1–8

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Environmental filtering drives community specific leaf area in Spanish forests and predicts relevant changes under future climatic conditions

MARK

⁎

José M. Costa-Sauraa, , Antonio Trabuccoa,b, Donatella Spanoa,b, Simone Mereua,b a b

Department of Sciences for Nature and Environmental Resources, University of Sassari, Sassari 07100, Italy Euro-Mediterranean Center on Climate Changes, IAFES Division, Sassari 07100, Italy

A R T I C L E I N F O

A B S T R A C T

Keywords: Community weighted mean Functional diversity Functional traits Forest Inventory Climate change Mediterranean

How functional traits at community level relate with environmental conditions is of great relevance to assess potential effects of climate change on ecosystem functioning. Species’ specific leaf area (SLA) is well recognised to be closely correlated with species drought resistance and with other forest functions such as productivity. Here, we used tree species abundance data from 44 501 forest plots from the Third Spanish National Forest Inventory and species SLA values from literature to assess how community weighted mean SLA (CWMSLA) and SLA diversity within communities (FDisSLA) of Spanish forests correlate with aridity. Later, using 19 climate change projections and following an approach that limits the values of CWMSLA along an aridity gradient, we assessed the potential climatic effects on CWMSLA for 2050 under the representative concentration pathways (RCP) 4.5 and 8.5. Results showed that CWMSLA and FDisSLA decreased significantly with aridity (deviance explained was 22 and 9%, respectively) suggesting an effect of climatic filtering at community level constraining the diversity of co-occurring strategies at harsher conditions. Up to 25% of plots were predicted to suffer changes in CWMSLA with these impacts being more common and of a greater magnitude in communities characterised by a high CWMSLA and located at humid and mid-altitude zones. Instead, communities already striving in arid areas appeared to be more resilient. The study proves useful for orienting forest management practices in current permanent forest stands based on trait ecology (e.g. promoting communities species composition with specific trait values), to increase their mitigation potential and adaptive capacity to current and future changing climate conditions.

1. Introduction Global change poses a threat to forest ecosystems and for the services they provide such as carbon sequestration, water regulation or nutrient cycling (Bartczak et al., 2014; Lindner et al., 2010; Schröter et al., 2005). Indeed, recent climate change is already influencing wood production and carbon storage of forests across biomes. Although this influence might be positive in several colder biomes that are experiencing warmer and wetter condition (e.g. increasing productivity, Briceño-Elizondo et al., 2006; Eggers et al., 2008), the effects of climate change are clearly negative in the Mediterranean region (e.g. decreasing tree growth and carbon stocks, Vayreda et al., 2012; RuizBenito et al., 2014a). In fact, among different regions, the Mediterranean basin is considered exceptionally vulnerable to climatic change (Lindner et al., 2010; Schröter et al., 2005) because of its location within the transitional zone between temperate and arid climates (i.e. between central Europe and north Africa) where a reduction in precipitation and a strong increase in temperatures is expected (Giorgi,

⁎

2006). Since ecosystem services of Mediterranean are in great extent determined by the functional attributes of communities (see RuizBenito et al., 2014b) how they may respond to climate change is extremely relevant to determine optimal management policies to increase resilience and optimize functionalities (Lavorel, 2013). A trait-based approach (i.e. based on plant morphological, physiological, or phenological attributes, Violle et al., 2007) is a powerful approach to assess climate change impacts, since traits represent both species adaptation to the environment and their effects on ecosystems (Lavorel et al., 2007). Both functional diversity (FD) and Community Weighted Mean (CWM) of functional traits have been shown to be relevant for ecosystem functioning at the community level (Jucker et al., 2014; Mokany et al., 2008; Paquette and Messier, 2011). CWM is based on the mass ratio hypothesis (Grime, 1998), which states that dominant species can determine, to a great extent, some ecosystem functions such a productivity via their dominant attributes (Lavorel et al., 2011; Roscher et al., 2012). In this context, dominant species are those that account for the highest proportion of biomass in the community

Corresponding author. E-mail addresses: [email protected] (J.M. Costa-Saura), [email protected] (A. Trabucco), [email protected] (D. Spano), [email protected] (S. Mereu).

http://dx.doi.org/10.1016/j.foreco.2017.09.023 Received 23 May 2017; Received in revised form 5 September 2017; Accepted 10 September 2017 Available online 28 September 2017 0378-1127/ © 2017 Elsevier B.V. All rights reserved.

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al.

ecosystem functioning. For instance, DGVM have been used to model and predict net primary productivity under future scenarios (Prentice et al., 2007) and SDM to project future distributions of functional groups exploring potential effects on ecosystem properties such as soil biogeochemistry or fire flammability (Thuiller et al., 2006). Despite their usefulness, one of the limitations of these approaches is that they rely on the functional simplifications of communities (Scheiter et al., 2013), wherein Yang et al. (2015) observed that there might be greater variations within PFT (in terms of functional responses and effects) than among them. To overcome these methodological limits, research is now pointing to model community traits values as a whole and later project these under future conditions, in order to assess more accurately how the impacts of climate change on vegetation translate into changes in ecosystem functioning. For instance, Frenette-Dussault et al. (2013), based on the Community Assembly by Traits-Selection approach (CATS, Shipley, 2009), predicted shifts from ruderal to stress-tolerant subshrub communities on Morocco steppes, which might reduce the pastoral value of vegetation. However, CATS approach needs to assume that species pool from the studied area will not change in the future (i.e. no migration of new species from other areas). This assumption is an important drawback since migration of new species and consequently changes in species pools are common output from most studies on the effects of climate change in species distribution (Bakkenes et al., 2002; Garzón et al., 2008; Thuiller et al., 2006). Here, we propose an approach that does not need to articulate potential changes on community mean traits considering relative changes in species abundances, species trait plasticity, adaptation, or new species arrivals (note, that all mechanism may lead to changes on CWM; Bussotti et al., 2014). Instead, based on the environmental filtering theory (Keddy, 1992), our study stresses the hypothesis that dominant trait values will vary with climatic conditions leading to changes in ecosystem functioning. Indeed, similar approaches have already credited potential impacts on ecosystem functioning as a consequence of climatic filtering effects on community traits over non woody alpine plant systems (Dubuis et al., 2013), Tasmanian forest communities (Mokany et al., 2015) and in Swedish wetlands (Moor et al., 2015). Thus, benefiting from the opportunity of a consolidated and accessible forest inventory data (Tomppo et al., 2010), we assessed climate dependency of CWMSLA for more than 44 000 permanent plots distributed across the whole forested territory of Spain. We used SLA as the main trait expression of environmental filtering, since it is wellrecognised that SLA reflects species aridity tolerance (Micco and Aronne, 2012; Niinemets, 2001) and particularly of Mediterranean tree species (Costa-Saura et al., 2016). Before assessing potential climate change effects on communities, we tested the following hypotheses over which our approach is based: 1) CWMSLA would decrease with aridity in the Mediterranean since low SLA values are required for drought adaptation, and 2) FDSLA and the range of CWMSLA would also decrease with aridity since increasing drought severity will constrain the number of suitable strategies. The following step was to assess the potential changes in CMWSLA on Spanish forest communities under multiple climate change scenarios in terms of vulnerability, magnitude of impact and likelihood of change. We hypothesized that communities from humid zones characterised by high CMWSLA will be impacted more frequently and will undergo a greater magnitude of change.

whereas transients species are those instable in time and space and account for less biomass (Grime, 1998). Thus, dominant traits combined in the form of CWM have been used to test the effect of different species compositions on ecosystem functions (Debouk et al., 2015; Lavorel et al., 2011). For instance, Garnier et al. (2004) studying Mediterranean successional series following vineyard abandonment showed how CWM traits (i.e. community specific leaf area and leaf nitrogen content) were positively correlated with functions such as net primary productivity and litter decomposition rate. Accordingly, they suggested these traits as functional markers for scaling up from species organs to ecosystems functions. FD (i.e. trait diversity within communities) has also been found to positively influence ecosystem functioning through mechanisms such as complementary and facilitation resource use (Hooper et al., 2005). For instance, it was observed that FD might improve functions such as productivity and stability across different forest systems (Jucker et al., 2014; Paquette and Messier, 2011; Ruiz-Benito et al., 2014b). Community mean traits are also seen as a trait expression of the environmental filtering effect on community composition (i.e. environmental selection at community level; Shipley et al., 2006). This means that environmental factors act as a filter by removing those species lacking a specific combination of traits (Keddy, 1992), and thus species of a given community at a given environmental condition tend to converge on specific trait values, but note that a certain degree of trait divergence is also expected for species coexistence because of niche differentiation (Chesson, 2000; Maire et al., 2012). For instance, Ackerly et al. (2002) found that in Mediterranean climates, solar radiation promoted communities mostly composed by species with low leaf area and low specific leaf area (SLA). They argued that species lacking this trait combination might have difficulties to perform properly under arid conditions. Indeed, low SLA values are commonly related with more tightly packed cells with thicker cell walls and few air spaces which might enhance leaf wilting resistance and reduce water losses (Poorter et al., 2009; Micco and Aronne, 2012; De la Riva et al., 2016a). Moreover, Pakeman et al. (2009) using data from different vegetation types across locations in Europe observed that variations in community traits values were mainly explained by differences in climatic and soil conditions. However, sometimes CWM might not entirely reflect species optimal strategies because the environment might more strongly constrain multivariate than univariate traits (Muscarella and Uriarte, 2016). In addition, environmental filtering is also expected to influence the variation of suitable functional responses along climatic gradients (i.e. a lower FD but also a narrower range of CWM values). It has been suggested that strong filtering operates at less favourable locations promoting greater functional similarity (i.e. constraining the trait values required for existence in harsh conditions, Swenson et al., 2012; Weiher and Keddy, 1995). Indeed Swenson and Enquist (2007), analysing data from North, Central and South America, showed that wood density varies less in forest communities from temperate and high elevation zones than those from tropical and low lands with more favourable abiotic conditions. Studies on climate change traditionally have focused on its impact on species distributions and the risk of specific species being lost including their functions and services they provide (e.g. Hanewinkel et al., 2013). For instance, Garzón et al. (2008) reported potential losses of genetic diversity of Iberian tree by some species reducing their range as a consequence of climate change. However, in general it is difficult to assess climate change impact on ecosystem functioning based on a single species analysis. An alternative approach used to overcome this limitation is to group species in plant functional types (PFT), assuming that a set of species share a similar eco-physiology and thus affect ecosystems in a similar way (Díaz and Cabido, 1997). Dynamic Global Vegetation Models (DGVM, Prentice et al., 2007) and Species Distribution Modelling (SDM, Thuiller et al., 2006) approaches have been used to translate climate change impacts on PFT into consequences for

2. Materials and methods 2.1. Study area, community and climatic data The study area covered the Spanish continental territory, located between 36° N and 44° N of latitude, and between 10° W and 3° E longitude. The main Köppen climatic domains are arid and temperate, with annual mean temperatures (AMT) ranging from ∼3 to ∼17 °C and annual mean precipitations (AMP) from ∼300 to ∼2200 mm/year (Chazarra, 2011). 2

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al.

variable and full model deviance explained). GAM were used because they do not assume an a priori curve type for the relationship between predictors and the response variable (Yee et al., 2007).

Species community composition and cover was retrieved from the Third Spanish National Forest Inventory (IFN3) (Direccion General de Conservacion de la Naturaleza, 2006). IFN includes periodical surveys, approximately every 10 years, of the entire Spanish forested area reporting information about stand composition and data on diameter at breast height and height for tree species. IFN3 accounts for more than 90,000 plots regularly distributed over a ∼1 km2 grid. We subsampled IFN3 for plots composed (within 25 m radius) only by tree species that usually form natural forests in Spain and that were not classified as plantations (i.e. we excluded plots composed by species such as Eucalyptus sp., Pinus radiata, Acacia sp. or Castanea sativa). Furthermore, we selected only those plots in which the stand canopy cover of the selected species was above 90%. Note that Pakeman and Quested (2007) showed that for quantitative traits at least 80% tree cover should be accounted to avoid accuracy lost in community trait metrics. At the end of this selection process, 44,501 plots remained and were used for further analysis. Species SLA values (in m2 kg−1) were retrieved from the literature (see Appendix 1 for data values and sources used). We selected those values that were measured on sun exposed leafs and on adult plants. When more than one value was found for a species, the average of the value was used. Community weighted mean SLA (CWMSLA) was calculated as the average SLA value of species inside the community weighted by their relative abundances (Garnier et al., 2004), where abundance was defined as the percentage cover of the species. SLA variability within communities (i.e. SLA functional diversity) was calculated using the Functional Dispersion index (FDis, Laliberté and Legendre, 2010). FDis represents, in a trait space, the mean distance of individual species to the centroid of all species accounting for species abundances. Thus, both centroid position and individual distances are weighted by species relative cover. Preliminary analysis showed that FDis was highly correlated (∼0.9) with functional metrics like Rao’s quadratic entropy and functional richness (see Laliberté and Legendre, 2010 for further details on these metrics). Environmental data for current conditions (average over 1960–2000) were downloaded at a resolution of 30 arc seconds (∼600 m for Spain) from the Consortium for Spatial Information of the Consultative Group for International Agricultural Research (CSICGIAR) and WorldClim websites (Hijmans et al., 2005, http://www.csi. cgiar.org and http://www.worldclim.org, respectively). Because of collinearity, environmental predictors selection was based on a priori ecological knowledge and on pairwise correlations (Dormann et al., 2013). The aridity index, which is the ratio of annual precipitation over potential evapotranspiration (PET), was preferred over other precipitation variables (e.g. AMP or precipitations of the driest quarter) because it integrates the evaporative effect of temperature and radiation and it better represents water availability over vegetation water demand (see Zomer et al., 2008 for further details). Mean temperature of the coldest quarter (MTCQ) was also included as a predictor, since low temperatures might also influence species SLA. For instance, Poorter et al. (2009) reported in a meta-analysis that species with low SLA, coupled with more cell layers, can slow down leaf freezing. MTCQ was preferred over AMT because it was less correlated with aridity (note that AMT is embedded on PET and thus on aridity index). Soil nutrient content was not considered in the analysis because of data unavailability, however, previous studies at landscape scale on Spain did not find a significant effect of soil nutrient content on CWMSLA (De la Riva et al., 2016b).

2.3. Climate change models A combination of 19 downscaled Earth System Models (ESM, see Appendix 2) projections of future climate developed in the Phase 5 of the Coupled Model Intercomparison Project (CIMP5; Meehl and Bony, 2011) for different representative concentration pathways (RCP’s; van Vuuren et al., 2011) were downloaded from the WorldClim website. The dataset provides projections for 2050 (monthly averages of precipitations and temperature over 2040–2060) downscaled at a resolution of 30 arc seconds using the Delta method (Ramirez-Villegas and Jarvis, 2010) which, using the present climatic data defined by the WorldClim, spatially resolves monthly ESM climate anomalies assuming that the change in climate is relatively stable over space within each ESM pixel under climate change. From the data retrieved from WorldClim for the year 2050 and for each plot we extracted the MTCQ, and both the projected annual precipitation and potential evapotranspiration to calculate the aridity index projected by each of the 19 ESM for the RCP 4.5 and 8.5 scenarios. Implementing such ensemble allowed us to account for diverging climate projections, and thus assessing the uncertainty inherent in different ESM simulations.

2.4. Predicting climate change impacts on CWMSLA Our study rely on setting CWMSLA upper limits for different aridity conditions to later asses if current CWMSLA values would fall above the upper limit under projected aridity conditions. First, the studied area was subdivided into 13 aridity classes, with a predominant aridity interval of 0.1 and at least including 1% of the data. Thus, aridity classes with few plots located at the extremes were not included in the analysis. We preferred an approach based on classes to a model fitted over the mean (as in Dubuis et al., 2013; Mokany et al., 2015) because it directly determines the upper and lower range values of each class. Afterwards, the CWMSLA of plots falling under the same aridity class were aggregated to determine the CWMSLA distribution within each class (Fig. 1). Then, we set two distinct CWMSLA upper limits for each class as the 75th and 90th distribution percentiles within each aridity class. Two distinct percentiles were used to assess the sensitivity of the approach for different percentiles in further analyses. Next, to determine if a plot was susceptible to a change in CWMSLA, we defined two conditions that had to be met simultaneously: 1) ESM projections have to predict a change in aridity class (i.e. hazard assessment, sensu IPCC, 2014; note that we did not consider plots shifting to more humid climates because they only accounted for ∼1%), and 2) current CWMSLA values have to be above the upper CWMSLA limit of the new aridity class (i.e. vulnerability assessment, sensu IPCC, 2014). In addition, we computed the difference between current CWMSLA and the upper CWMSLA limit of the new aridity class to assess the magnitude of the change (i.e. impact assessment, sensu IPCC, 2014). Since hazard, vulnerability and impact were calculated each time across all ESM projections, it allowed estimating the uncertainty inherent in 19 different ESM. Hazard, vulnerability and impact assessment were arranged over different land typologies in order to identify environmental conditions and geographical areas more at risk. In particular, relative frequency of plots by predicted change in aridity and CWMSLA, and the magnitude of change, were grouped in classes of elevation, aridity, and CWMSLA. We also estimated and mapped for each plot the likelihood of change (sensu IPCC, 2014) as the ratio between the number of models predicting change and the total number of used ESM models.

2.2. Climatic factors driving community SLA Generalized Additive Models (GAM), implemented in the mgcv package (Wood, 2004) in R statistical software 3.2.0 (R Core Team 2015, http://www.R-project.org/), were used to disentangle the relative importance of environmental variables (i.e. aridity index and MTCQ together and separately) on CWMSLA and FDisSLA (via single 3

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al.

Fig. 1. Community weighted mean SLA (CWM SLA, in m2 kg−1) and Functional Dispersion SLA (FDisSLA) across aridity classes (higher values represent greater water availability). Hinges represent the first and third quartiles and whiskers extend up to ± 1.5 the interquartile range. Upper panel shows the number of plots within each aridity class.

3. Results

the upper hinge and whisker of Fig. 1) reaches a maximum asymptote staying fairly constant at more humid classes. In addition, mean and range of FDisSLA increase with humidity, reaching a peak which decreases for aridity classes greater than 1.3. A positive effect of MTCQ on CWMSLA was only observed along the coldest part of the temperature gradient, whereas its effect on FDisSLA was also positive on the coldest part of the gradient, but negative on the warmest (Appendix 3).

3.1. Community SLA along environmental gradients CWMSLA and FDisSLA modelled relations to climate (i.e. including together aridity and MTCQ) explained 24.2% and 10.6% of the observed deviance, respectively. The most important variable explaining the variability of CWMSLA and FDisSLA was aridity (Table 1). Results showed that, with increasing humidity, communities tend to have a higher CWMSLA with wider distributions (i.e. grater range of CWMSLA values, Fig. 1). Notably, while the mean CWMSLA of each aridity class continuously increases with humidity, the upper limit of the class (see

3.2. Climate change impacts The hazard assessment revealed potential changes in aridity classes for 80 and 85% (N = 44 501) of plots under scenarios RCP 4.5 and RCP 8.5, respectively. Under both scenarios, changes in aridity were predicted to be more common in humid sites, at higher altitudes and in plots characterized by high CWMSLA values (Appendix 4). However, the vulnerability assessment showed that, when using the 75th percentile as CWMSLA upper limits, changes in CWMSLA might occur over 23 and 25% of plots under scenarios RCP 4.5 and RCP 8.5, respectively. Instead, when using the 90th percentile, changes in CWMSLA would be less common, namely 13 and 15% under scenarios RCP 4.5 and RCP 8.5, respectively. Overall, changes in CWMSLA were predicted to decrease with aridity, to increase with altitude, and to be more common in communities characterized by a higher CWMSLA. However, changes in CWMSLA were projected to be slightly less frequent

Table 1 Deviance explained by each single predictor on the generalized additive models. Response variable and predictor variables CWMSLA Aridity Mean temperature coldest quarter FDisSLA Aridity Mean temperature coldest quarter

Single-variable model deviance explained (%)

22.1 3.79 9.23 3.18

Note: all smoothing terms were significant (p-value < 0.001).

4

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al.

Fig. 2. Fraction of plots with predicted change in CWMSLA (i.e. the relative frequency within each class) across different community characteristics when using the 75th percentile for setting CWMSLA limits. Lines represent the mean value while vertical bar lines represent standard deviation. Numbers in the lower part of each single figure represent the number of plots within each thematic class (for aridity classes see Fig. 1) Units: aridity (dimensionless), CWMSLA (m2 kg−1), altitude (in meters).

filtering imposed by water availability over this trait at community level suggesting that this pattern, previously observed at the species level (Costa-Saura et al., 2016; Niinemets, 2001), also occurs at the community level. Such result was expected since low SLA values are usually associated with species capacity to resist drought with tightly package cells with thicker walls and low air spaces that may enhance leaf wilting resistance and reduce water losses (De la Riva et al., 2016a; Micco and Aronne, 2012; Poorter et al., 2009). Interestingly, the upper CWMSLA range tends to level off for the most humid classes, suggesting that SLA declines in relevance, and might vary because of other factors (Poorter et al., 2009), when water is not a limiting factor. A temperature effect on CWMSLA was observed for the coldest locations, but it did not exert by far the same ubiquitous importance as aridity. Indeed, in a previous study (De la Riva et al., 2016b) it was found that water availability was the main factor driving CWMSLA in forests from southern Spain. The results also show that in humid locations a wider spectrum of functional strategies may prove viable both within the same aridity class (i.e. a greater variability of CMWSLA values) and within the same community (i.e. a higher FDisSLA), whereas in arid locations the range

for the most humid and elevated plots (Fig. 2). Impact assessment showed that the magnitude of CWMSLA changes increases with humidity up to aridity classes between 0.9 and 1.1, and decreases for the most humid classes (Fig. 3). Instead, the magnitude of change gradually increases with altitude and for communities with high CWMSLA. Overall, changes in CWMSLA were predicted to be greater when using the 75th instead of the 90th percentile (see Appendix 4 for further details). Geographical distribution of results showed that regardless of the RCP scenario and percentile used to determine CWMSLA limits, predicted changes in CWMSLA were evenly distributed across the whole Spanish territory with high agreement across ESM projections (i.e. the likelihood of change, Fig. 4). However, a greater magnitude of change was predicted in the middle and-northern part of the country (Fig. 4). 4. Discussion 4.1. Environmental filtering on community SLA The reduction of CWMSLA at increasing aridity reflects the current

Fig. 3. Magnitude of change (i.e. the difference among current CWMSLA and the CWMSLA limit of the projected aridity class in 2050; in m2 kg−1) across different community characteristics when using the 75th percentile for setting CWMSLA limits. Bars represent the mean value while error bars are standard deviation (thin vertical lines).

5

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al.

Fig. 4. Maps of forest plots with the likelihood of change in CWMSLA (i.e. predictions agreement when using the 19 different ESM projections) and the magnitude of that change (i.e. the difference among current CWMSLA and the upper CWMSLA limit of the new arid class; in m2 kg−1) for scenario RCP 8.5 when using the 75th percentile for setting CWMSLA limits.

predicted increments in CWMHeight and CWMSLA of non woody species due to increasing temperatures in the Swiss Alps. However, the expected changes in the Mediterranean region are different from those projected for temperate or boreal zones, since warming will be associated to a reduction of the already limited precipitations (Giorgi and Lionello, 2008). Thus, in water limited environments, greater aridity will promote species with leaves adapted to resist drought (Micco and Aronne, 2012, Costa-Saura et al., 2016). Consistently, our results projected a significant decrease in CWMSLA: Communities from humid and mid elevation locations, and with high SLA (those mainly composed by deciduous species from the family Fagaceae, see Appendix 5 for main forests types associated with each CWMSLA class) were predicted to be the most vulnerable (i.e. the greatest percentage of plots with change in CWMSLA). Furthermore, our results were in agreement with those found by other authors working at species level (Garzón et al., 2008; RuizLabourdette et al., 2012). For instance, studies that analysed climate change impacts on species distribution in Spain observed that species like Fagus sylvatica, Quercus petraea or Quercus pyrenaica, characterised by high SLA and mainly located in mountainous regions, might suffer the greatest impacts whereas species better adapted to drought, i.e. Pinus halepensis or Quercus ilex, would be expected to increase their geographical ranges (Garzón et al., 2008; Ruiz-Labourdette et al., 2012). Our results also projected that the magnitude of the change will be greater in communities from humid and mid elevation locations, and currently characterized by high CWMSLA. Indeed, studies conducted at species level have observed that Quercus ilex (with SLA values between 4 and 7 m2 kg−1) is replacing Fagus sylvatica (SLA values between 12 and 14 m2 kg−1) in Catalonia mountains (north eastern Spain) as a consequence of global change (Peñuelas et al., 2007). Importantly, our results also showed that at the most humid locations changes in CMWSLA might be lower, most likely because these zones also host communities with low CWMSLA probably promoted by low temperatures (Poorter et al., 2009). Thus, despite communities with low CWMSLA that currently occur in these cold and humid zones, climate change may promote communities with higher SLA (see Thuiller et al., 2006 for results based on modelling forest types). Indeed, in the Spanish Pyrenees it was observed abundant recruitment of Quercus sp on declining stands of Pinus sylvestris as consequence of changes in temperature and precipitations (Galiano et al., 2010). Thus, our results together with those at species levels (despite the latter not being directly linked to functions) suggest significant alterations to those ecosystem process related to SLA. SLA together with climate, soil resources and other closely related traits (e.g. N content and photosynthetic rate, Wright et al., 2004) determine functions as net

of strategies is considerably narrower. This is consistent with the theory that predicts environmental filtering strongly operating in low resource environments limiting the range of suitable strategies (Weiher and Keddy, 1995). For example, Swenson et al. (2012) identified that for SLA and wood density the range of values that can be observed in America are smaller at low mean temperatures and high latitudes suggesting the same trends at community level (i.e. low variability on CWMSLA). Thus, by limiting diversity on SLA values, harsher environments constrain coping strategies (e.g. the slow-fast resource use spectrum, Díaz et al., 2004; Wright et al., 2004; Reich, 2014), potentially limiting the quantity of services that can be supplied (e.g. amount and quality of wood and fodder provisions, Bello et al., 2010). In addition, previous studies found that diversity might entail drought stress alleviation because of complementarity on resource use and/or facilitation process (Grossiord et al., 2014; Lebourgeois et al., 2013; Pretzsch et al., 2013). Thus, the low FDisSLA observed in dryer sites might limit the positive effects of diversity on community resistance to biotic and abiotic disturbances (Jactel and Brockerhoff, 2007). In addition, higher FD can entail a larger diversity of responses to environmental changes, conferring a higher resilience to climate change (Mori et al., 2013). Note that the range of possible CWMSLA found for a given level of aridity not only reflects alternative community structures but also embeds different successional stages, management practices as well as local climatic conditions that are not captured by the climate data used. 4.2. Climate change impacts on CWMSLA Results showed that a high percentage of forest plots are expected to experience changes in aridity. Interestingly, impacts under scenario RCP 8.5 were not much higher than under RCP 4.5, mainly due to the fact that the mean difference in aridity index among these two scenarios for the studied plots was ∼0.02. However, it is important to note that our predictions were made for 2050, where climatic differences among scenarios are rather small compared to those projected for year 2100 (IPCC, 2014). Overall, changes in aridity were predicted to be more common for plots currently characterized by a higher humidity and higher altitude, which are sites characterized by species with high SLA and thus less adapted to drought events. Indeed, our results predicted that climate change will mostly impact these forest communities due to climate filtering. Studies conducted in other climatic zones and vegetation types also predict changes on community traits values due to the impositions created by future climatic conditions. For instance Dubuis et al. (2013) 6

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al.

References

primary productivity (Stuart Chapin et al., 2009). Indeed, previous studies also using the Spanish Forest Inventory dataset showed that CWMSLA influences stand’s carbon storage and productivity (RuizBenito et al., 2014b). Therefore, strong changes on communities with high SLA imply a reduction in productivity and might also imply changes in other functions correlated with SLA as litter decomposition (Garnier et al., 2004; Cornwell et al., 2008). Importantly, changes in ecosystems functioning are likely to determine changes in management practices such as rotation lengths or thinning prescriptions in order to adapt to new productivity regimens (Millar et al., 2007). In addition, the role of management can be extremely crucial in order to reduce climate change impacts (Keenan, 2015; Resco de Dios et al., 2006; Spittlehouse and Stewart, 2003). For instance, previous studies have shown that thinning treatments might alleviate stress under climate changing conditions, because of lowering inter and intra-specific competition for water, and thus improving growth and potentially reducing mortality (Briceño-Elizondo et al., 2006; Kerhoulas et al., 2013; RuizBenito et al., 2013). Thus, these adaptation practices may be effective to maintain viability and/or resilience on communities with small projected changes in CWMSLA. Instead, in locations in which greater changes on CWMSLA are projected, more aggressive treatments may be required. Combination of management practices as selective thinning, species restoration and assisted migration should aim at reaching the target CWMSLA (Bussotti et al., 2014) and will depend on the FDisSLA of the stand. In case of monocultures, planting and restoring with species with trait values compatible with future climatic conditions, even assisted migration, could be an adequate option. Instead, in communities with a high FDisSLA the projected CWMSLA may fall within the SLA range of species composing the community or below the value of the species with the lowest SLA. In the first case, species abundances could be managed in order to reach the target CWMSLA (e.g. by selective thinning), whereas in the second case the introduction of new species, or within species adapted genotypes (see examples in Bussotti et al., 2014), will be required to reach the target CWMSLA and thus adapted to future climatic conditions.

Ackerly, D., Knight, C., Weiss, S., Barton, K., Starmer, K., 2002. Leaf size, specific leaf area and microhabitat distribution of chaparral woody plants: contrasting patterns in species level and community level analyses. Oecologia 130, 449–457. Bakkenes, M., Alkemade, J.R.M., Ihle, F., Leemans, R., Latour, J.B., 2002. Assessing effects of forecasted climate change on the diversity and distribution of European higher plants for 2050. Glob. Chang. Biol. 8, 390–407. Bartczak, A., Giergiczny, M., Jacobsen, Jette Bredahl, Mavsar, R., Mäntymaa, E., Ovaskainen, V., Varela, E., Pettenella, D., Thorsen, B.J., Tyrväinen, L., Vedel, S.E., Wunder, S., 2014. The provision of forest ecosystem services: quantifying and valuing non-marketed ecosystem services. European Forest Institue, Joensuu, FI. Bello, F., Lavorel, S., Díaz, S., Harrington, R., Cornelissen, J.H.C., Bardgett, R.D., Berg, M.P., Cipriotti, P., Feld, C.K., Hering, D., Martins da Silva, P., Potts, S.G., Sandin, L., Sousa, J.P., Storkey, J., Wardle, D.a., Harrison, P.a., 2010. Towards an assessment of multiple ecosystem processes and services via functional traits. Biodivers. Conserv. 19, 2873–2893. Briceño-Elizondo, E., Garcia-Gonzalo, J., Peltola, H., Matala, J., Kellomäki, S., 2006. Sensitivity of growth of Scots pine, Norway spruce and silver birch to climate change and forest management in boreal conditions. For. Ecol. Manage. 232, 152–167. Bussotti, F., Pollastrini, M., Holland, V., Bruggemann, W., 2014. Functional traits and adaptive capacity of European forests to climate change. Environ. Exp. Bot. 111, 91–113. Chazarra, A., 2011. Atlas Climático Ibérico. Ministerio de Medio Ambiente y Medio Rural y Marino, Madrid, SP. Chesson, P., 2000. Mechanism of maintenance of species diversity. Annu. Rev. Ecol. Syst. 31, 343–358. Cornwell, W.K., Cornelissen, J.H.C., Amatangelo, K., Dorrepaal, E., Eviner, V.T., Godoy, O., Hobbie, S.E., Hoorens, B., Kurokawa, H., Pérez-Harguindeguy, N., Quested, H.M., Santiago, L.S., Wardle, D.A., Wright, I.J., Aerts, R., Allison, S.D., Van Bodegom, P., Brovkin, V., Chatain, A., Callaghan, T.V., Díaz, S., Garnier, E., Gurvich, D.E., Kazakou, E., Klein, J.A., Read, J., Reich, P.B., Soudzilovskaia, N.A., Vaieretti, M.V., Westoby, M., 2008. Plant species traits are the predominant control on litter decomposition rates within biomes worldwide. Ecol. Lett. 11, 1065–1071. Costa-Saura, J.M., Martínez-Vilalta, J., Trabucco, A., Spano, D., Mereu, S., 2016. Specific leaf area and hydraulic traits explain niche segregation along an aridity gradient in Mediterranean woody species. Perspect. Plant Ecol. Evol. Syst. 21, 23–30. De la Riva, E.G., Olmo, M., Poorter, H., Ubera, J.L., Villar, R., 2016a. Leaf mass per area (LMA) and its relationship with leaf structure and anatomy in 34 mediterranean woody species along a water availability gradient. PLoS One 11, 1–18. De la Riva, E.G., Pérez-Ramos, I.M., Tosto, A., Navarro-Fernández, C.M., Olmo, M., Marañón, T., Villar, R., 2016b. Disentangling the relative importance of species occurrence, abundance and intraspecific variability in community assembly: A traitbased approach at the whole-plant level in Mediterranean forests. Oikos 125, 354–363. Debouk, H., De Bello, F., Sebastia, M.T., 2015. Functional trait changes, productivity shifts and vegetation stability in mountain grasslands during a short-term warming. PLoS One 10, 1–17. Díaz, S., Cabido, M., 1997. Plant functional types and ecosystem function in relation to global change. J. Veg. Sci. 8, 463–474. Díaz, S., Hodgson, J.G., Thompson, K., Cabido, M., Cornelissen, J.H.C., Jalili, A., Montserrat-Martí, G., Grime, J.P., Zarrinkamar, F., Asri, Y., Band, S.R., Basconcelo, S., Castro-Díez, P., Funes, G., Hamzehee, B., Khoshnevi, M., Pérez-Harguindeguy, N., Pérez-Rontomé, M.C., Shirvany, A., Vendramini, F., Yazdani, S., Abbas-Azimi, R., Bogaard, A., Boustani, S., Charles, M., Dehghan, M., de Torres-Espuny, L., Falczuk, V., Guerrero-Campo, J., Hynd, A., Jones, G., Kowsary, E., Kazemi-Saeed, F., MaestroMartínez, M., Romo-Díez, A., Shaw, S., Siavash, B., Villar-Salvador, P., Zak, M.R., 2004. The plant traits that drive ecosystems: Evidence from three continents. J. Veg. Sci. 15, 295–304. Direccion General de Conservacion de la Naturaleza, 2006. Tercer Inventario Forestal Nacional 1997–2006. Ministerio de Medio Ambiente, Madrid, SP. Dormann, C.F.C., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G., Marquéz, J.R.G., Gruber, B., Lafourcade, B., Leitão, P.J., Münkemüller, T., Mcclean, C., Osborne, P.E., Reineking, B., Schröder, B., Skidmore, A.K., Zurell, D., Lautenbach, S., 2013. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46. Dubuis, A., Rossier, L., Pottier, J., Pellissier, L., Vittoz, P., Guisan, A., 2013. Predicting current and future spatial community patterns of plant functional traits. Ecography. 36, 1–11. Eggers, J., Linder, M., Zuddin, S., Zaehle, S., Liski, J., 2008. Impact of changing wood demand, climate and land use on European forest resources and carbon stocks during the 21st century. Glob. Chang. Biol. 14, 2288–2303. Frenette-Dussault, C., Shipley, B., Meziane, D., Hingrat, Y., 2013. Trait-based climate change predictions of plant community structure in arid steppes. J. Ecol. 101, 484–492. Galiano, L., Martinez-Vilalta, J., Lloret, F., 2010. Drought-induced multifactor decline of scots pine in the pyrenees and potential vegetation change by the expansion of cooccurring Oak species. Ecosystems 13, 978–991. Garnier, E., Cortez, J., Billès, G., Navas, M., 2004. Plant functional markers capture ecosystem properties during secondary succession. Ecology 85, 2630–2637. Garzón, M.B., Sánchez de Dios, R., Sainz Ollero, H., 2008. Effects of climate change on the distribution of Iberian tree species. Appl. Veg. Sci. 11, 169–178. Giorgi, F., 2006. Climate change hot-spots. Geophys. Res. Lett. 33, 1–4. Giorgi, F., Lionello, P., 2008. Climate change projections for the Mediterranean region. Glob. Planet. Change 63, 90–104. Grime, J., 1998. Benefits of plant diversity to ecosystems: immediate, filter and founder effects. J. Ecol. 86, 902–910. Grossiord, C., Granier, A., Ratcliffe, S., Bouriaud, O., Bruelheide, H., Chećko, E., Forrester, D.I., Dawud, S.M., Finér, L., Pollastrini, M., Scherer-Lorenzen, M., Valladares, F., Bonal, D., Gessler, A., 2014. Tree diversity does not always improve resistance of

5. Conclusions Based on environmental filtering theory our approach provides an easy and logical procedure to assess climate change impacts on the functional structure of forests stands, potentially linked to the ecosystem services they provide (e.g. carbon storage or productivity). The study also proves useful for orienting forest management practices in current forest stands, such as thinning and pruning operations, while promoting species with those trait values more adapted to future potential conditions. In addition, we encourage continuity over time of forest inventories to support the adequacy of these actions in association of changing climate conditions. Finally, we also recommend improvements in trait databases for these ecosystems, that could also provide information on genetic and phenotypic plasticity, and remote sensing continuous availability to better address and resolve the links among traits, environment and ecosystem functioning. Acknowledgements We thank MAGRAMA for free access to the Spanish Forest Inventory dataset. Costa-Saura J.M. was supported by the mobility programme ERASMUS + and a PhD grant from the University of Sassari. We also thank J.M.C. Lopez for revising the manuscript and previous reviewers which suggestions that improved the manuscript. 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.2017.09.023. 7

Forest Ecology and Management 405 (2017) 1–8

J.M. Costa-Saura et al. forest ecosystems to drought. Proc. Natl. Acad. Sci. 111, 14812–14815. Hanewinkel, M., Cullmann, D.a., Schelhaas, M.-J., Nabuurs, G.-J., Zimmermann, N.E., 2013. Climate change may cause severe loss in the economic value of European forest land. Nat. Clim. Chang. 3, 203–207. Hijmans, R.J., Cameron, A., Parra, J.L., Jones, P.G., Jarvis, A., 2005. The WorldClim interpolated global terrestrial climate surfaces. Int. J. Climatol. 25, 1965–1978. Hooper, D.U., Chapin, F.S., Ewel, J.J., Hector, A., 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecol. Monogr. 75, 3–35. IPCC, 2014. Summary for Policymakers. In: Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J., Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S., Levy, A.N., MacCracken, S., Mastrandrea, P.R., L.L.W. (Eds.), Climate Change 2014: Impacts, Adaptation and Vulnerability - Contributions of the Working Group II to the Fifth Assessment Report. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 1–32. Jactel, H., Brockerhoff, E.G., 2007. Tree diversity reduces herbivory by forest insects. Ecol. Lett. 10, 835–848. Jucker, T., Bouriaud, O., Avacaritei, D., Coomes, D.a., 2014. Stabilizing effects of diversity on aboveground wood production in forest ecosystems: linking patterns and processes. Ecol. Lett. 17, 1560–1569. Keddy, P., 1992. Assembly and response rules: two goals for predictive community ecology. J. Veg. Sci. 3, 157–164. Keenan, R.J., 2015. Climate change impacts and adaptation in forest management : a review. Ann. For. Sci. 72, 145–167. Kerhoulas, L.P., Kolb, T.E., Hurteau, M.D., Koch, G., 2013. Managing climate change adaptation in forests: A case study from the U. S. Southwest. J. Appl. Ecol. 50, 1311–1320. Laliberté, E., Legendre, P., 2010. A distance-based framework for measuring functional diversity from multiple traits. Ecology 91, 299–305. Lavorel, S., 2013. Plant functional effects on ecosystem services. J. Ecol. 101, 4–8. Lavorel, S., Díaz, S., Cornelissen, J.H.C., Garnier, E., Harrison, S.P., Mcintyre, S., Pausas, J.G., Catherine, N.P., Carlos, R., Roumet, C., Pérez-harguindeguy, Natalia, Urcelay, C., 2007. Plant functional types: are we getting any closer to the Holy Grail? In: Canadell, J.G., Pataki, D.E., Pitelka, L.F. (Eds.), Terrestrial Ecosystems in a Changing World. Springer-Verlag, Berlin, Heidelberg, DE, pp. 149–160. Lavorel, S., Grigulis, K., Lamarque, P., Colace, M.-P., Garden, D., Girel, J., Pellet, G., Douzet, R., 2011. Using plant functional traits to understand the landscape distribution of multiple ecosystem services. J. Ecol. 99, 135–147. Lebourgeois, F., Gomez, N., Pinto, P., Mérian, P., 2013. Mixed stands reduce Abies alba tree-ring sensitivity to summer drought in the Vosges mountains, western Europe. For. Ecol. Manage. 303, 61–71. 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. For. Ecol. Manage. 259, 698–709. Maire, V., Gross, N., Börger, L., Proulx, R., Wirth, C., Pontes, L.daS., Soussana, J.F., Louault, F., 2012. Habitat filtering and niche differentiation jointly explain species relative abundance within grassland communities along fertility and disturbance gradients. New Phytol. 196, 497–509. Meehl, G.A., Bony, S., 2011. Introduction to CMIP5. CLIVAR Exch. 16, 4–5. Micco, V. De, Aronne, G., 2012. Morpho-anatomical traits for plant adaptation to drought. In: Aroca, R. (Ed.), Plant Responses to Drought Stress. Springer, Berlin Heidelberg, Berlin, Heidelberg, DE, pp. 37–61. Millar, C.I., Stephenson, N.L., Stephens, S.L., 2007. Climate change and forest of the future: Managing in the face of uncertanity. Ecol. Appl. 17, 2145–2151. Mokany, K., Ash, J., Roxburgh, S., 2008. Functional identity is more important than diversity in influencing ecosystem processes in a temperate native grassland. J. Ecol. 96, 884–893. Mokany, K., Thomson, J.J., Lynch, A.J.J., Jordan, G.J., Ferrier, S., 2015. Linking changes in community composition and function under climate change Linking changes in community composition and function under climate change. Ecol. Appl. 25, 2132–2141. Moor, H., Hylander, K., Norberg, J., 2015. Predicting climate change effects on wetland ecosystem services using species distribution modeling and plant functional traits. Ambio 44, 113–126. Mori, A.S., Furukawa, T., Sasaki, T., 2013. Response diversity determines the resilience of ecosystems to environmental change. Biol. Rev. 88, 349–364. Muscarella, R., Uriarte, M., 2016. Do community-weighted mean functional traits reflect optimal strategies? Proc. R. Soc. B 283, 20152434. Niinemets, Ü., 2001. Global-scale climatic controls of leaf dry mass per area, density, and thickness in trees and shrubs. Ecology 82, 453–469. Pakeman, R.J., Lepš, J., Kleyer, M., Lavorel, S., Garnier, E., Vista, C., 2009. Relative climatic, edaphic and management controls of plant functional trait signatures. J. Veg. Sci. 20, 148–159. Pakeman, R.J., Quested, H.M., 2007. Sampling plant functional traits: What proportion of the species need to be measured ? Appl. Veg. Sci. 10, 91–96. Paquette, A., Messier, C., 2011. The effect of biodiversity on tree productivity: from temperate to boreal forests. Glob. Ecol. Biogeogr. 20, 170–180. Peñuelas, J., Ogaya, R., Boada, M., Jump, S.A., 2007. Migration, invasion and decline: changes in recruitment and forest structure in a warming-linked shift of European beech forest in Catalonia (NE Spain). Ecography 30, 830–838. Poorter, H., Niinemets, U., Poorter, L., Wright, I.J., Villar, R., 2009. Causes and consequences of variation in leaf mass per area (LMA): a meta-analysis. New Phytol. 182, 565–588. Prentice, I.C., Bondeau, A., Cramer, W., Harrison, S.P., Hickler, T., Lucht, W., Sitch, S., Smith, B., Sykes, M.T., 2007. Dynamic global vegetation modelling : quantifying terrestrial ecosystem responses to large-scale environmental change. In: Canadell, J., Pataki, D.E., Pitelka, L.F. (Eds.), Terrestrial Ecosystems in a Changing World.

Springer, Berlin Heidelberg, DE, pp. 175–192. Pretzsch, H., Schütze, G., Uhl, E., 2013. Resistance of European tree species to drought stress in mixed versus pure forests: Evidence of stress release by inter-specific facilitation. Plant Biol. 15, 483–495. Ramirez-Villegas, J., Jarvis, A., 2010. Downscaling Global Circulation Model Outputs: The Delta Method Decision, Policy Analysis Working Paper No. 1. Reich, P., 2014. The world-wide “fast–slow” plant economics spectrum: a traits manifesto. J. Ecol. 102, 275–301. Resco de Dios, V., Fischer, C., Colinas, C., 2006. Climate change effects on mediterranean forests and preventive measures. New For. 33, 29–40. Roscher, C., Schumacher, J., Gubsch, M., Lipowsky, A., Weigelt, A., Buchmann, N., Schmid, B., Schulze, E.-D., 2012. Using plant functional traits to explain diversityproductivity relationships. PLoS One 7, 1–5. Ruiz-Benito, P., Gómez-Aparicio, L., Paquette, A., Messier, C., Kattge, J., Zavala, M.A., 2014a. Diversity increases carbon storage and tree productivity in Spanish forests. Glob. Ecol. Biogeogr. 23, 311–322. Ruiz-Benito, P., Lines, E.R., Gomez-Aparicio, L., Zavala, M.A., Coomes, D.A., 2013. Patterns and drivers of tree mortality in iberian forests: climatic effects are modified by competition. PLoS One 8, e56843. Ruiz-Benito, P., Madrigal-Gonzalez, J., Ratcliffe, S., Coomes, D.A., Kandler, G., Lehtonen, A., Wirth, C., Zavala, M.A., 2014b. Stand structure and recent climate change constrain stand basal area change in european forests: a comparison across boreal, temperate, and mediterranean biomes. Ecosystems 17, 1439–1454. Ruiz-Labourdette, D., Nogués-Bravo, D., Ollero, H.S., Schmitz, M.F., Pineda, F.D., 2012. Forest composition in Mediterranean mountains is projected to shift along the entire elevational gradient under climate change. J. Biogeogr. 39, 162–176. Scheiter, S., Langan, L., Higgins, S., 2013. Next generation dynamic global vegetation models: learning from community ecology. New Phytol. 198, 957–969. Schröter, D., Cramer, W., Leemans, R., Prentice, I.C., Araújo, M.B., Arnell, N.W., Bondeau, A., Bugmann, H., Carter, T.R., Gracia, C.A., de la Vega-Leinert, A.C., Erhard, M., Ewert, F., Glendining, M., House, J.I., Kankaanpää, S., Klein, R.J.T., Lavorel, S., Lindner, M., Metzger, M.J., Meyer, J., Mitchell, T.D., Reginster, I., Rounsevell, M., Sabaté, S., Sitch, S., Smith, B., Smith, J., Smith, P., Sykes, M.T., Thonicke, K., Thuiller, W., Tuck, G., Zaehle, S., Zierl, B., 2005. Ecosystem service supply and vulnerability to global change in Europe. Science 310, 1333–1337. Shipley, B., 2009. From Plant Traits to Vegetation Structure. Chance and selection in the assembly of ecological communities, Cambridge University Press, Cambridge, UK. Shipley, B., Vile, D., Garnier, É., 2006. From plant traits to plant communities: a statistical mechanistic approach to biodiversity. Science 314, 812–814. Spittlehouse, D.L., Stewart, R.B., 2003. Adaptation to climate change in forest management. BC J. Ecosyst. Manage. 4, 1–11. Stuart Chapin III, F., McFarland, J., David McGuire, A., Euskirchen, E.S., Ruess, R.W., Kielland, K., 2009. The changing global carbon cycle: linking plant-soil carbon dynamics to global consequences. J. Ecol. 97, 840–850. Swenson, N., Enquist, B., 2007. Ecological and evolutionary determinants of a key plant functional trait: wood density and its community-wide variation across latitude and elevation. Am. J. Bot. 94, 451–459. Swenson, N.G., Enquist, B.J., Pither, J., Kerkhoff, A.J., Boyle, B., Weiser, M.D., Elser, J.J., Fagan, W.F., Forero-Montaña, J., Fyllas, N., Kraft, N.J.B., Lake, J.K., Moles, A.T., Patiño, S., Phillips, O.L., Price, C.A., Reich, P.B., Quesada, C.A., Stegen, J.C., Valencia, R., Wright, I.J., Wright, S.J., Andelman, S., Jørgensen, P.M., Lacher Jr, T.E., Monteagudo, A., Núñez-Vargas, M.P., Vasquez-Martínez, R., Nolting, K.M., 2012. The biogeography and filtering of woody plant functional diversity in North and South America. Glob. Ecol. Biogeogr. 21, 798–808. Thuiller, W., Lavorel, S., Sykes, M.T., Araújo, M.B., 2006. Using niche-based modelling to assess the impact of climate change on tree functional diversity in Europe. Divers. Distrib. 12, 49–60. Tomppo, E., Gschwantner, T., Lawrence, M., McRoberts, R.E. (Eds.), 2010. National Forest Inventories Pathways for Common Reporting, Media. Springer, Heidelberg Dordrecht London New York. van Vuuren, D.P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., Hurtt, G.C., Kram, T., Krey, V., Lamarque, J.F., Masui, T., Meinshausen, M., Nakicenovic, N., Smith, S.J., Rose, S.K., 2011. The representative concentration pathways: An overview. Clim. Change 109, 5–31. Vayreda, J., Martinez-Vilalta, J., Gracia, M., Retana, J., 2012. Recent climate changes interact with stand structure and management to determine changes in tree carbon stocks in Spanish forests. Glob. Chang. Biol. 18, 1028–1041. Violle, C., Navas, M.L., Vile, D., Kazakou, E., Fortunel, C., Hummel, I., Garnier, E., 2007. Let the concept of trait be functional!. Oikos 116, 882–892. Weiher, E., Keddy, P., 1995. Assembly rules, null models, and trait dispersion: new questions from old patterns. Oikos 74, 159–164. Wood, S.N., 2004. Stable and efficient multiple smoothing parameter estimation for generalized additive models. J. Am. Stat. Assoc. 99, 673–686. Wright, I.J., Reich, P.B., Westoby, M., Ackerly, D.D., Baruch, Z., Bongers, F., CavenderBares, J., Chapin, T., Cornelissen, J.H.C., Diemer, M., Flexas, J., Garnier, E., Groom, P.K., Gulias, J., Hikosaka, K., Lamont, B.B., Lee, T., Lee, W., Lusk, C., Midgley, J.J., Navas, M.-L., Niinemets, U., Oleksyn, J., Osada, N., Poorter, H., Poot, P., Prior, L., Pyankov, V.I., Roumet, C., Thomas, S.C., Tjoelker, M.G., Veneklaas, E.J., Villar, R., 2004. The worldwide leaf economics spectrum. Nature 12, 821–827. Yang, Y., Zhu, Q., Peng, C., Wang, H., Chen, H., 2015. From plant functional types to plant functional traits: A new paradigm in modelling global vegetation dynamics. Prog. Phys. Geogr. 39, 514–535. Yee, T.W., Mitchell, N.D., Science, V., Dec, N., Thomas, W., Neil, D., 2007. Generalized additive models in plant ecology. J. Veg. Sci. 2, 587–602. Zomer, R.J., Trabucco, A., Bossio, D.A., Verchot, L.V., 2008. Climate change mitigation: A spatial analysis of global land suitability for clean development mechanism afforestation and reforestation. Agric. Ecosyst. Environ. 126, 67–80.

8