Soil conditions drive changes in a key leaf functional trait through environmental filtering and facilitative interactions

Acta Oecologica 86 (2018) 1–8

Contents lists available at ScienceDirect

Acta Oecologica journal homepage: www.elsevier.com/locate/actoec

Soil conditions drive changes in a key leaf functional trait through environmental filtering and facilitative interactions

T

Rafael Molina-Venegasa,∗, Abelardo Aparicioa, Sébastien Lavergneb, Juan Arroyoa a b

Departamento de Biología Vegetal y Ecología, Universidad de Sevilla, Apartado 1095, E-41080 Sevilla, Spain Laboratoire D’Écologie Alpine, CNRS Université Grenoble Alpes, BP 53, 38041 Grenoble Cedex 9, France

A R T I C L E I N F O

A B S T R A C T

Keywords: Dolomites Elevation Mediterranean Communities Nurse plants Specific leaf area

Non-random patterns in the functional structure of communities are often interpreted as evidence for different forces governing their assemblage. However, community assembly processes may act antagonistically, countering each other's signatures on the functional structure of communities, which may lead to spurious inferences on the underlying mechanisms. To illustrate this issue, we assessed the joint effects of environmental filtering and facilitative interactions on a key leaf functional trait (i.e. specific leaf area, SLA) in Mediterranean dwarfshrub communities, using a two-scale sampling approach. Specifically, we analyzed differences in communityweighted mean SLA values (CWM-SLA) between communities (community-scale) and between guilds within communities (guild-scale, i.e. individuals sampled in understorey, overstorey and open-ground conditions) across contrasted soil environments and elevational gradients. We found that communities on harsh edaphic conditions (i.e. dolomite habitats) showed significantly lower CWM-SLA values than communities on more fertile habitats. In contrast, elevation was a poor predictor of differences in CWM-SLA between the communities. This suggests that environmental filtering may influence leaf trait variation along soil gradients irrespective of elevation. On the other hand, communities on dolomite habitats showed strong differences in CWM-SLA between understorey (higher CWM-SLA) and either open-ground and overstorey guilds (lower CWM-SLA), whereas communities on more fertile soils showed no differences between the guilds. The strong differences in CWM-SLA between understorey and non-understorey guilds in dolomite communities suggest that facilitative interactions may be particularly at stake under stressful edaphic conditions, thus partially mitigating the effect of environmental filtering (i.e. low SLA values) on communities growing in harsh soils.

1. Introduction Community assembly theory is complex and comprises multifold neutral and niche-based processes (Weiher et al., 2011). Based on the widespread idea that these processes leave tractable imprints on the functional structure of communities, some studies have interpreted non-random patterns as evidence of different dominant forces governing their assemblage (Götzenberger et al., 2012; Kraft et al., 2015; Funk et al., 2017). Environmental filtering and historical contingency may determine the species in the regional pool that can thrive in a given community, according to their suitability to local conditions (Diamond, 1975; Keddy, 1992; Cornwell et al., 2006). Following this rationale, directional changes in community-weighted mean trait values (CWM; Garnier et al., 2004; Violle et al., 2007) along environmental gradients have been interpreted as evidence of habitat filtering processes (Cornwell and Ackerly, 2009; Lebrija-Trejos et al., 2010). However, once species have passed through the environmental filter, biotic ∗

interactions might also shape the functional structure of communities (de Bello et al., 2012). For example, community traits may converge towards certain values if highly competitive species tend to displace weak competitors (Mayfield and Levine, 2010). On the other hand, functional diversity may increase if limiting similarity is the dominant process (Stubbs and Wilson, 2004). Therefore, communities are expected to be shaped by the joint effects of abiotic filtering and biotic interactions (Cavieres et al., 2014; Michalet et al., 2015a). Although competitive interactions have long been the dominant biotic factor in conceptual models of community assembly (see Webb et al., 2002; Mayfield and Levine, 2010), facilitative interactions have recently gained prominence in the literature (e.g. Valiente-Banuet and Verdú, 2013; McIntire and Fajardo, 2014; Soliveres et al., 2015). Facilitation among species is highly dependent on the abiotic environment (Bertness and Callaway, 1994; Callaway, 2007), and is thought to be particularly relevant under harsh environmental conditions (Xiao et al., 2013; Bulleri et al., 2016). Communities in stressful habitats are usually

Corresponding author. Current address: Institute of Plant Sciences, University of Bern, Altenbergrain 21, Bern 3013, Switzerland. E-mail address: [email protected] (R. Molina-Venegas).

https://doi.org/10.1016/j.actao.2017.11.008 Received 11 May 2017; Received in revised form 3 November 2017; Accepted 13 November 2017 Available online 23 November 2017 1146-609X/ © 2017 Elsevier Masson SAS. All rights reserved.

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

Fig. 1. Conceptual model to explain the antagonistic effect of habitat filtering and facilitative interactions on community structure under harsh environmental conditions. At the pioneer stage (left-panel), propagules of species with different stress tolerances (i.e. wide array of trait values) arrive into the community. Then, harsh environmental conditions filter out stressintolerant species from the community, which converges towards optimal trait values (central-panel). Finally, “nurse” species reach a minimum size to provide effective microclimatic amelioration to stress-intolerant species, which are then able to locally increase their abundance and persist under the shrub canopy (right-panel). Thus, facilitative interactions between “nurse” and stress-intolerant species lead to strong differences in trait values between understorey and non-understorey individuals. The arrow represents the temporal dimension.

Elevation gradients are also known to be an important factor in structuring Mediterranean plant communities (Loidi et al., 2015; Molina-Venegas et al., 2016). Many components of regional climate and local environment vary in a non-random fashion along elevation gradients (Lomolino, 2001), and thus the intensity of biotic interactions may be affected accordingly (Callaway et al., 2002; Cavieres et al., 2016). In the context of Mediterranean-type ecosystems, facilitative interactions seem more relevant at low elevation, likely due to a stronger severity of summer drought in comparison to Mediterranean alpine habitats (Cavieres et al., 2006). Therefore, it is reasonable to think that both soil conditions and elevation may drive the functional structure of communities. In this study, we assessed the joint effects of environmental filtering and facilitative interactions on a stress-related functional trait (i.e. SLA) in Mediterranean dwarf-shrub communities, using a two-scale sampling approach. Specifically, we analyzed differences in community-weighted mean SLA values (CWM-SLA) across communities (community-scale) and between guilds within communities (guild-scale, i.e. individuals sampled in understorey, overstorey and open-ground conditions) along contrasted soil environments and elevational gradients. We hypothesized that communities growing on stressful dolomite habitats would show low CWM-SLA values (i.e. higher water-use efficiency and a conservative resource-use strategy), though they would exhibit strong differences in CWM-SLA values between understorey and non-understorey guilds if facilitative interactions are also at stake (Soliveres et al., 2012; see Fig. 1). In contrast, communities growing under more benign conditions (i.e. limestones and particularly mica-schists) would show higher CWM-SLA values, as well as little or no differences between understorey and non-understorey guilds. We also hypothesized a greater effect of environmental filtering and facilitative interactions at low elevation (i.e. low CWM-SLA values and strong differences between understorey and non-understorey guilds) due to a stronger severity of summer drought in comparison to higher elevations (i.e. high CWMSLA values and no differences between guilds).

dominated by species with stress-tolerance trait values (e.g. low specific leaf area, SLA; Wilson et al., 1999; Damschen et al., 2012; Hulshof et al., 2013). However, overstorey “nurse” plants are known to provide microclimatic amelioration in relation to surrounding areas, which may help species with less stress-tolerant trait values (e.g. high SLA) to locally increase their abundances and persist (Bruno et al., 2003; Michalet et al., 2006). Therefore, although harsh environmental conditions are expected to filter communities towards optimal CWM trait values (Mason et al., 2010; Laughlin et al., 2011), facilitative interactions occurring under the same environmental conditions may lead to different CWM values between species that exhibit contrasting niche requirements within the communities (i.e. between understorey and nonunderstorey species, see Fig. 1), thus mitigating the effect of environmental filtering on the functional structure of communities (Soliveres et al., 2012). Therefore, environmental filtering and facilitative interactions could act antagonistically, and may counter each other's signatures on the functional structure of communities. Facilitative interactions are particularly relevant in structuring Mediterranean woody plant communities (Verdú et al., 2009; Rey et al., 2016). Although water availability is the most obvious factor influencing positive interactions among Mediterranean plants (Armas et al., 2011; Pugnaire et al., 2011; Gross et al., 2013), other sources of stress may also be at stake. Mediterranean landscapes show a high incidence of stressful substrates which support highly peculiar and specific plant communities, such as those growing on serpentines (Kruckeberg, 1986; Anacker, 2011), gypsum (Escudero et al., 2015), sandstones (Arroyo and Marañón, 1990) and dolomites (Mota et al., 1993). Among these, dolomite communities have received considerably less attention (but see Pignatti and Pignatti, 2014). Dolomite is a sedimentary rock in which a good deal of calcium has been replaced by magnesium (Jones, 1950), and may impose a severe restriction for plant growth and establishment (Mota et al., 2008). In addition, dolomite soils are sandy and gravelly, prone to drain rapidly and provoke water stress (Michalet et al., 2002). Besides, other less stressful substrates are also common in the Mediterranean, such as those derived from limestones and micaschists. Thus, a directional change in the CWM of a trait related to stress-tolerance (e.g. SLA; Wilson et al., 1999) from communities that grow on harsh dolomite habitats to communities that thrive on less stressful soils (i.e. from low to high CWM-SLA values) may indicate that soil conditions act as an environmental filter influencing community functional structure (Spasojevic and Suding, 2012). On the other hand, facilitative interactions are expected to be particularly at stake in communities on harsh dolomite habitats (Bertness and Callaway, 1994; Callaway, 2007), which may lead to functional divergence between understorey and non-understorey species (Soliveres et al., 2012).

2. Materials and methods 2.1. Study area The Sierra Nevada is a core range of the Baetic-Rifan complex in southern Iberian Peninsula. Its rugged topography (up to 3482 m a.s.l.) harbours a heterogeneous lithology, with a broad central mica-schist core that extends linearly from west-southwest to east-northeast for about 100 km. This siliceous central core is surrounded by a discontinuous limestone border that forms the middle and lower slopes of 2

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

guild (n = 48) in the dataset weighted by their respective species abundances (i.e. community-weighted mean, CWM; Garnier et al., 2004; Violle et al., 2007) as:

the Sierra, together with an extensive dolomite outcrop along its western edge (Martín et al., 2008). The climate is predominantly Mediterranean with precipitation concentrated in spring and autumn (mean annual precipitation is ranged between 200 and 900 mm), while the warmest season coincides with effective drought in summer.

S

CWM =

∑ pi ∗ti i=1

2.2. Vegetation survey and data collection

where S is the total number of species in a given community or guild, ti is the SLA value of the corresponding population of the species i, and pi is the relative abundance of the species i in a given community or guild. In order to test for non-random patterns in the observed CWM-SLA values, we used two different null model-based approaches. First, we tested whether the observed CWM-SLA values of the communities departed from random expectation given SLA values across all populations in the dataset (community-scale analyses). To do so, we compared the observed CWM-SLA values of the communities to null distributions generated by randomly shuffling SLA values across all populations in the dataset 9999 times. Second, we tested whether the observed CWMSLA values of the guilds departed from random expectation given SLA values across populations within their corresponding communities (guild-scale analyses). To do so, we compared the observed CWM-SLA values of the guilds to null distributions generated by randomly shuffling SLA values across populations within each community 9999 times. Note that unlike the community-scale randomization scheme, the latter was constrained within the communities, which allowed us to estimate whether the observed differences in CWM-SLA between guilds within the communities depart from random expectation given the SLA values observed in the communities. We calculated an effect size CWM-SLA score (ES-CWM-SLA) for each community and guild based on the probability P for the observed CWMSLA values to be higher than expected given their corresponding null distributions as:

We selected two south-oriented elevation gradients for each of the three main types of bedrocks present in the Sierra Nevada (dolomite, limestone and mica-schist), being separated from one other along the west-to-east axis of the range. Within each elevation gradient, we drew at random three 75-m-long and 1-m-wide transects at three different elevations (∼1300, ∼1650 and ∼1950 m a.s.l.) that passed through dwarf-shrub vegetation (chamaephytes and few nanophanerophytes) on sunny slopes beyond the edge of the tree canopy, when present (see Molina-Venegas et al., 2016 for details). We only were able to sample one transect-site on the dolomites at ∼1950 m because there is only one dolomitic peak over 1900 m in the Sierra Nevada. Within each transect, we counted the total number of individuals of each population of woody species and subspecies. Given the survey was conducted within a single year, seedlings were excluded because they show a high inter-annual variation in emergence and mortality under Mediterranean conditions (and particularly in the Sierra Nevada, see Mendoza et al., 2009), and may therefore lead to spurious conclusions. In order to measure microhabitat preferences among plants in the communities, each individual was assigned to one of three exclusive categories (guilds); overstorey, understorey and open-ground. The overstorey guild was constituted by those plants that presented at least one individual of any species growing under their canopy projection. Individuals growing under the shrub canopy (i.e. under overstorey plants) were assigned to the understorey guild, and the remainder plants formed the open-ground guild. Individuals that could not be clearly assigned to any of these categories were excluded (< 5% of sampled individuals). We did not consider one limestone transect-site at ∼1950 m because the vegetation of this site was constituted almost exclusively by open-ground individuals, presumably due to recent disturbance. Thus, our final dataset consists in two different matrices of 16 communities (Fig. 2) with 261 populations (community-dataset) and 48 vegetation layers (i.e. guilds) with 261 populations (guild-dataset), respectively. The 261 sampled populations correspond to 74 woody species (see Table S1). Spatial association of juveniles with larger plants has been considered as observational evidence of facilitative interactions in former studies (Siles et al., 2008; Verdú et al., 2009). However, establishing a criterion to distinguish between adult plants and juveniles in our study system may be tricky, given the small size of the vegetation (dwarfshrub communities) and the effect of herbivory, which may increase resemblance between adult and juvenile plants. Although we recognize that looking for juvenile recruitment is the optimal approach to detect the signature of facilitation in natural communities, we decided not to distinguish between adult and juvenile plants. Our approach is rather conservative, and prevent us from making systematic errors in life-stage assignments. We measured the average specific leaf area (SLA) of each population of the species within transects using standardized protocols (Cornelissen et al., 2003). We sampled five healthy individuals (or part of them) per population, taking two fully developed and healthy leaves per individual (see Appendix 2 for details). Importantly, intraspecific trait variation can change the outcome of ecological interactions (Bolnick et al., 2011), and therefore incorporating intraspecific variance in community ecology studies is highly recommended.

P=

number (null < obs ) +

number (null = obs ) 2

10000

This one-side probability was used to calculate ES-CWM-SLA scores by subtracting 0.5 to P and multiplying by 2 (Kelt et al., 1995; Chase et al., 2011; Bernard-Verdier et al., 2012). This non-parametric calculation of effect sizes was preferred to the widely used standardized effect size (SES; e.g. Kembel, 2009; Verdú et al., 2009) due to the asymmetry or otherwise non-normality in many of the null distributions generated in the study. ES scores vary between −1 and 1, with values close to −1 and 1 indicating that the observed values are lower and higher than expected based on the null distributions, respectively. 2.4. Statistical analyses First, we conducted unpaired Student's t-tests to evaluate whether lithological origins of soils and elevation explain differences in ESCWM-SLA scores between the communities (community-scale analyses). The ES-CWM-SLA scores of the communities were used as the dependent variable, whereas bedrock (dolomite, limestone and micaschist) and elevation (low, medium and high) were used as two different explanatory factors, separately (i.e. three unpaired t-tests per explanatory factor). Second, we conducted paired t-tests to evaluate whether lithological origins of soils and elevation explain differences in ES-CWM-SLA scores between guilds within the communities (guildscale analyses). To do so, we grouped the communities (n = 48 guilds) in six different ways, according to the different bedrock types and elevational belts; (i) n = 15 guilds in dolomites, (ii) n = 15 in limestones, (iii) n = 18 in mica-schists, (iv) n = 18 at low elevation, (v) n = 18 at medium elevation and (vi) n = 12 at high elevation. Then, we conducted paired t-tests between the levels of the explanatory factor guild (i.e. understorey, overstorey and open-ground) within each group of communities, using the ES-CWM-SLA scores as the dependent variable (i.e. three paired t-tests per group of communities). For all tests, we

2.3. Community-weighted mean SLA values and null model testing We estimated the mean SLA value for each community (n = 16) and 3

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

Fig. 2. Map of the Sierra Nevada in southern Iberian Peninsula showing the location of the sampled communities. The arrow in the inset points to the location of the Sierra Nevada in the western Mediterranean. The pictures show typical Mediterranean dwarf-shrub communities growing under different edaphic conditions in the study area (from left to right; dolomites, limestones and mica-schists).

considered a nominal α level of 5%. Although our data might not fit the assumptions for parametric statistics (i.e. normality could not be tested due to low sample sizes), the Student's t-test is still appropriate when sample size is low (de Winter 2013). All the analyses were conducted in R version 3.2.2 (R Development Core Team, 2015).

At the guild-scale, communities on dolomite habitats showed strong differences in ES-CWM-SLA between the guilds (Fig. 4a and Table 2). Specifically, the understorey guilds showed significantly higher scores than either the open-ground (t = 6.8459, p = 0.002) and the overstorey guilds (t = 3.728, p = 0.02). Communities on limestone showed a similar pattern, although differences between the guilds were only marginally significant (t = 2.6863, p = 0.05 between understorey and open-ground; t = 2.4024, p = 0.07 between understorey and overstorey). In contrast, communities on mica-schist exhibited no significant differences between the guilds (Table 2). Communities at medium elevation showed significant differences between the understorey and either the open-ground (t = 4.4472, p = 0.007) and the overstorey guilds (t = 3.4856, p = 0.02) (Fig. 4b). Finally, high-elevation communities exhibited significant differences between the overstorey and the open-ground guilds (t = 10.59, p = 0.002).

3. Results Our dataset consisted of 11,779 individual plants. Overall, communities on limestones showed higher number of individuals, and so did communities at medium elevation (Table S2). Most individuals were assigned to the open-ground guilds across all substrates and elevations, whilst the overstorey guilds were formed by fewer individuals. At the community scale, we found significant differences in ESCWM-SLA between communities growing on different type of bedrocks (Fig. 3a and Table 1). Communities on dolomite habitats showed significantly lower scores than communities on mica-schists (t = −2.268, p < 0.001; Table 1), the latter showing the highest scores across all the communities. Dolomite communities also showed lower scores than communities on limestones, thought differences were only marginally significant (t = 2.2005, p = 0.06), and the latter showed significantly lower scores than communities on mica-schists (t = −2.7009, p = 0.04). We found no significant differences between communities growing at different elevations (Fig. 3b).

4. Discussion There is increasing evidence that facilitative interactions may enhance diversity in species-rich ecosystems (Hacker and Gaines, 1997; Michalet et al., 2006; Allesina and Levine, 2011; McIntire and Fajardo, 2014), and particularly under harsh environmental conditions (Bulleri et al., 2016). However, other community assembly mechanisms such as environmental filtering could act antagonistically and reduce diversity 4

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

Fig. 3. Box-and-whisker plots showing differences in ES-CWM-SLA scores between communities across bedrock types (a) and elevational belts (b). Values close to −1 and 1 indicate that the observed CWM-SLA values are lower and higher than expected for the given null distributions, respectively. Horizontal dotted lines represent visual references at y = 0 (random pattern). The letters indicate pairs of comparisons with significant differences for a nominal α level of 5% (unpaired t-tests).

sampling approach to provide evidence that both mechanisms may jointly shape the functional structure of plant communities, which caution about making inferences of community assembly mechanisms from the sole use of community-level functional patterns. We found that communities on dolomite habitats showed low CWMSLA values, suggesting than these communities are rather dominated by slow-growing and stress-tolerant species (Westoby et al., 2002; Wright et al., 2002). In contrast, communities on soils derived from limestones and mica-schists showed intermediate and high CWM-SLA values, respectively (Fig. 3a). This directional change in CWM-SLA suggests that the functional structure of communities may be determined by differential environmental filtering acting along contrasting edaphic conditions (Cornwell and Ackerly, 2009; Lebrija-Trejos et al., 2010; Spasojevic and Suding, 2012). Soils derived from dolomites are nutrient deficient and highly alkaline (Table S3), and show low water-holding capacity (Allison and Stevens, 2001; Michalet et al., 2002). Thus,

Table 1 Results of the unpaired t-tests conducted between communities growing in different bedrock types and elevational belts (differences in ES-CWM-SLA scores). In bold, significant p-values for a nominal α level of 5%. Factor

Comparison

t-statistic

p-value

bedrock

Dolomites – Limestones Dolomites – Mica-schists Limestones – Mica-schists Low – Medium Low – High Medium – High

2.2005 −2.2680 −2.7009 −0.9076 −0.2232 −1.1612

0.06 < 0.001 0.03 0.39 0.83 0.29

elevation

(Laliberté et al., 2014), which may erase the signature of facilitation on the functional structure of communities (Soliveres et al., 2012; Gross et al., 2013; Michalet et al., 2015b). In this study, we used a two-scale

Fig. 4. Box-and-whisker plots showing differences in ES-CWM-SLA scores between guilds (i.e. understorey, overstorey and open-ground) within communities grouped by bedrock types (a) and elevational belts (b). Values close to −1 and 1 indicate that the observed CWM-SLA values are lower and higher than expected for the given null distributions, respectively. Horizontal dotted lines represent visual references at y = 0 (random pattern). The letters indicate pairs of comparisons with significant differences for a nominal α level of 5% (paired ttests).

5

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

Table 2 Results of the paired t-tests conducted between guilds (i.e. understorey, overstorey and open-ground) within communities grouped by bedrock types and elevational belts, respectively (differences in ES-CWM-SLA scores). In bold, significant p-values for a nominal α level of 5%. Environmental axis

Bedrock

Elevation

Grouping factor

Dolomites Limestones Mica-schists Low Medium High

Understorey – Overstorey

Understorey – Open-ground

Overstorey – Open-ground

t-statistic

p-value

t-statistic

p-value

t-statistic

p-value

3.7280 2.4024 −0.3406 2.0091 3.4856 −1.5168

0.02 0.07 0.75 0.10 0.02 0.23

6.8459 2.6863 −0.5267 1.8808 4.4472 −0.1494

0.002 0.05 0.62 0.12 0.007 0.89

−1.7001 −1.2468 0.1413 −1.4094 −2.2753 10.5860

0.16 0.28 0.89 0.22 0.07 0.002

dolomite habitats may act as an effective environmental filter selecting slow-growing and stress-tolerant species (or populations) that show low SLA values. In contrast, communities that thrive in more benign soils such as those derived from mica-schists may converge towards high CWM-SLA values, as fast-growing competitive species displace weak competitors and become dominant (Mayfield and Levine, 2010). Further, the lower species richness recorded on mica-schist (n = 12.5 ± 4.09; mean ± SD) in comparison with dolomite habitats (n = 18.8 ± 3.11) supports the competitive exclusion hypothesis (Pausas and Austin, 2001). Finally, soils derived from limestones may occupy an intermediate position between both extremes of the edaphic gradient, allowing species with different resource-uptake strategies to entry the communities. Accordingly, limestone communities sit at the midpoint of the SLA spectrum (Fig. 3a). Communities on dolomite habitats showed strong differences in SLA between guilds, particularly between the understorey (higher SLA) and either the open-ground and the overstorey guilds (lower SLA). In contrast, communities on mica-schists exhibited no differences between guilds, whilst communities on limestones showed slight (though nonsignificant) differences (Fig. 4a). These results are much in line with the stress gradient hypothesis (Bertness and Callaway, 1994), which predicts that facilitative interactions would predominate under harsh environments (He et al., 2013; but see Michalet et al., 2006; Holmgren and Scheffer, 2010). Overstorey “nurse” species (or populations) adapted to the harsh edaphic conditions imposed by dolomite habitats (i.e. plants with low SLA values) may provide microclimatic amelioration to less stress-tolerant understorey individuals (i.e. plants with high SLA values), allowing the latter to persist under stressful local edaphic conditions that otherwise would exclude them from the communities (Bruno et al., 2003). As edaphic conditions become more benign (i.e. limestone and mainly mica-schist) the importance of facilitative interactions for recruitment may decrease, and thus differences in CWM-SLA between guilds would narrow. We admit that other processes may also lead to contrasting differences in CWM-SLA between guilds. For example, difference in light regimes for individuals from understorey (low intensity) versus non-understory (high intensity) may also explain such differences (Gómez et al., 2012). However, significant differences in CWM-SLA between guilds only occurred in dolomite communities, and given that light conditions were similar across edaphic environments, it is unlikely that competition for light is driving such differences. Despite elevation was a weak predictor of community structure, mid-elevation communities also showed significant differences in CWM-SLA between the guilds, particularly between the understorey (higher SLA) and either the open-ground and the overstorey guilds (lower SLA) (Fig. 4b). These results suggest that facilitative interactions may be also at stake in communities growing at medium elevation, which may be subject to a greater water-deficit stress in comparison with more alpine habitats (Cavieres et al., 2006). However, in contrast to our predictions, we found no differences in CWM-SLA between guilds in low-elevation communities. Although we hypothesized a negative linear relationship between facilitation and elevation (i.e. more facilitation at low elevation), differences between the guilds in low-

elevation communities may have diminished due to the collapse of facilitative interactions towards the stressful end of the elevation gradient, which might explain the humped-back shape of the relationship (i.e. maximum facilitation at medium elevation, Michalet et al., 2006). Finally, we found significant differences between the overstorey and the open-ground guilds in high-elevation communities, perhaps suggesting that open-ground individuals are subjected to a greater abiotic stress than “nurse” plants in these communities. However, this result should be interpreted with caution because of the low sample size for highelevation communities in the study. We have shown how soil conditions (and elevation to a minor extent) may drive changes in the functional structure of communities (at least for SLA) through the joint effects of environmental filtering and facilitative interactions. Facilitation among plants have been documented in other stressful substrates such as serpentines (e.g. Oviedo et al., 2014). Serpentine soils are long known to have a profound effect on plant cover, structure and composition of plant communities (Harrison and Rajakaruna, 2011), and they share some similarities with dolomite soils (e.g. nutrient deficiencies and low water-holding capacity; Kruckeberg, 1986). Also, serpentine soils are well-known for their distinctive flora, which consists to a large extent of serpentine habitat specialists (Anacker et al., 2011). Importantly, many species with clear preferences for dolomite outcrops are narrow-endemics (Mota et al., 1993) and share some characteristics with serpentine endemisms (e.g. vegetative growth potential, small leaves, sericeous indument; Mota and Valle, 1992; Mota et al., 2011), suggesting an active role of dolomite habitats as drivers for plant adaptation and speciation. Stressful substrates have long attracted the attention of evolutionary ecologists, as they serve as model systems for the study of plant adaptation, speciation, and species interactions (Anacker and Harrison, 2011; Escudero et al., 2015). However, dolomite habitats have passed mostly unnoticed, and future studies should pay particular attention to these characteristic species-rich environments.

Author contributions RMV, JA conceived the ideas, AA conducted the taxonomical identification, SL collaborated in the writing, RMV conducted the vegetation surveys and data collection, led the statistical analyses and wrote the manuscript.

Acknowledgments We would like to thank the whole Sierra Nevada National Park staff and Juan Lorite (University of Granada) for help in the field, Juan Carlos Rubio (IGME) for providing lithological information and to the EVOCA discussion group (http://grupo.us.es/grnm210/web/index. htm). This work was supported primarily by the Spanish Organismo Autónomo de Parques Nacionales via grant 296/2011 and secondarily by the Regional Government of Andalusia (Grant P09-RNM-5280). 6

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

Appendix A. Supplementary data

Garnier, E., Cortez, J., Billès, G., et al., 2004. Plant functional markers capture ecosystem properties during secondary succession. Ecology 85, 2630–2637. http://dx.doi.org/ 10.1890/03-0799. Gómez, S., Guenni, O., de Guenni, B., 2012. Growth, leaf photosynthesis and canopy light use efficiency under differing irradiance and soil N supplies in the forage grass Brachiaria decumbens Stapf. Grass Forage Sci. 68, 395–407. http://dx.doi.org/10. 1007/s10457-015-9858-y. Götzenberger, L., de Bello, F., Bråthen, K.A., et al., 2012. Ecological assembly rules in plant communities—approaches, patterns and prospects. Biol. Rev. 87, 111–127. http://dx.doi.org/10.1111/j.1469-185X.2011.00187.x. Gross, N., Börger, L., Soriano-Morales, S.I., et al., 2013. Uncovering multiscale effects of aridity and biotic interactions on the functional structure of Mediterranean shrublands. J. Ecol. 101, 637–649. http://dx.doi.org/10.1111/1365-2745.12063. Hacker, S.D., Gaines, S.D., 1997. Some implications of direct positive interactions for community species diversity. Ecology 78, 1990–2003. http://dx.doi.org/10.1890/ 0012-9658(1997)078[1990:SIODPI]2.0.CO;2. Harrison, S.P., Rajakaruna, N., 2011. Serpentine: the Evolution and Ecology of a Model System. University of California Press. He, Q., Bertness, M.D., Altieri, A.H., 2013. Global shifts towards positive species interactions with increasing environmental stress. Ecol. Lett. 16, 695–706. http://dx.doi. org/10.1111/ele.12080. Holmgren, M., Scheffer, M., 2010. Strong facilitation in mild environments: the stress gradient hypothesis revisited. J. Ecol. 98, 1269–1275. http://dx.doi.org/10.1111/j. 1365-2745.2010.01709.x. Hulshof, C.M., Violle, C., Spasojevic, M.J., et al., 2013. Intra-specific and inter-specific variation in specific leaf area reveal the importance of abiotic and biotic drivers of species diversity across elevation and latitude. J. Veg. Sci. 24, 921–931. http://dx. doi.org/10.1111/jvs.12041. Jones, H.T., 1950. Magnesium as a plant nutrient. Chem. Ind. 15, 1108–1110. Keddy, P.A., 1992. Assembly and response rules: two goals for predictive community ecology. J. Veg. Sci. 3, 157–164. http://dx.doi.org/10.2307/3235676. Kelt, D.A., Taper, M.L., Meserve, P.L., 1995. Assessing the impact of competition on community assembly: a case study using small mammals. Ecology 76, 1283–1296. http://dx.doi.org/10.2307/1940935. Kembel, S.W., 2009. Disentangling niche and neutral influences on community assembly: assessing the performance of community phylogenetic structure tests. Ecol. Lett. 12, 949–960. Kraft, N.J.B., Godoy, O., Levine, J.M., 2015. Plant functional traits and the multidimensional nature of species coexistence. PNAS 112, 797–802. http://dx.doi.org/ 10.1073/pnas.1413650112. Kruckeberg, A.R., 1986. An essay: the stimulus of unusual geologies for plant speciation. Syst. Bot. 11, 455–463. http://dx.doi.org/10.2307/2419082. Laliberté, E., Zemunik, G., Turner, B.L., 2014. Environmental filtering explains variation in plant diversity along resource gradients. Science 345, 1602–1605. http://dx.doi. org/10.1126/science.1256330. Laughlin, D.C., Fulé, P.Z., Huffman, D.W., et al., 2011. Climatic constraints on trait-based forest assembly. J. Ecol. 99, 1489–1499. http://dx.doi.org/10.1111/j.1365-2745. 2011.01885.x. Lebrija-Trejos, E., Pérez-García, E.A., Meave, J.A., et al., 2010. Functional traits and environmental filtering drive community assembly in a species-rich tropical system. Ecology 91, 386–398. http://dx.doi.org/10.1890/08-1449.1. Loidi, J., Campos, J.A., Herrera, M., et al., 2015. Eco-geographical factors affecting richness and phylogenetic diversity patterns of high-mountain flora in the Iberian Peninsula. Alp. Bot. 125, 137–146. http://dx.doi.org/10.1007/s00035-015-0149-z. Lomolino, M., 2001. Elevation gradients of species-density: historical and prospective views. Glob. Ecol. Biogeogr. 10, 3–13. http://dx.doi.org/10.1046/j.1466-822x.2001. 00229.x. Martín, J.M., Braga, J.C., Gómez-Pugnaire, M.T., 2008. Itinerarios geológicos por Sierra Nevada. Junta de Andalucía, Consejería de Medio Ambiente. Mason, N.W.H., Peltzer, D.A., Richardson, S.J., et al., 2010. Stand development moderates effects of ungulate exclusion on foliar traits in the forests of New Zealand. J. Ecol. 98, 1422–1433. http://dx.doi.org/10.1111/j.1365-2745.2010.01714.x. Mayfield, M.M., Levine, J.M., 2010. Opposing effects of competitive exclusion on the phylogenetic structure of communities. Ecol. Lett. 13, 1085–1093. http://dx.doi.org/ 10.1111/j.1461-0248.2010.01509.x. McIntire, E.J.B., Fajardo, A., 2014. Facilitation as a ubiquitous driver of biodiversity. New Phytol. 201, 403–416. http://dx.doi.org/10.1111/nph.12478. Mendoza, I., Gómez-Aparicio, L., Zamora, R., Matías, L., 2009. Recruitment limitation of forest communities in a degraded Mediterranean landscape. J. Veg. Sci. 20, 367–376. http://dx.doi.org/10.1111/j.1654-1103.2009.05705.x. Michalet, R., Gandoy, C., Joud, D., Pages, J.P., Choler, P., 2002. Plant community composition and biomass on calcareous and siliceous substrates in the northern French Alps: comparative effects of soil chemistry and water status. Arct. Antarct. Alp. Res. 34, 102–113. Michalet, R., Brooker, R.W., Cavieres, L.A., Kikvidze, Z., Lortie, C.J., Pugnaire, F.I., Valiente-Banuet, A., Callaway, R.M., 2006. Do species interactions shape both sides of the humped-back model of species richness in plant communities? Ecol. Lett. 9, 767–773. Michalet, R., Chen, S., An, L., et al., 2015a. Communities: are they groups of hidden interactions? J. Veg. Sci. 26, 207–218. http://dx.doi.org/10.1111/jvs.12226. Michalet, R., Maalouf, J.-P., Choler, P., et al., 2015b. Competition, facilitation and environmental severity shape the relationship between local and regional species richness in plant communities. Ecography 38, 335–345. http://dx.doi.org/10.1111/ ecog.01106. Molina-Venegas, R., Aparicio, A., Lavergne, S., Arroyo, J., 2016. How soil and elevation shape local plant biodiversity in a Mediterranean hotspot. Biodivers. Conserv. 25,

Supplementary data related to this article can be found at http://dx. doi.org/10.1016/j.actao.2017.11.008. References Allesina, S., Levine, J.M., 2011. A competitive network theory of species diversity. PNAS 108, 5638–5642. http://dx.doi.org/10.1073/pnas.1014428108. Allison, J.R., Stevens, T.E., 2001. Vascular flora of ketona dolomite outcrops in bibb county, Alabama. Castanea 66, 154–205. Anacker, B.L., 2011. Phylogenetic patterns of endemism and diversity. In: Harrison, S.P., Rajakaruna, N. (Eds.), Serpentine: the Evolution and Ecology of a Model System. University of California Press, pp. 49–79. Anacker, B.L., Harrison, S.P., 2011. Climate and the evolution of serpentine endemism in California. Evol. Ecol. 26, 1011–1023. http://dx.doi.org/10.1007/s10682-0119532-4. Anacker, B.L., Whittall, J.B., Goldberg, E.E., Harrison, S.P., 2011. Origins and consequences of serpentine endemism in the California flora. Evolution 65, 365–376. http://dx.doi.org/10.1111/j.1558-5646.2010.01114.x. Armas, C., Rodríguez-Echeverría, S., Pugnaire, F.I., 2011. A field test of the stress-gradient hypothesis along an aridity gradient. J. Veg. Sci. 22, 818–827. http://dx.doi.org/10. 1111/j.1654-1103.2011.01301.x. Arroyo, J., Marañón, T., 1990. Community ecology and distributional spectra of Mediterranean shrublands and heathlands in Southern Spain. J. Biogeogr. 17, 163–176. http://dx.doi.org/10.2307/2845324. Bernard-Verdier, M., Navas, M.-L., Vellend, M., et al., 2012. Community assembly along a soil depth gradient: contrasting patterns of plant trait convergence and divergence in a Mediterranean rangeland. J. Ecol. 100, 1422–1433. http://dx.doi.org/10.1111/ 1365-2745.12003. Bertness, M.D., Callaway, R., 1994. Positive interactions in communities. Trends Ecol. Evol. 9, 191–193. http://dx.doi.org/10.1016/0169-5347(94)90088-4. Bolnick, D.I., Amarasekare, P., Araújo, M.S., et al., 2011. Why intraspecific trait variation matters in community ecology. Trends Ecol. Evol. 26, 183–192. http://dx.doi.org/10. 1016/j.tree.2011.01.009. Bruno, J.F., Stachowicz, J.J., Bertness, M.D., 2003. Inclusion of facilitation into ecological theory. Trends Ecol. Evol. 18, 119–125. http://dx.doi.org/10.1016/S0169-5347(02) 00045-9. Bulleri, F., Bruno, J.F., Silliman, B.R., Stachowicz, J.J., 2016. Facilitation and the niche: implications for coexistence, range shifts and ecosystem functioning. Funct. Ecol. 30, 70–78. http://dx.doi.org/10.1111/1365-2435.12528. Callaway, R.M., 2007. Positive Interactions and Interdependence in Plant Communities. Springer, Dordrecht. Callaway, R.M., Brooker, R.W., Choler, P., et al., 2002. Positive interactions among alpine plants increase with stress. Nature 417, 844–848. http://dx.doi.org/10.1038/ nature00812. Cavieres, L.A., Badano, E.I., Sierra-Almeida, A., et al., 2006. Positive interactions between alpine plant species and the nurse cushion plant Laretia acaulis do not increase with elevation in the Andes of central Chile. New Phytol. 169, 59–69. http://dx.doi.org/ 10.1111/j.1469-8137.2005.01573.x. Cavieres, L.A., Brooker, R.W., Butterfield, B.J., et al., 2014. Facilitative plant interactions and climate simultaneously drive alpine plant diversity. Ecol. Lett. 17, 193–202. http://dx.doi.org/10.1111/ele.12217. Cavieres, L.A., Hernández-Fuentes, C., Sierra-Almeida, A., Kikvidze, Z., 2016. Facilitation among plants as an insurance policy for diversity in Alpine communities. Funct. Ecol. 30, 52–59. http://dx.doi.org/10.1111/1365-2435.12545. Chase, J.M., Kraft, N.J., Smith, K.G., et al., 2011. Using null models to disentangle variation in community dissimilarity from variation in α-diversity. Ecosphere 2 art24. Cornelissen, J.H.C., Lavorel, S., Garnier, E., et al., 2003. A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Aust. J. Bot. 51, 335–380. http://dx.doi.org/10.1071/BT02124. Cornwell, W.K., Ackerly, D.D., 2009. Community assembly and shifts in plant trait distributions across an environmental gradient in coastal California. Ecol. Monogr. 79, 109–126. http://dx.doi.org/10.1890/07-1134.1. Cornwell, W.K., Schwilk, D.W., Ackerly, D.D., 2006. A trait-based test for habitat filtering: convex hull volume. Ecology 87, 1465–1471. http://dx.doi.org/10.1890/00129658(2006)87[1465:ATTFHF]2.0.CO;2. Damschen, E.I., Harrison, S.P., Ackerly, D.D., et al., 2012. Endemic plant communities on special soils: early victims or hardy survivors of climate change? J. Ecol. 100, 1122–1130. http://dx.doi.org/10.1111/j.1365-2745.2012.01986.x. de Bello, F., Price, J.N., Münkemüller, T., et al., 2012. Functional species pool framework to test for biotic effects on community assembly. Ecology 93, 2263–2273. http://dx. doi.org/10.1890/11-1394.1. de Winter, J.C.F., 2013. Using the Student's t-test with extremely small sample sizes. Pract. Assess. Res. Eval. 18, 1–12. Diamond, J.M., 1975. Assembly of species communities. In: Cody, M., Diamond, J. (Eds.), Ecology and Evolution of Communities. Harvard University Press, Cambridge, pp. 342–444. Escudero, A., Palacio, S., Maestre, F.T., Luzuriaga, A.L., 2015. Plant life on gypsum: a review of its multiple facets. Biol. Rev. 90, 1–18. http://dx.doi.org/10.1111/brv. 12092. Funk, J.L., Larson, J.E., Ames, G.M., et al., 2017. Revisiting the Holy Grail: using plant functional traits to understand ecological processes. Biol. Rev. 92, 1156–1173. http://dx.doi.org/10.1111/brv.12275.

7

Acta Oecologica 86 (2018) 1–8

R. Molina-Venegas et al.

822–836. http://dx.doi.org/10.1111/j.1654-1103.2012.01410.x. Soliveres, S., Smit, C., Maestre, F.T., 2015. Moving forward on facilitation research: response to changing environments and effects on the diversity, functioning and evolution of plant communities. Biol. Rev. Camb Philos. Soc. 90, 297–313. http://dx.doi. org/10.1111/brv.12110. Spasojevic, M.J., Suding, K.N., 2012. Inferring community assembly mechanisms from functional diversity patterns: the importance of multiple assembly processes. J. Ecol. 100, 652–661. http://dx.doi.org/10.1111/j.1365-2745.2011.01945.x. Stubbs, W.J., Wilson, J.B., 2004. Evidence for limiting similarity in a sand dune community. J. Ecol. 92, 557–567. http://dx.doi.org/10.1111/j.0022-0477.2004.00898.x. Valiente-Banuet, A., Verdú, M., 2013. Plant facilitation and phylogenetics. Annu. Rev. Ecol. Evol. S 44, 347–366. http://dx.doi.org/10.1146/annurev-ecolsys-110512135855. Verdú, M., Rey, P.J., Alcántara, J.M., et al., 2009. Phylogenetic signatures of facilitation and competition in successional communities. J. Ecol. 97, 1171–1180. Violle, C., Navas, M.-L., Vile, D., et al., 2007. Let the concept of trait be functional!. Oikos 116, 882–892. http://dx.doi.org/10.1111/j.0030-1299.2007.15559.x. Webb, C.O., Ackerly, D.D., McPeek, M.A., Donoghue, M.J., 2002. Phylogenies and community ecology. Annu. Rev. Ecol. Evol. S 33, 475–505. http://dx.doi.org/10.1146/ annurev.ecolsys.33.010802.150448. Weiher, E., Freund, D., Bunton, T., et al., 2011. Advances, challenges and a developing synthesis of ecological community assembly theory. Philos. T R. Soc. B 366, 2403–2413. http://dx.doi.org/10.1098/rstb.2011.0056. Westoby, M., Falster, D.S., Moles, A.T., Vesk, P.A., Wright, I.J., 2002. Plant ecological strategies: some leading dimensions of variation between species. Annu. Rev. Ecol. Evol. S 33, 125–159. http://dx.doi.org/10.1146/annurev.ecolsys.33.010802.150452. Wilson, P.J., Thompson, K., Hodgson, J.G., 1999. Specific leaf area and leaf dry matter content as alternative predictors of plant strategies. New Phytol. 143, 155–162. http://dx.doi.org/10.1046/j.1469-8137.1999.00427.x. Wright, I.J., Westoby, M., Reich, P.B., 2002. Convergence towards higher leaf mass per area in dry and nutrient-poor habitats has different consequences for leaf life span. J. Ecol. 90, 534–543. http://dx.doi.org/10.1046/j.1365-2745.2002.00689.x. Xiao, S., Zhao, L., Zhang, J.-L., et al., 2013. The integration of facilitation into the neutral theory of community assembly. Ecol. Model 251, 127–134. http://dx.doi.org/10. 1016/j.ecolmodel.2012.12.018.

1133–1149. http://dx.doi.org/10.1007/s10531-016-1113-y. Mota, J.F., Valle, F., 1992. Notas fitosociológicas sobre los blanquizares béticos. Actas del Simp. Int. Botánica Pius Font Quer 2, 283–290. Mota, J.F., Cabello, J., Valle, F., 1993. Dolomitic vegetation of the baetic ranges. Plant Ecol. 109, 29–45. Mota, J.F., Medina-Cazorla, J.M., Navarro, F.B., et al., 2008. Dolomite flora of the baetic ranges glades (south Spain). Flora 203, 359–375. http://dx.doi.org/10.1016/j.flora. 2007.06.006. Mota, J.F., Sánchez-Gómez, P., Guirado, J.S., 2011. Diversidad vegetal de las yeseras ibéricas. In: El reto de los archipiélagos edáficos para la biología de la conservación. ADIF y Mediterráneo Asesores-Consultores. Oviedo, R., Faife-Cabrera, M., Noa-Monzón, A., et al., 2014. Facilitation allows plant coexistence in Cuban serpentine soils. Plant Biol. J. 16, 711–716. http://dx.doi.org/ 10.1111/plb.12116. Pausas, J.G., Austin, M.P., 2001. Patterns of plant species richness in relation to different environments: an appraisal. J. Veg. Sci. 12, 153–166. http://dx.doi.org/10.2307/ 3236601. Pignatti, E., Pignatti, S., 2014. Plant Life of the Dolomites. Vegetation Structure and Ecology. Springer, New York. Pugnaire, F.I., Armas, C., Maestre, F.T., 2011. Positive plant interactions in the Iberian Southeast: mechanisms, environmental gradients, and ecosystem function. J. Arid. Environ. 75, 1310–1320. http://dx.doi.org/10.1016/j.jaridenv.2011.01.016. R Development Core Team, 2015. R: a Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www. R-project.org/. Rey, P.J., Alcántara, J.M., Manzaneda, A.J., Sánchez-Lafuente, A.M., 2016. Facilitation contributes to Mediterranean woody plant diversity but does not shape the diversityproductivity relationship along aridity gradients. New Phytol. 211, 464–476. http:// dx.doi.org/10.1111/nph.13916. Siles, G., Rey, P.J., Alcántara, J.M., Ramírez, J.M., 2008. Assessing the long-term contribution of nurse plants to restoration of Mediterranean forests through Markovian models. J. App Ecol. 45, 1790–1798. http://dx.doi.org/10.1111/j.1365-2664.2008. 01574.x. Soliveres, S., Torices, R., Maestre, F.T., 2012. Environmental conditions and biotic interactions acting together promote phylogenetic randomness in semi-arid plant communities: new methods help to avoid misleading conclusions. J. Veg. Sci. 23,

8