Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2018 Dec 12;123(4):569–578. doi: 10.1093/aob/mcy207

Functionally distinct assembly of vascular plants colonizing alpine cushions suggests their vulnerability to climate change

Jiri Dolezal 1,2,✉, Miroslav Dvorsky 1, Martin Kopecky 1, Jan Altman 1, Ondrej Mudrak 1, Katerina Capkova 1, Klara Rehakova 1, Martin Macek 1, Pierre Liancourt 1,3
PMCID: PMC6417476  PMID: 30541052

Abstract

Background and Aims

Alpine cushion plants can initially facilitate other species during ecological succession, but later on can be negatively affected by their development, especially when beneficiaries possess traits allowing them to overrun their host. This can be reinforced by accelerated warming favouring competitively strong species over cold-adapted cushion specialists. However, little empirical research has addressed the trait-based mechanisms of these interactions. The ecological strategies of plants colonizing the cushion plant Thylacospermum caespitosum (Caryophyllaceae), a dominant pioneer of subnival zones, were studied in the Western Himalayas.

Methods

To assess whether the cushion colonizers are phylogenetically and functionally distinct, 1668 vegetation samples were collected, both in open ground outside the cushions and inside their live and dead canopies, in two mountain ranges, Karakoram and Little Tibet. More than 50 plant traits related to growth, biomass allocation and resource acquisition were measured for target species, and the phylogenetic relationships of these species were studied [or determined].

Key Results

Species-based trait–environment analysis with phylogenetic correction showed that in both mountain ranges Thylacospermum colonizers are phylogenetically diverse but functionally similar and are functionally different from species preferring bare soil outside cushions. Successful colonizers are fast-growing, clonal graminoids and forbs, penetrating the cushion by rhizomes and stolons. They have higher root-to-shoot ratios, leaf nitrogen and phosphorus concentrations, and soil moisture and nutrient demands, sharing the syndrome of competitive species with broad elevation ranges typical of the late stages of primary succession. In contrast, the species from open ground have traits typical of stress-tolerant specialists from high and dry environments.

Conclusion

Species colonizing tight cushions of T. caespitosum are competitively strong graminoids and herbaceous perennials from alpine grasslands. Since climate change in the Himalayas favours these species, highly specialized subnival cushion plants may face intense competition and a greater risk of decline in the future.

Keywords: Abiotic stress, alpine, arid, competition, facilitation, global warming, Himalaya, Ladakh, plant–plant interactions, subnival plant communities

INTRODUCTION

Alpine cushion plants are characterized by some of the most suitable adaptive strategies in cold regions (Aubert et al., 2014) and ecologists have long been interested in both the adaptation mechanisms for survival in harsh environments and in the plants’ role in ecological interactions (Rauh, 1939; Körner, 2003; Parsons and Gibson, 2009; Schöb et al., 2014). Despite the evidence supporting the positive impact of cushion plants on the local biota and environment (Arroyo et al., 2003; Cavieres et al., 2006; Yang et al., 2010; Mihoč et al., 2016), their role in biotic interactions and ecological succession is complex (Michalet et al., 2014, 2016; Molina-Montenegro et al., 2015; Anthelme et al., 2017), often context-dependent and species-specific (Anthelme et al., 2014; Soliveres et al., 2015; Liancourt et al., 2017). For instance, nurse cushions can initially facilitate other species during primary succession after glacier retreat, and in turn can be negatively affected by their development, especially when beneficiaries possess traits allowing them to overrun their host (Connell and Slatyer, 1977). This can be reinforced by ongoing climate warming favouring competitively strong species over cold-adapted cushion specialists.

Global warming has been shown to threaten species living in the highest and coldest places. High-mountain cushion plants are potentially endangered because competitively stronger species expanding their range from the lower alpine zone may outcompete them (Erschbamer et al., 2011; Gottfried et al., 2012; Doležal et al., 2016); this may eventually lead to biodiversity loss in alpine flora (Pauli et al., 2007; Chen et al., 2011). However, expanding species may face multiple challenges that are similar to those described during primary succession (Doležal et al., 2008). They can be less well adapted to harsh conditions and their establishment and long-term survival may depend on the presence of safe sites (Stöcklin and Bäumler, 1996; Cavieres et al., 2005), a process that can be facilitated or hindered by resident species (Connell and Slatyer, 1977).

Despite the apparently vast colonizable area, successful colonization in the upper alpine zone may rely on only a few microsites where microtopographic, climatic, physical and chemical characteristics are most suitable (Stöcklin and Bäumler, 1996). The large canopies of alpine cushion plants can potentially provide these better growing conditions and serve as nurse habitats for less well adapted species (Antonsson et al., 2009; Haussmann et al., 2010; Cavieres et al., 2016; Liancourt et al., 2017). Alternatively, alpine specialists, the first pioneers of newly deglaciated areas, can pre-empt the most suitable patches and prevent other species from establishing or thriving (Connell and Slatyer, 1977; Dvorský et al., 2013). They could therefore slow down warming-induced species expansion. Understanding the outcome of plant–plant interactions between resident species and species likely to expand their range in the context of safe site limitation is crucial for developing realistic predictions of the consequences of climate change on mountain biota (Anthelme et al., 2014; Michalet et al., 2016).

Dry mountain regions are ideally suited to the examination of this question (Wang, 1988). In the arid Himalayas, most habitable sites are confined to concave topographies that are sheltered from strong winds and have readily available water, such as along glacial river banks, lakes, snowbeds and flat valley bottoms (Řeháková et al., 2017). During primary succession after glacial retreat, these safe sites are first colonized by low-growing hemicryptophytes, often with a deep root anchoring the plants in unstable substrates (Hartmann, 2009; Angel et al., 2016). Thylacospermum caespitosum is a prominent tap-rooted pioneer and is the dominant cushion plant in the Himalayan subnival zones (Klimešová et al., 2011). Interestingly, recent studies from extreme elevations between 5000 and 6000 m found no facilitative effects in Thylacospermum (de Bello et al., 2011a), with more species and individuals growing outside than inside the cushions (Dvorský et al., 2013). Competitive rather than facilitative interactions prevailed within the safe sites occupied by the cushions. Alpine cushion plants can therefore represent a biotic filter that late-successional or range-expanding species have to pass (Cavieres et al., 2005; Michalet et al., 2016; Anthelme et al., 2017). This overall setting could make upper alpine zones in arid Himalayan mountains more resistant than other, wetter mountainous regions of the world.

Himalayan cushion plants are predicted to become increasingly less resistant to outside colonization due to ongoing climate change, which can be inferred from the ecological strategies of species successfully colonizing them. The subnival zone in the arid north-western Himalayas dominated by cushions of T. caespitosum has experienced an unprecedented rise in summer temperature over the past two decades (Bhutiyani et al., 2007), accompanied by an extension of the growing season (Doležal et al., 2016) and rapid glacial retreat (Schmidt and Nüsser, 2017). In addition to warming, the region has also experienced increased summer precipitation (Shrestha et al., 2012; Thayyen et al., 2013). Previous studies documented that clonally spreading species from wetter habitats increased in abundance in response to increasing temperature and precipitation, while non-clonal tap-rooted subnival specialists, typical of arid Himalayan environments, declined in abundance (Dvorský et al., 2016). If these trends continue, pioneer subnival cushion plants, including Thylacospermum, are predicted to be invaded and gradually replaced by competitive clonal species from the lower-elevation alpine grasslands (Doležal et al., 2016). Moreover, the colonization could be enhanced by the frequent dead tissue and cracks observed in old, senescing Thylacospermum canopies (Dvorský et al., 2013).

Despite these predictions, our understanding of ecological strategies associated with biotic interactions in the rapidly warming Himalayas is still limited and requires further in-depth study (Michalet et al., 2016). Previous studies have shown that Thylacospermum is colonized by a phylogenetically diverse set of species across its entire elevation range in the arid north-western Himalayas (Le Bagousse-Pinguet et al., 2018). However, it remains an open question whether colonizers are also functionally diversified species, indicating that there is no specific strategy behind successful colonization. Alternatively, species colonizing cushion plants possess ecophysiological and morphological attributes distinct from those of less successful colonizers that prefer bare soils outside the cushions (Butterfield and Callaway, 2013; Soliveres et al., 2014; Schöb et al., 2017). We hypothesized that plants colonizing Thylacospermum are phylogenetically diverse but functionally similar species, characteristic of more competitive strategies ensuring fast growth and clonal multiplication, along with an affinity for higher soil moisture and nutrients.

In this study, we aimed to explore which traits enhance the species’ success in becoming established inside compact cushion canopies, and whether different traits are associated with species colonizing their living versus dead parts. We also tested whether the success of a species invading a cushion plant is related to its ecological preferences, including optima and ranges along major environmental gradients. Based on theoretical and field evidence (Griggs, 1956; Alliende and Hoffmann, 1985; Michalet et al., 2016), we expected clonal graminoids from lower-elevation grasslands, as well as herbaceous perennials, to be more successful in colonizing the cushion plants than non-clonal annuals, subshrubs or tap-rooted species. Furthermore, we examined the role of species’ evolutionary history (phylogeny) in the trait-based interactions. To test whether the predictive power of functional traits remains significant after taking phylogeny into account, we employed species-based trait–environment analysis with phylogenetic correction. To provide evidence of how general these patterns are, we explored functional trait composition across entire elevation ranges of species in two mountains: the more glaciated, wetter Karakoram and the less glaciated, more continental Little Tibet.

MATERIALS AND METHODS

Study area and sampling design

We collected vegetation samples and plant traits along the entire Thylacospermum elevational range (4800–5850 m) in the Ladakh region of the north-western Indian Trans-Himalayas (Fig. 1). Much of the study area is covered by cold deserts and steppes, alpine grasslands form a belt above the steppe zone, and subnival vegetation is developed at the highest elevations (Dvorský et al., 2011). The elevation gradient with Thylacospermum correlates with a significant decrease in mean annual/summer temperatures from −1.6/7.7 to −10.4/4.4 °C (Dvorský et al., 2013).

Fig. 1.

Fig. 1.

Open in a new tab

Thylacospermum caespitosum is one of the most prominent high-alpine cushion plants in the Himalayas. It is a perennial plant with a deep tap root and it forms very dense and solid cushions. (A) It is a long-lived pioneer and subnival specialist, colonizing infertile glacial substrata early after deposition, and then occupying the spot for a long time (Little Tibet, 5850 m). (B) With the accelerated rate of warming in the Himalayas, Thylacospermum is being gradually replaced by clonally spreading species, as seen around glacial lakes and streams, where alpine grasslands dominated by Cyperaceae and Poaceae form the late stage of postglacial succession (Karakoram, 5250 m).

To assess which traits are associated with species growing inside live and dead cushion canopies (hereafter termed ‘successful colonizers’) and those preferring outside bare soil, and how general these trait differences are, we collected vegetation and trait data across contrasting elevations in two mountain ranges differing in glaciation, geology and partly in plant species pool, but with similar main vegetation types. We sampled four locations in the Nubra Valley (34°45′ N, 77°35′ E) in Eastern Karakoram (4850, 5000, 5100 and 5250 m), and four locations in the Chamser-Lungser range (32°59′ N, 78°24′ E) above Tso Moriri Lake in Little Tibet (5350, 5600, 5750 and 5850 m). Karakoram is less arid (~250 mm annually) due to the stronger influence of winter westerlies; it is characterized by sharp and rugged ridges, narrow and steep valleys ending in vast glaciers, the fronts of which usually start at ~5300 m. Situated in the rain shadow of the High Himalayas and consequently rarely affected by monsoonal precipitation, Little Tibet has a climate that is predominantly continental and arid, with annual precipitation of around 150 mm and a snowline at 5900 m. Hence, these two ranges have different lower and upper distributional limits for vascular plants (Dvorský et al., 2013, 2015).

Zonation of the vegetation is similar between the two mountain ranges, but the particular belts are shifted downwards in Karakoram because of extensive glaciation and higher precipitation (Bhutiyani et al., 2007). In Karakoram, Thylacospermum caespitosum can be found from 4600 to 5480 m, from dry alpine steppes at lower elevations (with poorly developed sandy soils dominated by Tanacetum tibeticum, Artemisia minor and Elymus schrenkianus) to rocky outcrops at higher elevations surrounded by glacier moraines and mesic alpine meadows (dominated by Potentilla pamirica, Poa attenuata and Astragalus confertus). In Little Tibet, the cushions occur from 5100 to 5960 m, from dry alpine screes at lower elevations (dominated by Poa attenuata, Urtica hyperborea and Dracocephalum heterophyllum) to the subnival zone with poorly developed soils covered by algal crusts and few vascular plants (Saussurea gnaphalodes, Draba altaica and Stellaria decumbens). Unlike other studies showing pronounced altitudinal changes in cushion canopy compactness (e.g. Bonanomi et al., 2015), the Thylacospermum cushions studied had tightly knit, stone-like mat-forming canopies along the whole elevation gradient.

The vegetation survey was carried out during the peak of the growing season (August) in 2011. At each location, cushions were systematically surveyed within an area of ~1 ha to allow sufficient replications (n = 66, 61, 69, 77, 70, 70, 70 and 73 in the different locations). The surveyed cushions represented all size classes, from 5 to 132 cm in diameter, the most common size class being 30–50 cm. Three vegetation samples were taken at each cushion, representing three contrasting microhabitats located next to each other (hereafter termed ‘blocks’): two samples were taken from inside the cushion on live and dead parts of the canopy, and the third was the same size and shape as the cushion and was taken at a random location in the open ground outside the cushion at a distance equalling the cushion diameter. We recorded vascular plant species rooting within the respective sample areas and their percentage cover. We also recorded the proportion of dead tissue within the cushion canopy; in total we recorded plant composition in 1668 samples. Our sampling design inevitably led to the three microhabitat samples having different sizes. For this reason, we performed a species-based multivariate analysis (see the Statistical analysis section) that explicitly accounted for sample size differences by filtering them out before calculating the unbiased (unique) species preferences for three microhabitats and assessing the significance of particular traits in predicting these preferences.

Trait measurement

For all 56 species recorded in the vegetation samples, we recorded several whole plant traits, organ traits, clonal growth strategies and species ecological information, making altogether 53 morphological and ecophysiological traits and ecological indicator values measured both in the field and in the laboratory, based on Cornelissen et al. (2003). More than ten individuals were collected for each species (881 individuals in total) at different elevations covering most of the species’ altitudinal ranges in Ladakh during several expeditions in the years 2008–11. Plant individuals were first excavated from the soil, the roots were washed, the shoots and flowers were counted, and plant height was measured before the plant was separated into individual organs for further laboratory analyses.

We measured several plant traits relevant to competitive ability (plant height), allocation (total, leaf, stem and root biomass), growth [specific leaf area (SLA), leaf dry matter content (LDMC) and stem dry matter content (StDMC), leaf and root N and P concentrations], C storage (starch, fructans as non-structural carbohydrates), drought and frost tolerance (free sugars and sugar alcohols), water use efficiency (δ13C), generative (seed mass) propagation and vegetative (clonality and lateral spread) propagation. Phosphorus was determined after digestion in HClO4 using a Shimadzu UV-1650PC spectrophotometer. We measured δ13C and total C and N using an elemental analyser coupled to an isotope ratio mass spectrophotometer (IRMS) at the Stable Isotope Facility, University of California, Davis, USA. Starch and fructans were determined using the Megazyme total starch assay procedure (www.megazyme.com), and ethanol-soluble sugars were determined by anion exchange chromatography with pulsed amperometric detection. Soluble sugars included sugar alcohols such as glycerol, xylitol and arabitol, and simple sugars such as glucose, fructose, sucrose and galactose (for a complete list see Supplementary Data Appendix S1). Total non-structural carbohydrate (NSC) was calculated as the sum of all analysed carbohydrates. These traits are important indicators of plant resource-use strategy, reflecting a fundamental trade-off between the rapid production of biomass (e.g. high SLA, high foliar N and P, low LDMC, long rhizomes) and the efficient conservation of nutrients (low SLA, less negative δ13C, low foliar N and high LDMC). Additional information on the plant traits is given in Supplementary Data Appendix S1.

Clonal propagation

In addition to quantitative traits, each species was classified into one of 20 clonal growth forms based on which organ (rhizomes versus primary tap roots) provides connections between offspring shoots, whether this organ is short or long, whether a plant is able to form adventitious roots, and whether there are special storage organs. There were three exceptions: annuals and biennials, woody plants and cushion plants were assessed according to their whole morphology. Furthermore, species were divided into four space occupancy strategies, based on the rate of lateral spread (spreading, >10 cm per year; non-spreading, <10 cm per year) and persistence of connections between ramets [splitters were plants producing adventitious roots with the main root decaying; integrators were plants not producing adventitious roots and/or with a perennial main root (Klimešová et al., 2011)].

Species’ elevational optima and ecological indicator values

To obtain a robust estimate of the elevational optima and ranges of the species, we calculated response curves fitted with Huisman–Olff–Fresco (HOF) models (Dvorský et al., 2017). Species response curves were derived from 4150 vegetation plots (each 100 m × 100 m) sampled over the entire Ladakh between 1999 and 2014. The dataset contains >122 000 records of occurrence of vascular plant species along the exceptional elevational gradient from 2800 to 6150 m. Species’ optima on five environmental gradients (ecological indicator values) were derived from the vegetation composition of 369 plots (each 100 m2) sampled in a stratified design to cover major vegetation types over the study area.

Species phylogeny

To assess the possible effect of phylogenetic relatedness among species on the species’ trait–habitat relationship, we reconstructed the phylogeny of studied species from nucleotide sequences obtained from GenBank (www.ncbi.nlm.nih.gov/nuccore/). All the cushion taxa had to be covered, yet many of them were never sequenced. A combined multigene approach was therefore applied. The maximum data density was achieved with four loci: internal transcribed spacer (ITS), trnT-trnL intergenic spacer, matK + trnK region, and the gene for rubisco large subunit (rbcL). The L-INS-i algorithm implemented in the online version of MAFFT 6 (http://mafft.cbrc.jp/alignment/server) was employed to align the sequence datasets. The phylogenetic analysis itself was represented by the Bayesian inference, conducted in MrBayes 3.1.2 (Ronquist and Huelsenbeck, 2003). The tree inferred by MrBayes served as groundwork for phylogenetically independent contrasts in the trait–environment analysis.

Statistical analyses

To determine which set of traits best explain the species’ preferences for three studied microhabitats (outside cushion, inside living and inside dead canopy), we used a species-based multivariate analysis of the trait–environment relation (ter Braak and Šmilauer, 2012), which places individual taxa into focus. It is based on a two-stage analysis in which the preferences of individual species for three studied microhabitats were first quantified as their scores on partial canonical correspondence analysis (CCA) ordination axes after filtering out the effect of elevation, cushion size and block, using the CANOCO 5 software (ter Braak and Šmilauer, 2012). As a second step, the CCA scores were related to the species’ traits using partial redundancy analysis (RDA).

However, since part of the explanatory power uncovered by relating the traits to ecological preferences might be alternatively explained by phylogenetic inertia affecting both the similarity of traits among closely related taxa and the similarity of environmental niches that such taxa occupy, phylogenetic corrections based on the methods of Diniz-Filho (1998) and modified by Desdevises et al. (2003) were used. The variation explained by the phylogenetic relatedness of species was removed from the model, using species coordinates on selected axes of a principal coordinates analysis calculated from a patristic distance matrix corresponding to the MrBayes phylogenetic tree. Selected principal coordinates, which were used as covariates during tests including phylogenetic correction, were also used as predictors of functional traits to estimate the amount of variation in the trait values explained by species phylogeny (Desdevises et al., 2003).

We first analysed all traits together in separate analyses for Little Tibet and Karakoram, followed by analyses specific for each group of traits (whole plant traits, organ traits, clonal growth forms and species’ ecological information; Table 2). Statistical tests were based on 9999 Monte Carlo permutations restricted within the blocks. The results were visualized with an ordination diagram. To test further which traits significantly predicted the species’ preferences for individual cushion habitats, the global RDA analyses were followed by a t-value biplot approach with van Dobben circles (ter Braak and Šmilauer, 2012; Table 1 and Supplementary Data Appendixes S2 and S3). We also used an alternative plot-based approach, which placed individual plots (their community-weighted trait averages) into focus instead of individual taxa. Since the species-based and plot-based analyses provided the same results, we only present the former (for the outcome of the plot-based analysis see Supplementary Data Appendixes S4 and S5).

Table 2.

Results of species-based multivariate analyses, with or without phylogenetic correction (PC), testing the predictive power of life-history traits and ecological indicator values for species’ preferences of contrasting microhabitats, using partial RDA, after accounting for differences in cushion size, elevation and block. Models included all traits or were run separately for mountain ranges and groups of functional traits. Individual cells show pseudo-F values and type I error probability estimates (adjusted by false discovery rates). The percentage of variation explained by ecological preferences (Ecol) and phylogenetic relatedness (Evol) between the species are shown

Analysis Without PC With PC Ecol Evol
F P F P % %
1. Karakoram: all traits 4.2 0.001 1.6 0.028 13.4 23.1
2. Little Tibet: all traits 3.8 0.001 1.6 0.040 18.2 22.6
3. Karakoram: whole plant traits 4.2 0.001 1.3 n.s. 6.5 54.2
4. Karakoram: organ traits 4.2 0.001 1.0 n.s. 9.3 31.1
5. Karakoram: clonal growth forms 4.1 0.001 2.3 0.024 13.7 24.5
6. Karakoram: ecological indicators 4.2 0.001 0.9 n.s. 5.3 11.8
7. Little Tibet: whole plant traits 3.6 0.001 0.5 n.s. 4.4 37.8
8. Little Tibet: organ traits 4.2 0.001 0.9 n.s. 14.4 31.5
9. Little Tibet: clonal growth forms 3.8 0.001 2.2 0.004 27.3 12.6
10. Little Tibet: ecological indicators 3.7 0.001 1.1 n.s. 12.8 11.6
Open in a new tab

n.s., not significant.

Table 1.

List of plant functional traits and ecological indicator values measured on 56 species, their abbreviations, mean values and significant relationships with contrasting microhabitats (open ground outside the Thylacospermum cushion, inside their living and dead canopies) in two mountain ranges, Karakoram and Little Tibet

Traits Abbreviation Mean Karakoram Little Tibet
Whole plant traits
 Plant height (cm) Height 15.8 Inside Inside
 Total dry biomass (g) TotalBiomass 2.79 Outside Outside
 Root biomass (g) RootBiomass 0.87 Outside Outside
 Root biomass investment (%) RootInv 35.8 Inside dead Inside
 Leaf biomass (g) LeafBiomass 0.99 Outside Outside
 Leaf biomass investment (%) LeafInv 28 Outside Outside
 Stem biomass (g) StemBiomass 0.78 Outside Outside
 Stem biomass investment (%) StemInv 26.8
 Flower/fruit biomass (g) FlowerBiomass 0.18 Outside Outside
 Flower/fruit biomass investment (%) FlowerInv 9.4 Inside live
 Below-/above-ground biomass ratio R/S 1.5 Inside dead Inside
 Seed weight (g) SeedWeight 0.07 Inside live
 Number of flowers FlowerNo 12.8 Outside Inside
 Number of above-ground tillers ShootNo 25.5 Outside Outside
 Rooting depth (cm) RootDepth 13.6 Outside Outside
Organ traits
 Leaf nitrogen content (%) LNC 2.6 Inside live
 Leaf phosphorus content (%) LPC 0.18 Inside live
 Leaf carbon content (%) LCC 41.5 Outside
 Leaf carbon/nitrogen ratio C:N 18.3 Inside live
 Leaf δ 13 C (‰) d13C -26.5 Outside
 Leaf dry matter content (mg/g) LDMC 363 Outside
 Stem dry matter content (mg/g) STDMC 399 Inside
 Root nitrogen content (%) RNC 1.5 Outside Outside
 Root phosphorus content (%) RPC 0.16 Outside Outside
 Starch content (%) Starch 3.2 Outside
 Fructan content (%) Fructan 2.7 Outside
 Soluble sugars (%) FreeSugars 4.4 Outside
 Total non-structural carbohydrates (%) TNC 10.3 Outside
 Root water content (%) RootWater 58 Outside
 Stem water content (%) StemWater 66 Inside Live Inside live
 Leaf water content (%) LeafWater 59 Inside live
 Total water content (%) TotWater 66 Inside live
Clonal growth forms
 Hypogeogeic rhizomes >10 cm (%) LongRhizome 7 Inside Inside live
 Hypogeogeic rhizomes <10 cm (%) ShortRhizome 1
 Short rhizomes of turf graminoids (%) TurfGraminoids 24 Inside Live Inside
 Deep tap roots (%) Taprooted 33 Outside Outside
 Long belowground branches (%) LongBgBranches 17 Inside Live
 Cushions with no adventitious roots (%) Cushions 14 Outside
 Plants with bulbs Bulbs <1 Inside Dead
 Annual plants (%) Annual 1 Outside
 Non-spreading integrators (%) IntegratorsNS 48 Outside Outside
 Spreading integrators (%) IntegratorsS 17 Inside live Inside live
 Non-spreading splitters (%) SplittersNS 28 Inside dead Inside dead
 Spreading splitters (%) SplittersS 7 Inside Inside live
Species’ ecological information
 Elevation optima (m a.s.l) Altmean 4975 Outside Outside
 Elevation range (m) Altrange 1616 Inside dead Inside dead
 Elevation minima (m) Altmin 4166 Outside
 Elevation maxima (m) Altmax 5783
 Indicator value_Stability Stability 2.2 Inside dead
 Indicator value_Moisture Moisture 2.5 Inside
 Indicator value_Salinity Salinity 0.3
 Indicator value_Nutrient Nutrient 0.9 Inside
 Indicator value_Shade Shade 0.1
Open in a new tab

RESULTS

In both mountain ranges, the partial RDA analyses showed significant trait differences between the three microhabitats, showing that colonizers of Thylacospermum cushions are functionally different from those of species preferring bare soil outside cushions (Analyses 1 and 2 in Table 2). The first RDA axis was dominant in both regions, explaining more variability than the second axis, allowing one to distinguish between traits associated with outside versus inside the cushion (Fig. 2). The main functional differences along the first axis were associated with a higher representation of more clonally spreading species inside the cushions and non-clonal tap-rooted species outside. The second RDA axis was related to trait differences between species inhabiting live and dead cushion canopies. The pattern remained significant after accounting for phylogenetic relatedness among the species in both mountain ranges (Table 2). This indicated that evolutionarily distant taxa with similar strategies of vegetative propagation are more successful in colonizing cushion plants than non-clonal tap-rooted species.

Fig. 2.

Fig. 2.

Open in a new tab

Life-history traits and ecological indicator values predicting species’ preferences for open ground outside the Thylacospermum cushion, inside their living and dead canopies, using species-based phylogenetically corrected partial RDA analysis, for (A) Karakoram and (B) Little Tibet. Total explained variation (ExV) and P values are shown together with variation explained by the first two ordination axes. For trait abbreviations and significant trait responses to particular habitats, see Table 1.

Species from open-ground assemblages outside the Thylacospermum cushion were mostly composed of non-clonal herbs and subshrubs with deep tap roots, belonging to non-spreading integrators with dense, often cushiony rosettes (Fig. 2). They had a significantly higher leaf dry matter content, leaf C content, higher leaf water use efficiency, lower soil moisture demand, higher elevation optima and narrower elevational ranges than species growing inside cushions. Outside-growing species had significantly higher total dry biomass and, accordingly, stem and root biomass when compared with species growing inside cushions (Table 1). In addition, the outside-growing species had significantly more shoots, a higher number of flowers and deeper roots with more stored water. Finally, open-ground assemblages had significantly more P, N and non-structural carbohydrates in their roots than species growing inside.

While the first RDA axis both in Karakoram and Little Tibet reflects the gradient from clonal species inside cushions to non-clonal species inhabiting open ground outside them, the second axis is related to trait differences between species inhabiting live and dead cushion canopies. There were some common traits but also specific responses in Karakoram and Little Tibet. The plant assemblages occupying the cushion canopies in both mountain ranges were taller plants with higher root-to-shoot ratios, higher stem water contents, lower elevation optima and broader elevational ranges (Fig. 2). Karakoram had more diversity in terms of clonal growth-form strategies in species growing inside cushion canopies than Little Tibet. In Karakoram, living cushion canopies were colonized by clonal forbs with long underground branches bearing above-ground leaf rosettes (spreading integrators), while rhizomatous graminoids with higher affinity for substrate stability, nutrients and moisture were significantly more represented in dead cushions (Fig. 2). Species growing in the live cushion canopies in Little Tibet were long-rhizomatous graminoids and forbs with significantly higher leaf P and N concentrations and stem dry matter contents. Plant assemblages inhabiting the dead cushions in Little Tibet were taller turf graminoids (non-spreading splitters), with broader elevational ranges.

In separate analyses of distinct functional groups (whole plant traits, organ traits, clonal growth traits and species’ ecological information), only the clonal growth forms retained predictive power after accounting for phylogeny (Analyses 3–10 in Table 2).

DISCUSSION

Understanding the ecological role of alpine residents in rapidly warming mountains is important for properly assessing the effects of climate change on species interactions, range shifts and ecological succession (Griggs, 1956; Stöcklin and Bäumler, 1996; Michalet et al., 2011; Schöb et al., 2012, 2014; Butterfield et al., 2013; Bonanomi et al., 2015; Molina-Montenegro et al., 2015). This study provides a detailed exploration of the functional trait differences of entire interacting communities. By determining which set of traits explain the species’ preferences for three contrasting microhabitats (outside cushion, inside living and inside dead canopy), and hence assessing which ecological strategies distinguish plants colonizing the T. caespitosum cushion from those that prefer growing outside, we were able to identify the mechanistic background of plant–plant interactions, and therefore improve our prediction about the successional changes following the ongoing climate change in the Himalayas.

Ecological strategies of plants colonizing the alpine cushions and those staying outside

Our results corroborated glacier forefront successional studies in which stress-tolerant plants are giving way to more competitive ones with increasing time since deglaciation, along with increasing productivity and decreasing disturbances (Walker and del Moral, 2003). A common pattern observed in many arctic and alpine systems is that pioneer, often tap-rooted species colonizing unstable infertile substrates are gradually replaced by clonally spreading late-successional species characteristic of stable substrates and deeper soils (Doležal et al., 2008; Klimešová et al., 2012; Angel et al., 2016). Accordingly, in this study, Thylacospermum, as a dominant pioneer of concave and wetter topographies representing more productive safe sites in arid Himalayan subnival zones, is invaded by species capable of penetrating the cushion by rhizomes and stolons, with mother ramets potentially rooting outside the cushion (Griggs, 1956). Interestingly, we found similar functional properties and ecological preferences in successful colonizers of the Thylacospermum cushions in both mountain ranges, despite differences in glaciation, geology, orography and floristic composition between Karakoram and Little Tibet (Hartmann, 2009). Altogether, these species share the syndrome of more competitive species with broad elevation ranges usually growing in mesic/productive alpine grasslands (Dvorský et al., 2011).

There were also functionally similar assemblies of vascular plants occupying bare ground outside the Thylacospermum cushions in both mountain ranges. Open-ground assemblies were mostly dominated by tap-rooted forbs and subshrubs, similar to what was found with the cushion Laretia acaulis in the Andes (Alliende and Hoffmann, 1985), with a trait syndrome typical of stress-tolerant species specialists of high, cold and dry elevations. For instance, the outside-growing species had higher water use efficiency, more stored carbohydrates for regrowth and osmoprotection (Valluru and van den Ende, 2008) and slower growth (smaller radial increments; Doležal et al., 2016), but also high below- and above-ground biomasses. Those species are often long-lived taxa, such as Artemisia, Oxytropis and Potentilla, that can occupy the spot for several decades (Doležal et al., 2016).

Site-specific responses

Together with the overall trait differences between successful and unsuccessful cushion colonizers in both regions, there were some context-dependent deviations from the general pattern (Schöb et al., 2014). For instance, in Karakoram the live cushion canopies were invaded by species with more diverse clonal growth strategies in both graminoids and forbs, while in Little Tibet graminoids were more exclusively represented. These subtle discrepancies may reflect a difference in the ‘cushion quality effect’ between the two regions (Bonanomi et al., 2015; Chen et al., 2015). The compactness of the cushion canopy plays a crucial role by restricting species unable to penetrate through compact tillers (Soliveres et al., 2015; Michalet et al., 2016), and compactness can vary with environmental conditions (Bonanomi et al., 2015; Chen et al., 2015).

However, our field observations and other studies indicated that Thylacospermum cushions were exceptionally hard and compact throughout their distribution range, naturally hindering germination and penetration by roots of other species and thus reducing their opportunity to establish inside (Jiang et al., 2017). We believe these differences between the two regions are more likely related to higher soil moisture in the strongly glaciated Karakoram (Shrestha et al., 2012) supporting higher floristic diversities (Le Bagousse-Pinguet et al., 2018). In the drier and colder Little Tibet, with extensive semi-deserts and steppes (Wang, 1988), graminoids mostly colonized the cushion plants (de Bello et al., 2011b). The higher continentality and larger species pool of steppic graminoids in Little Tibet (Miehe et al., 2002) may also explain the close association of turf grasses with dead cushion tissues, the pattern being less visible in wetter Karakoram.

Cushion colonizers are phylogenetically diverse but functionally similar

The clonal traits of below-ground organs best predicted the successful colonizers of Thylacospermum even after phylogenetic corrections. Evolutionarily distant taxa with similar strategies of vegetative propagation were growing within the cushion plants (Le Bagousse-Pinguet et al., 2018). These included the graminoids Carex pseudofoetida (Cyperaceae), Poa attenuata and Elymus schrenkianus (Poaceae) and the forbs Dracocephalum heterophyllum (Lamiaceae), Waldheimia tridactylites (Asteraceae) and Potentilla bifurca (Rosaceae). The other studied functional traits, such as ecophysiological leaf and root properties and biomass allocations, were also significant in predicting cushion colonizers, but these differences were mostly driven by evolutionary contrasts. Consequently, though they might be, we cannot ascertain whether these traits are also causally related to successful colonization of the cushion.

The role of cushion plants in ecological succession

Previous studies have shown that the main cushion-facilitative mechanisms involve thermal amelioration (Arroyo et al., 2003), water and nutrient provision (Mihoč et al., 2016) and protection from strong desiccant winds (Cavieres et al., 2006). Although richness and abundance of species were greater outside Thylacospermum cushions (de Bello et al., 2011a; Dvorský et al., 2013), suggesting that competitive rather than facilitative interactions prevail at the community level, we cannot rule out the possibility that improved soil conditions inside cushions contributed to the selective pattern of cushion colonization by vegetatively propagating species with higher moisture and nutrient demands. The Thylacospermum soils with higher organic matter, water content and more neutral pH (more readily available nutrients than in highly alkaline soils outside cushions; Dvorský et al., 2013) could be favourable environments for clonal species with long rhizomes.

Likewise, the dead plant material inside cushions can overheat as surface temperatures can occasionally reach 25–30 °C, even at 5800–6000 m (Dvorský et al., 2015; Doležal et al., 2016). In a companion study, we also documented that dead tissues inside cushions undergo anaerobic fermentation processes through the rich bacterial communities present (Aschenbach et al., 2013; Řeháková et al., 2015). Tall turf graminoids colonizing the decomposing dead tissues, such as Elymus schugnanicus, Stipa subsessiliflora (Poaceae) and Kobresia schoenoides (Cyperaceae), may take advantage of the enhanced N mineralization of cushion plant litter (He et al., 2014). Altogether, competitive and less stress-tolerant species able to cope with the cost of the cushion could also benefit from it (Liancourt et al., 2005), but no conclusion can be drawn without the quantification of a fitness benefit demonstrating that facilitation occurs for these species.

Implications for future changes under a warming scenario

The observed ecologically distinct assembly of plant colonizers on alpine cushions has important implications for future successional changes in the rapidly warming Himalayas (Shrestha et al., 2012). On the one hand, the persistent and competitive cushions likely hinder succession towards alpine grasslands under particularly stressful conditions (Dvorský et al., 2013), either in alpine screes or on very cold sites in the high-elevation subnival zone (de Bello et al., 2011a). On the other hand, our results revealed that the wet clonal grassland species, those that are most likely to expand their range with warming and increased precipitation (Doležal et al., 2016), have the traits required to compete for safe sites occupied by Thylacospermum. This pattern can already be observed around glacial lakes and streams, where alpine grasslands dominated by Cyperaceae and Poaceae form the late stage of postglacial succession (Supplementary Data Appendixes S6–S8).

Paradoxically though, this may not lead to major biodiversity loss because grasslands are the most species-rich communities in the arid Himalayas (Miehe et al., 2002; Dvorský et al., 2011). Moreover, it is unlikely that ‘unsafe’ drier sites occupied by stress-tolerant species would become safer under climate change (i.e. a ridge or a convex topography will not turn into a snowbed). That could lead to spatial coexistence of contrasting strategies. Future studies investigating the effect of climate change in high alpine systems should build on our knowledge of primary succession (Walker and del Moral, 2003), consider the spatial context and safe site limitation (Stöcklin and Bäumler, 1996) and examine further the possibility of resident species to facilitate or hinder the process of range expansion (Anthelme et al., 2014).

SUPPLEMENTARY DATA

Supplementary data are available online at https://academic.oup.com/aob and consist of the following.

Appendix S1: majority rule consensus of trees sampled by the Bayesian analysis of a concatenated partitioned dataset.

Appendix S2: results of the RDA analyses for Karakoram testing which traits are significantly related to particular cushion habitats, using t-value biplots with Van Dobben circles.

Appendix S3: results of the RDA analyses for Little Tibet testing which traits are significantly related to cushion habitats, using a t-value biplot with Van Dobben circles.

Appendix S4: results of the partial plot-based RDA showing differences in species’ functional traits and ecological indicator values.

Appendix S5: plot-based RDA of community-weighted means testing differences in functional traits between contrasting cushion microhabitats.

Appendix S6: Pictures showing Thylacospermum caespitosum as one of the most prominent high-alpine cushion plants in the Himalayas. It is a perennial plant with a woody taproot and it forms very dense and solid cushions (Klimešová et al., 2011). The largest cushions can be more than 150 cm in diameter (Dvorský et al., 2013).

Appendix S7: Pictures depicting Thylacospermum colonising infertile glacial substrata at the species upper elevation limit, Chamser Kangri plateau above Tso Moriri lake in Eastern Ladakh, SW extension of Tibetan Plateau, elevation 5900 m.

Appendix S8: Pictures show Thylacospermum being gradually replaced by clonally spreading species due to accelerated rate of warming in Himalayas. This can be observed around glacial lakes and streams, where alpine grasslands dominated by Cyperaceae and Poaceae form the late stage of postglacial succession. Here are pictures from 5200 m, from Nubra Valley, Eastern Karakoram in northern Ladakh.

Supplementary Appendix
Click here for additional data file. (6.3MB, docx)

ACKNOWLEDGEMENTS

We thank our technicians Karla Kunertová and Monika Průšová for laboratory work. This study was supported by the Czech Science Foundation (GACR 17-19376S). We thank Dr Brian G. McMillan for linguistic improvements.

LITERATURE CITED

  1. Alliende MC, Hoffmann AJ. 1985. Plants intruding Laretia acaulis (Umbelliferae), a high Andean cushion plant. Vegetatio 60: 151–156. [Google Scholar]
  2. Angel A, Conrad R, Dvorský M, et al. 2016. The root-associated microbial community of the world’s highest growing vascular plants. Microbial Ecology 72: 394–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anthelme F, Cavieres LA, Dangles O. 2014. Facilitation among plants in alpine environments in the face of climate change. Frontiers in Plant Science 5: 387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Anthelme F, Meneses RI, Huaman Valero NN, Pozo P, Dangles O. 2017. Fine nurse variations explain discrepancies in the stress-interaction relationship in alpine regions. Oikos 126: 1173–1183. [Google Scholar]
  5. Antonsson H, Björk RG, Molau U. 2009. Nurse plant effect of the cushion plant Silene acaulis (L.) Jacq. in an alpine environment in the subarctic Scandes, Sweden. Plant Ecology 2: 17–25. [Google Scholar]
  6. Arroyo MTK, Cavieres LA, Peñaloza A, Arroyo-Kalin MA. 2003. Positive associations between the cushion plant Azorella monantha (Apiaceae) and alpine plant species in the Chilean Patagonian Andes. Plant Ecology 169: 121–129. [Google Scholar]
  7. Aschenbach K, Conrad R, Řeháková K, Doležal J, Janatková K, Angel R. 2013. Methanogens at the top of the world: occurrence and potential activity of methanogens in newly deglaciated soils in high-altitude cold deserts in the Western Himalayas. Frontiers in Microbiology 4: 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Aubert S, Boucher F, Lavergne S, Renaud J, Choler P. 2014. 1914–2014: a revised worldwide catalogue of cushion plants 100 years after Hauri and Schröter. Alpine Botany 124: 59–70. [Google Scholar]
  9. de Bello F, Doležal J, Dvorský M, et al. 2011. a Cushions of Thylacospermum caespitosum (Caryophyllaceae) do not facilitate other plants under extreme altitude and dry conditions in the north-west Himalayas. Annals of Botany 108: 567–573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. de Bello F, Doležal J, Ricotta C, Klimešová J. 2011b Plant clonal traits, coexistence and turnover in East Ladakh, Trans-Himalaya. Preslia 83: 315–327 [Google Scholar]
  11. Bhutiyani MR, Kale VS, Pawar NJ. 2007. Long-term trends in maximum, minimum and mean annual air temperatures across the Northwestern Himalaya during the twentieth century. Climate Change 85: 159–177. [Google Scholar]
  12. Bonanomi G, Stinca A, Chirico GB, Ciaschetti G, Saracino A, Incerti G. 2015. Cushion plant morphology controls biogenic capability and facilitation effects of Silene acaulis along an elevation gradient. Functional Ecology 30: 1216–1226. [Google Scholar]
  13. ter Braak CJF, Šmilauer P. 2012. CANOCO reference manual and users’s guide: software for ordination (version 50) Ithaca, NY: Microcomputer Power. [Google Scholar]
  14. Butterfield BJ, Callaway RM. 2013. A functional comparative approach to facilitation and its context dependence. Functional Ecology 27: 907–917. [Google Scholar]
  15. Butterfield BJ, Cavieres LA, Callaway R, et al. 2013. Alpine cushion plants inhibit the loss of phylogenetic diversity in severe environments. Ecology Letters 16: 478–486. [DOI] [PubMed] [Google Scholar]
  16. Cavieres LA, Quiroz LC, Molina-Montenegro MA, Muñoz AA, Pauchard A. 2005. Nurse effect of the native cushion plant Azorella monantha on the invasive non-native Taraxacum officinale in the high-Andes of central Chile. Perspectives in Plant Ecology and Evolution 7: 217–226. [Google Scholar]
  17. Cavieres LA, Badano EI, Sierra-Almeida A, Gomez-Gonzalez S, Molina-Montenegro MA. 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 Phytologist 169: 59–69. [DOI] [PubMed] [Google Scholar]
  18. Cavieres LA, Hernández-Fuentes C, Sierra-Almeida A, Kikvidze Z. 2016. Facilitation among plants as an insurance policy for diversity in Alpine communities. Functional Ecology 30: 52–59. [Google Scholar]
  19. Chen IC, Hill JK, Ohlemuller R, Roy DB, Thomas CD. 2011. Rapid range shifts of species associated with high levels of climate warming. Science 333: 1024–1026. [DOI] [PubMed] [Google Scholar]
  20. Chen J, Schöb C, Zhou Z, et al. 2015. Cushion plants can have a positive effect on diversity at high elevations in the Himalayan Hengduan Mountains. Journal of Vegetation Science 26: 768–777. [Google Scholar]
  21. Connell J, Slatyer R. 1977. Mechanisms of succession in natural communities and their role in community stability and organization. American Naturalist 111: 1119–1144. [Google Scholar]
  22. Cornelissen JHC, Lavorel S, Garnier E, et al. 2003. A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Australian Journal of Botany 51: 335–380. [Google Scholar]
  23. Desdevises Y, Legendre P, Azouzi L, Morand S. 2003. Quantifying phylogenetically structured environmental variation. Evolution 57: 2647–2652. [DOI] [PubMed] [Google Scholar]
  24. Diniz-Filho JAF, de Sant’Ana CER, Bini LM. 1998. An eigenvector method for estimating phylogenetic inertia. Evolution 52: 1247–1262. [DOI] [PubMed] [Google Scholar]
  25. Doležal J, Homma K, Vyatkina MP, Yakubov V, Vetrova VP, Hara T. 2008. Primary succession following deglaciation at Koryto Glacier Valley, Kamchatka. Arctic Antarctic and Alpine Research 40: 309–322. [Google Scholar]
  26. Doležal J, Dvorský M, Kopecký M, et al. 2016. Vegetation dynamics at the upper elevational limit of vascular plants in Himalaya. Scientific Reports 6: 24881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Dvorský M, Doležal J, de Bello F, Klimešová J, Klimeš L. 2011. Vegetation types of East Ladakh: species and growth form composition along main environmental gradients. Applied Vegetation Science 14: 132–147. [Google Scholar]
  28. Dvorský M, Doležal J, Kopecký M, et al. 2013. Testing the stress-gradient hypothesis at the roof of the world: effects of the cushion plant Thylacospermum caespitosum on species assemblages. PLoS ONE 8: e53514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Dvorský M, Altman J, Kopecký M, et al. 2015. Vascular plants at extreme elevations in eastern Ladakh, northwest Himalayas. Plant Ecology and Diversity 8: 571–584. [Google Scholar]
  30. Dvorský M, Chlumská Z, Altman J, et al. 2016. Gardening in the zone of death: an experimental assessment of the absolute elevation limit of vascular plants. Scientific Reports 6: 24440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Dvorský M, Macek M, Kopecký M, Wild J, Doležal J. 2017. Niche asymmetry of vascular plants increases with elevation. Journal of Biogeography 44: 1418–1425. [Google Scholar]
  32. Erschbamer B, Unterluggauer P, Winkler E, Mallaun M. 2011. Changes in plant species diversity revealed by long-term monitoring on mountain summits in the Dolomites (northern Italy). Preslia 83: 387–401. [Google Scholar]
  33. Gottfried M, Pauli H, Futschik A, et al. 2012. Continent-wide response of mountain vegetation to climate change. Nature Climate Change 2: 111–115. [Google Scholar]
  34. Griggs RF. 1956. Competition and succession on a Rocky Mountain fellfield. Ecology 37: 8–20. [Google Scholar]
  35. Hartmann H. 2009. A summarizing report on the phytosociological and floristical explorations (1976 – 1997) in Ladakh (India) www.ladakhflowers.com Accessed 10 April 2016.
  36. Haussmann NS, McGeoch MA, Boelhouwers JC. 2010. Contrasting nurse plants and nurse rocks: the spatial distribution of seedlings of two sub-Antarctic species. Acta Oecologica 36: 299–305. [Google Scholar]
  37. He YT, Xu XL, Kueffer C, Zhang XZ, Shi PL. 2014. Leaf litter of a dominant cushion plants shifts nitrogen mineralization to immobilization at high but not low temperature in an alpine meadow. Plant and Soil 383: 415–426. [Google Scholar]
  38. Jiang X, Michalet R, Chen S, et al. 2017. Phenotypic effects of the nurse Thylacospermum caespitosum on dependent plant species along regional climate stress gradients. Oikos 127: 252–263. [Google Scholar]
  39. Klimešová J, Doležal J, Dvorský M, de Bello F, Klimeš L. 2011. Clonal growth forms in eastern Ladakh, Western Himalayas: classification and habitat preferences. Folia Geobotanica 46: 191–217. [Google Scholar]
  40. Klimešová J, Doležal J, Prach K, Košnar J. 2012. Clonal growth forms in Arctic plants and their habitat preferences: a study from Petuniabukta, Spitsbergen. Polish Polar Research 33: 421–442. [Google Scholar]
  41. Körner C. 2003. Alpine plant life: functional plant ecology of high mountain ecosystems. Berlin: Springer. [Google Scholar]
  42. Le Bagousse-Pinguet Y, Liancourt P, de Bello F, et al. 2018. A multi-scale approach reveals random phylogenetic patterns at the edge of vascular plant life. Perspectives in Plant Ecology, Evolution and Systematics 30: 22–30. [Google Scholar]
  43. Liancourt P, Callaway RM, Michalet R. 2005. Stress tolerance and competitive-response ability determine the outcome of biotic interactions. Ecology 86: 1611–1618. [Google Scholar]
  44. Liancourt P, Le Bagousse-Pinguet Y, Rixen C, Doležal J. 2017. SGH: stress or strain gradient hypothesis? Insights from an elevation gradient on the roof of the world. Annals of Botany 120: 29–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Michalet R, Xiao S, Touzard B, et al. 2011. Phenotypic variation in nurse traits and community feedbacks define an alpine community. Ecology Letters 14: 433–443. [DOI] [PubMed] [Google Scholar]
  46. Michalet R, Schob C, Lortie CJ, Brooke RW, Callaway RM. 2014. Partitioning net interactions among plants along altitudinal gradients to study community responses to climate change. Functional Ecology 28: 75–86. [Google Scholar]
  47. Michalet R, Schob C, Xiao S, et al. 2016. Beneficiary feedback effects on alpine cushion benefactors become more negative with increasing cover of graminoids and in dry conditions. Functional Ecology 30: 75–83. [Google Scholar]
  48. Miehe G, Miehe S, Dickoré WB. 2002. Alpine deserts in high Asia. In: Yiang XP, ed. Deserts and Alpine Environments. Advances in Geomorphology and Paleoclimatology. Beijing: China Ocean Press, 59–79. [Google Scholar]
  49. Mihoč MAK, Giménez-Benavides L, Pescador DS, Sánchez AM, Cavieres LA, Escudero A. 2016. Soil under nurse plants is always better than outside: a survey on soil amelioration by a complete guild of nurse plants across a long environmental gradient. Plant and Soil 408: 31–41. [Google Scholar]
  50. Molina-Montenegro MA, Oses R, Acuña-Rodríguez IS, et al. 2015. Positive interactions by cushion plants in high mountains: fact or artifact? Journal of Plant Ecology 9: 117–123. [Google Scholar]
  51. Parsons RF, Gibson N. 2009. The cushion plants of lowland southern Australia. Cunninghamia 11: 177–184. [Google Scholar]
  52. Pauli H, Gottfried M, Reisser K, Klettner C, Grabherr G. 2007. Signals of range expansions and contractions of vascular plants in the high Alps: observations (1994–2004) at the GLORIA master site Schrankogel, Tyrol, Austria. Global Change Biology 13: 147–156. [Google Scholar]
  53. Rauh W. 1939. Uber polsterformigen Wuchs Nova. Acta Leopoldina 7: 267–508. [Google Scholar]
  54. Řeháková K, Chroňáková A, Krištůfek V, et al. 2015. Bacterial community of cushion plant Thylacospermum caespitosum on elevational gradient in the Himalayan cold desert. Frontiers in Microbiology 6: 304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Řeháková K, Dvorský M, Kopecký M, Altman J, Šmilauer P, Doležal J. 2017. Interactions between soil phototrophs and vascular plants in Himalayan cold desert. Soil Biology and Biochemistry 115: 568–578. [Google Scholar]
  56. Ronquist F, Huelsenbeck JP. 2003. MRBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. [DOI] [PubMed] [Google Scholar]
  57. Schmidt S, Nüsser M. 2017. Changes of high altitude glaciers in the trans-Himalaya of Ladakh over the past five decades (1969–2016). Geosciences 7: 27. [Google Scholar]
  58. Schöb C, Butterfield BJ, Pugnaire FI. 2012. Foundation species influence trait-based community assembly. New Phytologist 196: 824–834. [DOI] [PubMed] [Google Scholar]
  59. Schöb C, Michalet R, Cavieres L, et al. 2014. A global analysis of bidirectional interactions in alpine plant communities shows facilitators experiencing strong reciprocal fitness costs. New Phytologist 202: 95–105. [DOI] [PubMed] [Google Scholar]
  60. Schöb C, Macek P, Pistón N, Kikvidze Z, Pugnaire FI. 2017. A trait-based approach to understand the consequences of specific plant interactions for community structure. Journal of Vegetation Science 28: 696–704. [Google Scholar]
  61. Shrestha UB, Gautam S, Bawa KS. 2012. Widespread climate change in the Himalayas and associated changes in local ecosystems. PLoS ONE 7: e36741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Soliveres S, Maestre FT, Bowker MA, et al. 2014. Functional traits determine plant co-occurrence more than environment or evolutionary relatedness in global drylands. Perspectives in Plant Ecology, Evolution and Systematics 16: 164–173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Soliveres S, Smit C, Maestre FT. 2015. Moving forward on facilitation research: response to changing environments and effects on the diversity, functioning and evolution of plant communities. Biological Reviews of the Cambridge Philosophical Society 90: 297–313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Stöcklin J, Bäumler E. 1996. Seed rain, seedling establishment and clonal growth strategies on a glacier foreland. Journal of Vegetation Science 7: 45–56. [Google Scholar]
  65. Thayyen RJ, Dimri AP, Kumar P, Agnihotri G. 2013. Study of cloudburst and flash floods around Leh, India, during August 4–6, 2010. Nature Hazards 65: 2175–2204. [Google Scholar]
  66. Valluru R, van den Ende W. 2008. Plant fructans in stress environments: emerging concepts and future prospects. Journal of Experimental Botany 59: 2905–2916. [DOI] [PubMed] [Google Scholar]
  67. Walker LR, del Moral R. 2003. Primary succession and ecosystem rehabilitation. Cambridge: Cambridge University Press. [Google Scholar]
  68. Wang J. 1988. The steppes and deserts of Xizang Plateau (Tibet). Vegetatio 75: 135–142. [Google Scholar]
  69. Yang Y, Niu Y, Cavieres LA, Sun H. 2010. Positive associations between the cushion plant Arenaria polytrichoides (Caryophyllaceae) and other alpine plant species increase with elevation in the Sino-Himalayas. Journal of Vegetation Science 21: 1048–1057. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Appendix
Click here for additional data file. (6.3MB, docx)

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES