1. Introduction

The positive contributions of cushion plants to alpine community structures, biodiversity and species interaction networks have been widely evidenced around the globe (Cavieres and Badano, 2009; Chen et al., 2015a, b; Losapio and Schöb, 2017; Losapio et al., 2019; Gavini et al., 2020). A generally accepted conclusion is that a proportional alpine plant diversity has been increased and/or sustained by the facilitation through the nurse effect of cushion plants (Cavieres et al., 2014, 2016; Chen et al., 2015a, b; Kikvidze et al., 2015; Gavini et al., 2020). For example, cushion plants can increase the plant diversity to ca. 40% in the high Andes (Gavini et al., 2020). In the Himalaya–Hengduan mountains of southwestern China, cushion plants can increase plant diversity of local communities from 5% to 59%, depending on the severity of surrounding environments (Yang et al., 2010, 2017; Chen et al., 2015a, b, 2019). At the global scale, Cavieres et al. (2014, 2016) suggested that facilitation of cushion plants can significantly increase the alpine plant diversity and hence acts as an important insurance for alpine biodiversity.

Consequently, the persistence of alpine cushion plants could be of great importance for the long-term maintenance of alpine biodiversity. However, possibly because cushion plants own very long life histories (Chen et al., 2017 and references therein) and hence difficult to monitor their entire life cycles, up to date, few studies have tried to reveal cushion population dynamics and the underlying mechanisms (but see Chen et al., 2020, 2024). Chen et al. (2020) noted that some populations of the typical alpine cushion plant Arenaria polytrichoides has been experiencing degradation. Later, Chen et al. (2024) confirmed that the high mortality of adult individuals and the difficulties in seedling recruitment jointly lead to the serious population degradation of A. polytrichoides.

To predict a species population dynamics, the first and may be the most important step is to detect what factors affect the seedling establishing processes, such as seed germination and seedling growth. Because seed germination and seedling survival are key processes that directly determine the probability of population recruitment and distribution range dynamics (Gimenez-Benavides et al., 2008; Donohue et al., 2010; Oldfather et al., 2021). Numerous factors may act in a concert to determine seed germination and seedling growth (Fenner and Thompson, 2005). Of those factors, temperature, water and light availability might be the most important ones (Fenner and Thompson, 2005; Koutecká and Lepš, 2009; Baskin and Baskin, 2014; Batlla and Benech-Arnold, 2015; Peng et al., 2017). Particularly, plants may need a cardinal/base temperature to trigger the initial germination, and then germination rate and percentage may increase with increasing temperature (Garciahuidobro et al., 1982; Fenner and Thompson, 2005; Peng et al., 2017). In addition, higher temperature may accelerate seedling growth, but lower water availability could constrain seedling growth and lower light availability may change the resource allocation strategy which would influence the seedling survival (Chen et al., 2020, 2024).

Alpine ecosystems are extremely sensitive to climate warming (Grabherr et al., 1994; Elmendorf et al., 2012; Gottfried et al., 2012; Elsen and Tingley, 2015; Steinbauer et al., 2018). The increasing temperature might cause changes in alpine plant seeds' germination and seedling growth and hence influence their establishment in situ (Briceño et al., 2015 and references therein). We hence are much more interested in how increased temperature will affect the seed germination and seedling growth of cushion plants. In addition, many studies have shown that global warming has led to increased surface aridity and more droughts due to decreased precipitation and increased evaporative demand associated with higher vapor pressure deficit under warmer temperatures (Dai, 2013; Trenberth et al., 2014; Dai et al., 2018). Especially, in alpine sub-nival ecosystems where the soils are poorly developed and have low capacity to hold water, droughts induced by climate warming may exert negative effects in the seedling recruiting processes (Chen et al., 2024). Furthermore, climate warming has persistently driven increases in vegetation cover (Gottfried et al., 2012; Liang et al., 2018), which could exert high pressures of competition for light availability and resources on seedling growth. Our previous studies suggested that high temperature, water and light availability could accelerate seed germination of cushion Arenaria polytrichoides, but also could accelerate the mortality of seedlings (Chen et al., 2020, 2024), implying that the influence of climate warming on alpine cushion population recruitment could be seriously negative.

The high Himalaya–Hengduan mountains region in southwestern China is one of the global biodiversity hotspots (Myers et al., 2000) and it also is a diversity center of alpine cushion plants (Zhang et al., 2022). Various cushion species in this region jointly play important roles in sustaining the local alpine plant diversity (Yang et al., 2010, 2017; Chen et al., 2015a, b, 2019). Of those various cushion plants, Arenaria oreophila Hook. f. (Caryophyllaceae) is a typical one. According to the FOC (Flora of China; http://www.iplant.cn/foc), A. oreophila can distribute from ca. 3500 m to 5300 m a.s.l. However, to our field observation, within a same mountain range on the Baima Snow Mountains in northwestern Yunnan province, China, A. oreophila mainly occurs above 4600 m in alpine screes. We indeed observed quite few individuals of A. oreophila within lower-elevation communities, but the number is extremely small (authors' unpublished data). We hence speculate that those individuals might be residues of populations which previously existed. That might imply that A. oreophila has extincted in local lower-elevation communities due to, possibly, climate warming and associated environmental changes. Since cushion A. oreophila, like many other alpine cushion plants, can sustain alpine plant diversity (Chen et al., 2015a), its population degradation might induce changes in alpine community structures and diversity (Chen et al., 2024). However, we know nothing about what factors drive its population shrinkage, especially at the seedling establishment stages.

In this study, we conducted a laboratory simulation experiment to detect the potential effects of climate warming on seed germination and seedling growth in the cushion Arenaria oreophila. We aimed to detect what factors constrain the population recruitment in early life history stages. By disclosing such issues, we would try to explain why this species mainly and currently distributes in the alpine screes and to predict its population dynamics in the future climates. Specifically, we asked: i) how do temperature, light and water availability affect seed germination, including initial germination time and the maximum germination percentage; ii) how do these factors affect the seedling performance, including seedling growth rate, survival and biomass accumulation?

2. Material and methods 2.1. Study species

Arenaria oreophila Hook. f. (Caryophyllaceae) is a typical cushion-forming plant which mainly occurs in high mountains in southwestern China. It has been proven to act as ecosystem engineer which can facilitate other non-cushion plants (Chen et al., 2015a). In our study region, the species mainly distributes in the high-elevation alpine screes (ca. 4600 m–4800 m), while only few individuals can be found in lower-elevation alpine meadows below 4300 m.

Seeds of Arenaria oreophila were collected in the late growing season (mid-October) of 2022 from the population at 4920 m on the mountain top of Pujin pasture which is located on the Baima Snow Mountains. Specifically, mature fruits from different individuals were collected, mixed, stored in paper envelopes and taken back to laboratory. In the laboratory, fruits were air-dried for one month and seeds were separated. Since our preliminary experiments suggested that the seeds of this species have no dormancy, we just stored the seeds in 4 ℃ until used.

2.2. Laboratory experiments

We carried out experiments in three artificial climate chambers, taking soils collected from the seed source community (Chen et al., 2020) as seedling cultivating medium. According to our previous field investigation, the minimum, average and maximum temperatures of the soil (2-cm depth) in the growing season (mid-May to late-September) of the seed source community were 0.1 ℃ below zero, 6.0 ℃ and 17.2 ℃, respectively (authors' unpublished data). Thus, to assess the effects of temperature, we set temperatures for the three chambers of 0/5 ℃ (low), 10/15 ℃ (intermediate) and 20/25 ℃ (high), respectively, and set 12 h of full light (day; 7000 lx) and 12 h of full dark (night) cycles. In addition, to test the effects of light availability on seed germination and seedling performance, we set the light availability in sections of each chamber to 0, 50 and 100% (7000 lx) using polypropylene shading net (Song et al., 2013; Chen et al., 2020, 2024). Finally, two levels of water treatments were assigned to each treatment of light availability under each temperature condition. We used plastic seedling hole trays, with six holes of 5 cm × 5 cm × 8 cm of each tray, to do the seed germination and seedling cultivating experiments. Firstly, soils were filled into the trays, and we got 18 plastic trays in total. Then, 30 matured seeds were sown into each hole of the trays. Then, six trays were randomly selected and assigned to each of the three climate chambers. In each chamber, two trays were randomly selected and assigned to each of the three light availability (0%, 50% and 100%) where again one tray was randomly selected and assigned to each of the two water availability (100% and 50%). For the 100% water treatment (‘full water’ hereafter), we fully watered the trays, with ca. 500 ml tap water, once every week to keep the soil sufficiently watered; while for the 50% water treatment (‘half water’ hereafter), we slightly watered the trays, with ca. 200 ml tap water, once every two weeks to exert certain water stress.

We checked the trays to record the number of germinated seeds every two days until no more seeds germinated for at least ten days. Then, we checked the trays every five days and recorded the following parameters regarding seedling performance: 1) number of surviving seedlings; 2) seedling length (from shoot base to leaf apex) of five randomly chosen seedlings in each hole of the trays; 3) number of leaves of five randomly chosen seedlings in each hole. Such conduction lasted for about three months, i.e., from 6th December 2022 to 14th March 2023.

At the end of the experiments, we carefully removed any adhering soils in the surviving seedlings. Then, seedlings from one replicating hole of the trays of different treatments were put into a small envelope, dried for 48 h in an oven with 70 ℃ and weighed by an electronic balance. Average dry mass of seedlings were gained by dividing the total seedling dry mass by the number of seedlings from related tray holes.

2.3. Data analyses

We analyzed the effects of temperature, water, light, and all possible interactions on the performances of seeds and seedlings by means of mixed-effect models. The time (days) to initial germination (IGT) and to reach the maximum germination percentage (MGT), the maximum germination percentage (MG), the seedling mortality (SM; number of died seedlings/total number of initially emerged seedlings × 100%), and the length, dry mass and number of leaves of survival seedlings at the end of experiments were set up as response variables; temperature, water, light, and their interactions were set up as fixed factors; the IDs of plastic seedling hole tray (i.e., replicate) was set up as a random effect. All data were square-root transformed to satisfy the assumption of homogeneity of variance. Because seedlings under some certain conditions, such as “high temperature plus zero light”, had become completely dead before the end of experiment, we did not achieve any values of seedling dry mass and leaf number for those seedlings. As a result, for these parameters, we only considered the main effects of temperature, water and light availability in the mixed-effect models, whilst their interactions were not taken into consideration.

All the above analyses were conducted with R v.4.1.3 (https://www.r-project.org/). The mixed-effect models were conducted using the lme() function in the ‘nlme’ package (R Core Team, 2022) and the ‘emmeans’ function to do pairwise comparisons between treatments. The ‘ggplot2’ package (Wickham, 2016) was used to plot all reported figures and the layout of figures was designed with Adobe Illustrator 2021.

3. Results

Results of the mixed-effect models showed that temperature, light and water availability and their interactions all had significant effects on seed and seedling performances (Table 1), but the effects also were dependent on the specific performance of seeds or seedlings (Table 1).

3.1. Seed germination

Higher temperatures (10/15 ℃ and 20/25 ℃) and higher light availability (50% and 100%) quickly initialized seed germination (at least one germinated seed was recorded ca. 4–6 days after sowing) and accelerated them to reach the maximum germination percentage (ca. 9–19 days after sowing); whilst, low temperature (0/5 ℃) significantly delayed both the initial time (ca. 29–32 days after sowing) of seed germination and the time that reached maximum germination percentage (ca. 57–86 days after sowing; P < 0.01; Table 1A, B; Fig. 1). Although the effect of water availability was not significant (P > 0.05, Table 1A, B), at low temperature condition, higher water availability generally accelerated seed germination (Fig. 1a, b). In other words, water interacted with other factors to affect the seed germination time (Table 1A, B).

Generally, higher temperature, water and light availability increased the final germination percentage (P < 0.01; Table 1C; Fig. 1c). However, their effects were dependent on each other (Table 1C; Fig. 1c), implying a complicated determining mechanism of seed germination of our target cushion Arenaria oreophila. Interestingly, regardless of water availability, lower light availability generally increased the final germination percentage at all temperature conditions (Fig. 1c). Particularly, when only considering seeds at the lowest temperature condition, even so light availability was zero, seeds even showed the highest final germination percentages (Fig. 1c). In addition, at given temperature and light availability, higher water availability increased seed germination percentage (Fig. 1c).

3.2. Seedling growth

Higher temperature, light and water availability significantly increased seedling length (P ≤ 0.05; Table 1D; Fig. S1) and accelerated seedling growth rate (Figs. 2 and 3). However, higher temperature simultaneously accelerated the mortality process of seedlings, especially when coupled with lower water and light availability (Fig. 4). When light was absolutely absent (zero), the seedlings died at a faster rate under the high temperature treatment (Fig. 4a, b). Only a few seedlings with full (100%) or half (50%) light availability under higher temperatures survived to the end of experiments (Fig. 4a, b). However, under low temperature, all seedlings survived to the end of experiments, and lower light availability accelerated the seedling growth rate, i.e., increased seedling length (Fig. 2c). Additionally, for those survival seedlings, they produced more leaves at higher temperatures, water and light availability than seedlings at lower temperatures, water and light availability (P < 0.05; Table 1G; Fig. S2).

The survived seedlings with higher temperatures accumulated much more biomass than those with low temperatures (P < 0.01; Table 1F; Fig. 3). Meanwhile, seedlings experienced high light availability showed higher dry mass than seedlings that experienced low light availability under intermediate temperature (P < 0.001; Table 1F; Fig. 3). Although seedlings under low temperature showed different length (Figs. 2c and S1), they actually showed similar dry mass (Fig. 3). Interestingly, water availability seemed to have no effect on the dry mass accumulation of seedlings if we do not consider its interacting effect with temperature and/or light (P = 0.14; Table 1F).

3.3. Seedling mortality

Generally, higher temperatures increased seedling mortality (Fig. 5) and accelerated their mortality process (P < 0.05; Table 1E; Fig. 4). Firstly, when experiencing the highest temperature (20/25 ℃), lower water and light availability led to higher seedling mortality, no seedlings under lower water and light availability survived to the end of experiment; and, more than 80% seedlings under full water (100%) and full light (100%) conditions had died at the end of experiment (Fig. 5). Secondly, when experiencing intermediate temperature (10/15 ℃), seedlings under lower light availability persisted for longer time than those under highest temperature, but similarly suffered very high mortality, especially with lower water availability (Figs. 4 and 5). Thirdly, when experiencing low temperature (0/5 ℃), although different water and light treatments had survival seedlings at the end of experiment, lower light availability generally increased seedling mortality (Figs. 4 and 5). Interestingly, under low temperature, higher water availability increased seedling mortality which was contradictory with those under higher temperatures.

4. Discussion

Although many studies have been conducted to assess the potential effects of climate warming on alpine plant seed germination and seedling growth (Graae et al., 2009; Milbau et al., 2009; Shevtsova et al., 2009; Mondoni et al., 2012; Briceño et al., 2015), few studies focused on alpine cushion plants of which population dynamics may have important impacts on alpine community structure (but see Chen et al., 2020, 2023). Cushion Arenaria oreophila can positively contributes to alpine plant diversity (Chen et al., 2015a), but some of its populations seem to be experiencing degradation (see Introduction). This study evidenced that the seeds and seedlings of cushion A. oreophila are quite sensitive to simulated climate warming. In addition, we also showed that increasing temperature and associated changes in water and light conditions constrain seed germination and seedling growth of A. oreophila, hence providing very important implication in understanding the underlying mechanisms of its populations shrinkage.

Seeds of Arenaria oreophila showed relatively high germination percentage, generally higher than 50% once temperature permits (Fig. 1c), suggesting that its seedling establishment may not be limited by seed dormancy. Additionally, a single cushion individual can produce a large number of seeds in a single growing season (Chen et al., 2017, 2024) and the seeds are quite small hence should be easily dispersed by winds or rain flows. Therefore, seed limitation, which is commonly caused by low seed production, high seed predation and/or low seed dispersal capacity (Gómez et al., 2003; Rey et al., 2006; Song et al., 2013), should not be the main constraint on the population recruitment of cushion A. oreophila. Consequently, other factors possibly play key roles in the population degradation process of this species.

We found that, either with full or half water supplies, higher temperature could initialize seed germination very quickly (ca. 4–6 days after sowing; Fig. 1a), implying that once water availability permits, higher temperature can rapidly initialize seed germination. Such result may further imply that under the condition of climate warming on alpine ecosystems (Grabherr et al., 1994), water from snow melts in the early growing season (late May to early June) may quickly trigger the germination of soil seeds in the field. Furthermore, with relatively higher temperatures, seeds could reach the maximum germination percentage within a short period (ca. 9–19 days after sowing; Fig. 1b). Such explosive germination could be advantageous for seedlings to take sufficient time to establish within a short growing season in alpine ecosystems. However, dangers for such explosive seed germination are also apparent, because risks, such as periodic drought or extreme climate events (Wang, 2006; Chen et al., 2024), for those newly emerged seedlings may happen in the subsequent growing periods (Briceño et al., 2015). In alpine sub-nival ecosystems in the Hengduan mountains in southwestern China, the soils are poorly developed and own low capacity of holding soil water. Moreover, the rains brought by the monsoon from the Indian Ocean usually reaches in late June or even middle July, meaning that there might be a short period of climate drought (Zhang et al., 1997; Wang, 2006). In other words, after the snow melt, there may be a period (ca. one week to one month) of zero rainfall until the rains from the Indian Ocean reach. As a result, the newly emerged seedlings triggered by snow melts might be confront with a period of strong drought stress. Combined with warming temperature, such drought stress may, on the one hand, constrain seedling growth; on the other hand, lead to massive mortality of seedlings (Figs. 4 and 5). Given this is true in the natural field, the sensitivity of seeds and seedlings of A. oreophila to climate warming is indeed disadvantageous for its population recruitment in situ.

Previous studies proved that besides temperature and water, light availability is another important factor that highly affect seed germination and seedling performance (Fenner and Thompson, 2005; Mondoni et al., 2009; Wu et al., 2013). Some plants even have strict light requirements for seed germination (Hitchmough et al., 2011; Peng et al., 2019). Contrary to these studies, we found that seeds with low light availability could reach the highest germination percentage at all temperature levels (Fig. 1c), and even took shorter time to reach the maximum germination percentage than seeds with higher light availability when experiencing low temperature (Fig. 1b). Such results suggested that light availability seems not a constraining factor on seed germination of cushion Arenaria oreophila, at least light is not an essential trigger for its seed germination. However, similar with many other plants, light plays important role in the seedling growth (Lei et al., 2006; Kupferschmid et al., 2014; Bianchi et al., 2021; Yin et al., 2023). On the one hand, low light availability significantly accelerated the shoot growth of seedlings (Figs. 2 and S1). A reasonable explanation is that seedlings allocated more resources to shoot growth to compete for light resource. Theories suggest that plants will allocate more resource to organs that experience high stresses (Thornley, 1972; Bloom et al., 1985). When shaded, the main stress experienced by seedlings might be light stress, hence resulting in more resource allocation to shoot length. Similar results were found by Tingstad et al. (2015) and Chen et al. (2020) who found that plants present lower root: shoot ratios in shaded environments but higher values when light stress was absent. On the other hand, lower light availability significantly accelerated seedling mortality (Figs. 4 and 5). This could be because the rapid shoot growth may need much more resources, but the low photosynthetic efficiency due to low light availability (leaves were dark yellow compared with leaves under full light condition) and the poorly developed root system may not supply enough resources. In natural field, climate warming has dramatically increased vegetation cover (Gottfried et al., 2012; Liang et al., 2018; Rumpf et al., 2022), which would increase shadiness on the soil seeds and hence exerting high light stress on newly emerged seedlings. For the target cushion A. oreophila in this study, seedlings emerge in dense vegetation at the low-elevation meadow might be confronted with intense competition for light, leading to relatively more investment in shoot growth at the expense of stable stems and roots. Together with the frequent drought event in early growing season (Wang, 2006), seedlings might be easily injured and damaged by drought- or temperature-related stresses (Smith et al., 2003; Munier et al., 2010; Chen et al., 2020). That could be a major driver that lead to the rapid mortality of seedlings and ultimately to the degradation of cushion A. oreophila in the natural field, especially in low-elevation but high vegetation covering communities. Although low temperature can delay seed germination and growth rate (Figs. 1a, 2 and 3), together with high water and light availability, it can increase seedling survival (Fig. 1c). This result may partly explain the larger population size of A. oreophila in high-elevation communities where vegetation cover is lower than low-elevation communities, because at least light could be sufficient for seedlings in high-elevation communities.

Cushion Arenaria polytrichoides is a sibling species of A. oreophila, although the populations of A. polytrichoides also are experiencing degradation at low elevations (Chen et al., 2024), its population size are currently much larger than that of cushion A. oreophila (authors' unpublished data). We speculate that A. oreophila is much more sensitive to climate warming than A. polytrichoides, and this possibly had resulted in its faster population degradation than A. polytrichoides. Contradictory with the findings in this study, our previous study showed that seed germination percentage of A. polytrichoides was lower at higher temperatures (10/15 and 20/25 ℃) than that at lower temperatures (0/5 ℃), and seeds of A. polytrichoides took longer time (ca. seven days; Chen et al., 2024) to initialize germination even at high temperatures than seeds of A. oreophila (ca. 4–6 days; Fig. 1a). Although lower light availability can accelerate seedling mortality of these two cushion plants, the mortality rate of A. oreophila is much faster than that of A. polytrichoides (Chen et al., 2024 and this study). Accordingly, these results suggested that A. oreophila is quite sensitive to climate warming, and it had led to its faster population degradation in the natural field. That possibly is the main reason why we observed smaller A. oreophila populations compared with lager populations of cushion A. polytrichoides.

5. Summary and perspectives

In this study, we showed that simulated climate warming can exert strong constraints on the first stage of population recruitment of cushion Arenaria oreophila. We suggested that A. oreophila is much more sensitive to climate warming than its sibling cushion A. polytrichoides. When A. polytrichoides is currently and slowly escaping from warmer environments, i.e., lower elevations (Chen et al., 2023), A. oreophila might had reached its highest elevation range (mountain summits) where the climate conditions is currently suitable for sustaining its populations. That means, there will be no more space for them to escape in the long future. With the ongoing climate warming, local extinction of A. oreophila would possibly be inevitable, and it would happen earlier than the extinction of cushion A. polytrichoides. However, to comprehensively confirm such speculations, it is urgently necessary to take further studies, especially field tests of how climate warming and associated changes in other ecological components influence this species in all life history stages, including responses of pollination efficiency, reproduction, seed, seedling and adult individuals.

Acknowledgments

This study was supported by the Second Tibetan Plateau Scientific Expedition and Research (STEP) program (2019QZKK0502 to H.S.), the Yunnan Applied Basic Research Project (202001AT070060 to J.G.C.), the CAS “Light of West China” Program (J.G.C.), the Young Academic and Technical Leader Raising Foundation of Yunnan Province (202205AC160053 to J.G.C.).

Availability of data and material

All data and reproducible codes used in this study can be offered on reasonable request.

Author contributions

J.G.C. and H.S. conceptualized the idea, J.G.C. designed methodology, R.Y.Z., P.F.Y., X.F.C., J.G.C. and M.S.S. conducted field and laboratory investigations and collected data, R.Y.Z. performed the statistical analyses with inputs from J.G.C., J.G.C. wrote the first manuscript. All authors contributed critically to the revision of the current manuscript and gave final approval for publication.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pld.2023.11.003.