DISCUSSION
Foundation species, by virtue of its special structural or functional attributes, can create an entire ecological community or ecosystem, sustain ecosystem functions and biodiversity; thus, the loss of foundation species can induce broad consequences for associated biota and ecosystem functions (Ellison et al., 2005; Schöb et al., 2012; Thomsen et al., 2018). Thus, explicitly elucidating how foundation species respond to climate warming and how the associated biodiversity changes are crucial for robustly predicting changes in ecosystems, and thus their long-term sustainable management.
As foundation species, cushion plants sustain a prominent proportion of alpine biodiversity (Cavieres & Badano, 2009; Butterfield et al., 2013; Chen et al., 2015a; Kikvidze et al., 2015). Thus, together with climate change, their population dynamics may imply future biodiversity changes in alpine ecosystems. We found that large numbers of cushion individuals that once existed in low-elevation communities had been excluded, making warmer populations now more fragmented than colder populations (Figure 1). This was supported by the facts that the warmer populations show lower population density but more old individuals than those colder populations (Figure 1A-D; Figure 2A; Figure S2) which indicate a lower demographic rate in warmer conditions. After exclusion from communities, the previously cushion-dominated communities gradually turned to climax communities that are overwhelmingly dominated by sedge species (Figure 3; Box 1). All these results suggest that populations of the cushion A. polytrichoides that are experiencing climate warming are simultaneously experiencing degeneration, while populations under cold conditions are maintaining recruitment and keep expanding (Chen et al., 2020a).
Newly established cushion populations in high-elevation habitats, especially those newly uncovered by glacial retreat (Baker & Moseley, 2007), could facilitate the establishment of other species, thereby increasing local plant diversity (Cavieres & Badano, 2009). In contrast, population degeneration after climate warming could potentially induce a thorough change in community structures and (possibly reduced) biodiversity (Figure 3). In our study sites, the mean annual temperature in the past decades has increased at a rate of 0.06 oC yr-1, resulting inca. 70 m in elevation upward shift of the treeline (Baker & Moseley, 2007), which could induce negative influences on high-elevation vegetation (Grabherr et al., 1994). Moreover, upward shifts of lowland species will inevitably induce increases in the diversity, cover and productivity of alpine plant communities (Chen et al., 2011; Gottfried et al., 2012; Steinbauer et al., 2018) which could impose strong competition on cushion plants. Consequently, high-elevation populations of the A. polytrichoides cushion plant would face the same challenges that low-elevation populations face now. Thus, in the long-term, low-elevation populations could become locally extinct and there could be serious risks of degeneration of the high-elevation populations. If cushion plants disappear, secondary extinctions are likely to occur, especially of species exclusively sustained by cushion plants (Losapio & Schöb, 2017). Furthermore, species interaction networks that are now mainly sustained by cushion plants (Losapio & Schöb, 2017; Losapio et al., 2019) would inevitably collapse.
Plants’ dynamics are influenced by ecological factors at different life history stages (Gimenez-Benavides et al., 2008; Oldfather et al., 2021). We found that all study populations except the Yulong population produced sufficient seeds in a single growing season (Figure 2B), suggesting that the species is not subject to seed limitation (Turnbull et al., 2000). Thus, the seed germination and seedling establishment rates in the following growing season could strongly affect recruitment rates. We found that, generally, low temperature and light availability can delay seed germination and reduce final germination percentages, indicating that the frequently low temperatures in the early growing season and cover by snow (Wang, 2006) or surrounding vegetation/litter, may delay and/or reduce soil seed germination in the field. In addition, very few seedlings survived for 17 weeks (Figure S4) which equals to the length of the growing season in their natural alpine ecosystems. These results imply large difficulties for A. polytrichoides seedlings to establish successfully in situ .
Even worse, A. polytrichoides seeds cannot persist in the soil for more than a year (Table S4; Figure S5), indicating that the species has a transient soil seed bank. Thus, if seeds cannot germinate or seedlings cannot survive through the first winter, population recruitment will be extremely constrained. Moreover, extreme climate events (e.g. , drought and coldness) frequently occur in our study region (Wang, 2006). We found that short periods of extreme water stress and low temperature can significantly reduce seedling survival (Figure 4A), but prior growth in mild conditions for sufficient time (90 to 120 days) significantly increases their ability to survive long-term extreme climate events (simulating a long winter; Figure 4; Table S3). Thus, timely germination early in the growing season and the occurrence of extreme climate events during the growing season may be key determinants of recruitment rates. These findings are consistent with expectations that a sufficiently long growing season without damaging events could be essential for seedlings to accumulate resources (e.g. , carbohydrates and various nutrients), allocate them appropriately, and maintain a healthy physiological status, thereby establishing high over-wintering capacity (Luscher et al., 2001). The finding that all transplanted seedlings died by the end of the growing season (Table S5) corroborates this conclusion, because the growing season was too short (ca. 60 days) for them to develop over-wintering capacity.
Furthermore, our transplantation experiment implied that inter-specific competition can hinders seedlings’ establishment. This could be because that other competitive species can delay seed germination and accelerate the death of A. polytrichoides seedlings (Chen et al., 2020a). In addition, surrounding vegetation may also impose certain allelopathic constraints on seed germination and seedling growth (Table S6; Figure S6).
Competitive stress imposed by beneficiary plants plays key role in the decay of previously established individual cushions. Firstly, beneficiary plants generally reduced the cushions’ nutrient contents and two stable isotope (δ13C and δ15N) ratios that provide information on plants’ water-use efficiency and carbon, water and nitrogen balances (Table S7; Figure S7) (Dawson et al., 2002). Secondly, beneficiary plants significantly constrain physiological traits (SLA and LDMC) of cushion individuals (Table S8; Figure S7A, C), which implies short leaf longevity and low resource use efficiency (Wright et al., 2004). SLA is also positively correlated with temperature, light availability (Poorter et al., 2009) and soil nitrogen availability (Ordonez et al., 2009). High cover of beneficiary plants can clearly reduce light availability on cushions’ surfaces, thereby reducing SLA (Figure S8A). Thirdly, beneficiary species could generally reduce the LDMC of individual cushions (Figure S7C), indicating that they may inhibit cushion individuals partially through competition for water and associated stress (Cornelissen et al., 2003). However, SLA increases but LDMC decreases with increases in beneficiary cover (Figure S7B, D). Nevertheless, the combined effects of high SLA and low LDMC induced by increases in beneficiary species probably accelerate the competitive exclusion process of individual cushions (Figure 5I-K) and hence degeneration of cushion populations.
Furthermore, possibly due to the trophic and physiological constraints, beneficiary species significantly reduced flower production of cushion individuals (Figure 5A-H), and had context-dependent, but generally no or slightly negative, effects on their fruit production (Figure 5A-H). This may be partly because the vegetative growth of the beneficiary plants hindered pollinators’ visitations of the flowers, and consequently reduced pollination efficiency and fruit set. In addition, increases in beneficiaries’ cover may induce changes in cushions’ resource allocation patterns, through competition-driven increases in allocations to growth and/or defenses (Schöb et al., 2014b). These results demonstrate that beneficiary species may constrain reproductive functions of cushion plants, and hence future population recruitment.
In conclusion, taking communities organized by typical foundational cushion species as a model system, we here explicitly revealed that when foundation species degenerate due to climate warming, the associated species composition and diversity both change with a possible biodiversity collapse. Additionally, such changing processes are influenced by the changes in series ecological factors which are induced by climate warming. As a result, we strongly suggest that to assess the processes of the dynamics of foundation species and the associated biodiversity changes is critic and effective for future biodiversity conservation concerns.