FIGURE 4. Effects of simulated extreme climate events on seedling survival. (A) Survival rates of seedlings assigned to indicated short-term extreme climate events starting at indicated ages; (B) survival rates of seedlings that survived short-term extreme climate events subjected to indicated long-term extreme climate events after growth for 120 days in ‘conducive’ conditions. n = 6 (replicate pots with 30 initially sown seeds), data are mean and standard error (s.e.). See Table S3 for the complete statistical details (Linear mixed-effects models). Very few seedlings in the control (‘conducive’) group died from the start to the end of the experiment, so only their final survival rate is shown.
Germination rates of seeds buried in the field (2 cm deep) were relatively high (more than 60%) in the following May and July. Seeds buried at lower elevation (warmer condition) generally showed higher frequencies of germination than those buried at higher elevation (colder condition; Table S4; Figure S5). No viable seeds remained in July, because if they did not germinate, seeds rotted in the soil. Most seeds and seedlings had rotted in September, and only a small proportion of the seedlings survived to the end of September if they did not break out of the soil (Figure S5).
All transplanted seedlings were still living a week after transplantation, suggesting that the establishment of seedlings in the field was not inhibited by transplantation. However, two months later few seedlings survived possibly due to constraints from surrounding environments (Table S5). More specifically, in late September survival rates of seedlings transplanted within vegetation and in bare ground micro-habitats were similar at high elevation, but at low elevation the seedlings transplanted into bare ground had higher survival rates than those transplanted into vegetation. Surprisingly, no seedlings survived until mid-November in any of the micro-habitats used in the experiment.
Finally, surrounding vegetation also imposed certain allelopathic constraints on seed germination and seedling growth (Table S6; Figure S6).
Beneficiary plants generally reduced the cushions’ nutrient contents and two stable isotope ratios (δ13C and δ15N), although not all effects were significant (Table S7, Figure S7). Beneficiary plants also reduced the specific leaf area (SLA) and leaf dry matter content (LDMC) of individual cushions in all study populations, and these effects were independent of the dominating beneficiary species (Table S8; Figure S8A, C). However, when beneficiary plants’ cover increased, both SLA and LDMC significantly increased (Figure S8B, D).
Beneficiary plant cover and cushion flower production were not correlated in the CM1 population, weakly negatively correlated (P= 0.118) in the PJ4 population and strongly negatively correlated in the other populations (Figure 5A-H). These findings clearly indicate that beneficiary species can significantly reduce the cushion plant’s flower production. A significant negative correlation between beneficiary cover and fruit production was also detected in the PY population (Figure 5), but this relationship was neutral in the PJ3, CM1, CM2 and YL populations, and weakly negative in the others (PJ1, PJ2 and PJ4). Thus, beneficiary plants seem to generally have slightly negative effects on the cushion plant’s fruit production.
Moreover, beneficiary plant cover also positively correlated with the mortality of individual cushions in all study populations (Figure 5I-K), indicating that increases in this variable can promote the mortality of cushion plants and accelerate their exclusion.