Community-wide patterns in pollen and ovule production, their ratio (P/O), and other floral traits along an elevation gradient in southwestern China.

BACKGROUND: As the male and female gametophytes of flowering plants, pollen and ovules largely determine the upper and lower boundaries of plant reproductive success. It is commonly predicted that pollen and ovule number per flower should increase, and pollen-ovule ratio (P/O) per flower should decrease with increasing elevation in response to a more stochastic pollination environment. Here, we aimed to determine the response of pollen number, ovule number, and P/O to other floral traits and elevation gradients for 84 insect-pollinated herbaceous flowering plant species in five sub-alpine and alpine communities (2709 to 3896 m a.s.l.) on Yulong Snow Mountain, southwestern China. RESULTS: Six floral traits, including P/O, floral display area, flower number, tube depth, flower shape, and pollen presentation, were highly correlated with pollen and ovule number per flower. With increasing elevation, pollen number and P/O per flower increased marginally and significantly, respectively; ovule number per individual, flower number per individual, stigma stamen separation, and inflorescence height decreased significantly. However, ovule number per flower and other floral traits (i.e., floral display area, tube depth, stigma height, stamen height, and pollen and P/O per individual) did not change with elevation. We detected significant phylogenetic signals for pollen number, ovule number, and P/O, suggesting that these traits may be highly conserved and with limited response to changing environmental conditions. CONCLUSIONS: Results revealed patterns of plant reproductive character evolution along elevation gradients and the potential factors governing their spatial variation in high-elevation environments. Plant species at high elevations are more likely adapted to cross-pollination, indicated by increased P/O per flower at high elevations on Yulong Mountain. Combined effects of phylogenetic history and plant-pollinator interactions should determine plant trait evolution.

Flowering phenology differs among wet and dry sub-alpine meadows in southwestern China.

The effect of floral traits, floral rewards and plant water availability on plant-pollinator interactions are well-documented; however, empirical evidence of their impact on flowering phenology in high-elevation meadows remains scarce. In this study, we assessed three levels of flowering phenology, i.e. population-, individual- and flower-level (floral longevity), in two nearby but contrasting (wet versus dry) sub-alpine meadows on Yulong Snow Mountain, southwestern China. We also measured a series of floral traits (pollen number, ovule number, and the ratio of pollen to ovule number per flower, i.e. pollen:ovule ratio [P/O]) and floral rewards (nectar availability and pollen presentation) as plausible additional sources of variation for each phenological level. Floral longevity in the wet meadow was significantly longer than that for the dry meadow, whereas population- and individual-flowering duration were significantly shorter. Our results showed a significant positive relationship between flowering phenology with pollen number and P/O per flower; there was no relationship with ovule number per flower. Further, we found a significant effect of flowering phenology on nectar availability and pollen presentation. Our findings suggest that shorter floral longevity in dry habitats compared to wet might be due to water-dependent maintenance costs of flowers, where the population- and individual-level flowering phenology may be less affected by habitats. Our study shows how different levels of flowering phenology underscore the plausible effects of contrasting habitats on reproductive success.