Many hypotheses have been proposed to explain elevational species richness patterns; however, evaluating their importance remains a challenge, as mountains that are nested within different biogeographic regions have different environmental attributes. Here, we conducted a comparative study for trees, shrubs, herbs, and ferns along the same elevational gradient for 22 mountains worldwide, examining the performance of hypotheses of energy, tolerance, climatic variability, and spatial area to explain the elevational species richness patterns for each plant group. Results show that for trees and shrubs, energy-related factors exhibit greater explanatory power than other factors, whereas the factors that are associated with climatic variability performed better in explaining the elevational species richness patterns of herbs and ferns. For colder mountains, energy-related factors emerged as the main drivers of woody species diversity, whereas in hotter and wetter ecosystems, temperature and precipitation were the most important predictors of species richness along elevational gradients. For herbs and ferns, the variation in species richness was less than that of woody species. These findings provide important evidence concerning the generality of the energy theory for explaining the elevational species richness pattern of plants, highlighting that the underlying mechanisms may change among different growth form groups and regions within which mountains are nested.

Southwest China is an important biodiversity hotspot in the world and is controlled by the Pacific and Indian Ocean monsoon in the east and west part respectively. However, how abiotic and biotic factors affect the response of vegetation to climate change in different monsoon regions is still not clear. Here we used the annual change rate of growing-season normalized difference vegetation index (NDVI trend) during 1982-2015 to explore the vulnerability of vegetation (forests and shrubs) activity to climate change in southwest China. We examined NDVI trend in relation to: 1) climate change trends, i.e. annual change rate of water and energy availability, indicated by the Palmer Drought Index (PDSI) and potential evapotranspiration (PET), respectively; 2) climatic condition, i.e. mean PDSI and PET during 1982-2015; 3) vegetation height; 4) biome type; 5) monsoon region. The results showed that NDVI generally increased in the Pacific monsoon region, especially in the southern areas, probably because the vegetation under more productive climate were more resistant to climate change, and also because decreased temperature lead to lower evapotranspiration which alleviated the slight drought trend in this region. In contrast, NDVI generally decreased in the Indian Ocean monsoon region which showed more pronounced drought trend, especially in the tall subalpine and tropical forests of Southeast Tibetan Mountains, which supports the \"hydraulic limitation hypothesis\" that vegetation height interacted with climate change in affecting vegetation vulnerability. Our analysis highlighted the critical roles of different monsoon systems, climate condition and vegetation height in affecting ecosystem vulnerability. We suggest that the (sub)tropical forests in the Pacific monsoon region may have act as an important carbon sink during the past decades, while the tall forests in Southeast Tibetan mountains (a biodiversity center with high carbon stock) are highly vulnerable to climate change and should have priority in ecosystem protection.