Abstract
Leaf traits are important indicators of plants' adaptive strategy to environmental changes. It is an established fact that leaf traits are jointly regulated by climatic and edaphic factors besides genetic factors. However, the relative importance of these abiotic forces in determining the general patterns of herbaceous plant leaf traits across different climatic regions in China is far from clear. We collected 1,653 observations of 542 species of herbaceous plant leaf traits including leaf mass per area, leaf nitrogen, and leaf phosphorus from 316 sampling sites across four climatic regions. We found that the leaf mass per area in the arid region was apparently larger than the others, whereas the smallest mass-based leaf nitrogen and mass-based leaf phosphorus were found in the humid region. Increased growing season temperature and evapotranspiration consistently promoted a conservative growth strategy indicated by higher relative benefit of leaf mass per area, especially in the arid region. Solar radiation in growing season promoted an acquisitive growth strategy indicated by higher relative benefits of mass-based leaf nitrogen and phosphorus in the humid region, but opposite patterns were found in the arid region and semi-humid region. Of all the soil nutrients including soil organic matter, total nitrogen, total phosphorus, and available nitrogen, soil available nitrogen was the strongest predictor of relative benefits of leaf traits associated with a nutrient acquisitive strategy, except in the nutrient-rich semi-humid region. There was a relatively larger number of abiotic factors contributing to relative benefits of leaf traits in the arid and humid regions. We concluded that plant functionality could respond divergently to the same factor facing different habitat conditions. Moreover, the relative benefits of leaf traits tended to be more vulnerable to abiotic filtering in more stressful conditions. Our findings have important implications for understanding the context-dependency of plant functionality to environmental filtering and further improving the predictability of plant dynamics under global change.