摘要：
Aims Understanding the drivers for leaf traits is critical to improving our predictions on ecosystems' responses to global changes. Geographic patterns of leaf traits are shaped by phylogenetic, biological and environmental factors simultaneously. However, till now few studies have examined how these factors influenced leaf traits together, and how their effects differed at the within- and among-site levels.
Methods We sampled leaf traits from a 1100 km shrub-biome transect across central Inner-Mongolia, including leaf mass per area (LMA), mass-based photosynthetic rate, nitrogen (N) and phosphorus (P) concentrations. We examined the effects of phylogenetic, biological (height and growth rate) and environmental (climate and soil) factors on leaf traits with mixed-model analyses of variance. Variation partitioning method was used to separate the joint and independent effects of these three types of factors.
Important findings (i) Climate and soil fertility (total or available nutrient concentrations) together explained 11.4–41.4% of among-site variations, with remarkable difference among traits. (ii) Height and height growth rate together explained 0.4–31.9% of trait variations (mostly among-site variations). Our results could only weakly support the ability of leaf traits as predictors for whole-plant growth. (iii) LMA was negatively related to height, which was consistent with the resource-use strategy hypothesis but inconsistent with the hypotheses proposed for coexisting trees, suggesting that the LMA–height relationship is shaped by rather different mechanisms between the within- and among-communities levels. (iv) The variation partitioning analysis showed that, the relationships between leaf traits and biological characters largely reflected the differences in both leaf traits and biological characters among species that occupying different sites. The relative importance of phylogenetic, biological and environmental factors differed remarkably among leaf traits, between the within- and among-communities levels, and between different biomes. (v) Our results strongly suggest the necessity of examining the three types of factors simultaneously, and at both the within- and among-communities levels, for a better understanding of the drivers for leaf traits patterns.
