Leaf morphological traits (LMTs) can directly influence carbon-uptake and water-loss of plants and are therefore sensitive indicators of plant adaptation to climate. Quantifying the patterns in LMTs across large geographical scale and the underlying mechanisms are fundamental for understanding plants’ responses to climate and for paleo-climate reconstruction. Recent studies find that LMTs-climate relationships vary among different life-forms and are affected by evolutionary history, which remains controversial. Here, using the dataset of leaf morphological traits (leaf margin states, leaf length, leaf width, and length-width product/ratio) and geographical distributions of 10,480 woody dicot species in China, we evaluated the phylogenetic signals of LMTs, and demonstrated the geographical patterns in LMTs and their relationships with climate across different life-forms (evergreen and deciduous; trees, shrubs and lianas) and species with different family-ages. Furthermore, we tested those findings using data on floras of North America.
Results/Conclusions
Significant phylogenetic signals were found for LMTs across families. Both the percentage of untoothed species and grid-mean leaf size for all Chinese woody dicots decreases with latitude, especially in eastern China. The grid-mean leaf size and shape were all sensitive to precipitation and temperature, but whether the relationships were influenced by plant life-form or evolutionary history depended on specific morphological traits. Our findings revealed great uncertainties in leaf margin-MAT relationship induced by life-form, precipitation and evolutionary history, and suggested that the widely-used method (leaf margin analysis, LMA) should be applied cautiously when reconstructing paleo-temperature. In contrast, leaf size variation was highly correlated with actual annual evapotranspiration and satellite-derived primary productivity, and the predictive power of leaf size for primary productivity was not influenced by plant life-form or family ages, even higher than leaf area index. Besides, the transfer functions of primary productivity‒leaf size based on Chinese data well predicted ecosystem primary productivity in North American, which further indicate that leaf size is a good proxy for reconstructing paleo-primary productivity of terrestrial ecosystems.