Leaves are morphologically and functionally diverse organs fundamental to the survival of photosynthesizing plants. Botanists have long attempted to explore the biological significance of leaf traits and their environmental drivers. Yet, studies of large-scale variation in leaf traits across wide geographical ranges are limited due to the quality of available data. Here, leaf morphological trait data of 11,405 Chinese woody species from online databases, herbarium specimens (>6000 specimens of 7 species spanning ~100 years since collection), published floras, and unpublished data from experts were combined with species distribution maps to generate spatial patterns of leaf traits, with the aim of exploring their environmental drivers, and evaluating their potential to predict ecosystem productivity. Results show that spatiotemporal variation of leaf size and shape responds to both temperature and precipitation, two major environmental axes influenced by climate change. Further, leaf size was strongly linked with ecosystem primary productivity showing consistent patterns among species with different life-forms and evolutionary histories. The leaf size linkage with primary productivity may reflect mechanistically how natural selection optimizes leaf size to maximize photosynthesis through the physics of leaf energy balance and how climate limits the total leaf area of canopies in different environments. As both current and fossil leaf size is easy to measure, our findings identify key plant indicators scaling from individual adaptation to productivity of terrestrial ecosystems and provide an important methodological approach for studying plant structure to elucidate ecological patterns and future predictions using herbarium specimens housed in botanical institutions in situ plant collections.