Abstract: Intraspecific trait variation (ITV) is a critical mechanism for species adaptation to heterogeneous environments, yet its relative contribution and variation patterns across different ecosystems remain poorly understood. This study investigated 28 dominant species and 910 individuals in (Xishuangbanna), subtropical evergreen broad-leaved forest (Ailaoshan), and savanna (Yuanjiang). We measured leaf thickness (LT), leaf area (LA), specific leaf area (SLA), leaf dry matter content (LDMC), and leaf chlorophyll content (LCC), combining phylogenetic signal testing and variation partitioning to quantify the effects of site, species, and individual scales on leaf trait variation. We detected significant phylogenetic signals among those leaf traits. Plants in the tropical rainforest exhibited resource-acquisitive strategies (high LA and SLA, but low LDMC), whereas species in the subtropical evergreen broad-leaved forest and savanna showed more resource-conservative strategies. ITV accounted for 9%–47% of the total trait variation at the regional scale. At local scale, the relative contribution of ITV to total community trait variance was highest in the resource-rich forest and negligible in the savanna. Our findings demonstrate that the variance of functional traits is highly context-dependent and scale-dependent across ecosystems. Given that intraspecific variation contributes substantially to total community-level trait variance, explicitly incorporating ITV into predictive models is critical for accurately shifting our understanding from static trait means to dynamic mechanisms of community assembly and ecosystem responses to global climate change.

Key words: climate, functional traits, intraspecific trait variation, ITV, common species, phylogenetic signal, variation partitioning