1. Disentangling ecological processes that influence community assembly and species diversity across spatial scales remains a major goal of community ecology. Community assembly processes influence spatial patterns of species diversity through their interactions with key functional traits. Hence, quantifying spatial patterns of functional trait diversity (FD) represents a useful tool for disentangling the relative importance of abiotic filtering, biotic interactions, random assembly, and dispersal limitation across spatial scales. 2. Here we measured 12 traits of 112 study species in a 20-ha fully-mapped subtropical forest plot. The individuals of the 112 study species account for 99% of all living stems with diameter at breast height (DBH) ≥ 1 cm. We studied important functional traits related to physiological processes of plants including resource acquisition (e.g., CO2 assimilation rate and leaf nutrient concentration) and drought tolerance (e.g., stem hydraulic conductivity and leaf turgor loss point) are measured. Additionally, species abundance, spatial locations (x- and y- coordinate for each individual of the 112 study species), as well as topographic and soil variables that represent potentially important attributes of the physical environment of the plot, were also included in our dataset. 3. We employed two FD-based tests (comparing FD within communities to those from random communities, distance- based Moran’s eigenvector maps (MEM) and redundancy analysis based variance partitioning), and one spatial analysis (inhomogeneous bivariate pair correlation analysis) to quantify the spatial patterns of FD of the plot at multiple spatial scales (400, 900, 1600, 2500, and 10000 m2). 4. We demonstrate that abiotic filtering is the major determinant responsible for trait convergence at relatively small scales (400, 900 and1600 m2), whereas dispersal limitation becomes dominant, causing the weakening of trait convergence at relatively large scales (2500 and 10000 m2). 5. Our results highlight the relative contributions of different ecological processes to community assembly at different spatial scales, which can be distinguished using the diversity patterns of key functional traits. Also, our integrated approaches constitute a useful study design to disentangle variable ecological processes in shaping community assembly across spatial scales.