1. The dissimilarity and hierarchy of trait values that characterize niche and fitness differences, respectively, have been increasingly applied to infer mechanisms driving community assembly and to explain species co-occurrence patterns. Here, we predict that limiting similarity should result in the spatial segregation of functionally similar species, while functionally similar species will be more likely to co-occur either due to environmental filtering or competitive exclusion of inferior competitors (hereafter hierarchical competition).

2. We used a fully mapped 50-ha subtropical forest plot in southern China to explore how pairwise spatial associations between saplings and between adult trees were influenced by trait dissimilarity and hierarchy in order to gain insight into assembly mechanisms. We assessed pairwise spatial associations using two summary statistics of spatial point patterns at different spatial scales and compared the effects of trait dissimilarity and trait hierarchy of different functional traits on the interspecific spatial associations. These comparisons allow us to disentangle the effects of limiting similarity, environmental filtering and hierarchical competition on species co-occurrence.

3. We found that trait dissimilarity was generally negatively related with interspecific spatial associations for both saplings and adult trees across spatial scales, meaning that species with similar trait values were more likely to co-occur and thus supporting environmental filtering or hierarchical competition. We further found that trait hierarchy outweighed trait dissimilarity in structuring pairwise spatial associations, suggesting that hierarchical competition played a more important role in structuring our forest community than environmental filtering across life stages.

4. This study employed a novel method, by offering the integration of pairwise spatial association and trait dissimilarity as well as trait hierarchy, to disentangle the relative importance of multiple assembly mechanisms in structuring co-occurrence patterns, especially the mechanisms of environmental filtering and hierarchical competition, which lead to indistinguishable co-occurrence patterns. This study also reinforced the importance of trait hierarchy rather than trait dissimilarity in driving neighborhood competition.

Adult (DBH > 10 cm) and saplings (1 cm < DBH <3 cm) were measured, identified and mapped in a 50 ha plot established in 2013.

Data on these traits (leaf area (LA; cm²), specific leaf area (SLA; cm² g^-1, calculated as leaf area/dry mass), leaf dry matter content (LDMC; g g^-1, calculated as leaf dry mass/fresh mass), wood density (WD; g cm^-3, calculated as trunk wood dry mass/fresh volume), wood dry matter content (WDMC; g g^-1, calculated as dry wood mass/fresh wood mass) and tree maximum height (H_max; m) ) on species-level were collected and measured from the HSD plot, in which H_max was estimated by averaging the top 1% tallest trees for each species in the plot.