Relationships between spatial pattern and absolute trait distances
By fitting the linear mixed regression model (2) using absolute trait distances of individual traits exclusively, we found statistical support for negative effects of the absolute trait distances measured by individual traits on pairwise spatial associations assessed by the two summary statistics (gij (r ) and Dij (r )) across different spatial scales (Fig. 2). The coefficients of absolute trait distances of individual traits of LA, WD and Hmax were consistently negative for both summary statistics across spatial scales. SLA (atr =5 m) and WDMC showed slightly negative effects on pairwise spatial associations, while LDMC did not show significant effects. As for absolute trait distances based on multiple traits, negative effects on spatial associations were also observed consistently across summary statistics and spatial scales (Fig. S1).
For the 80 focal species, we found that the effects of absolute trait distances were non-significant for most of the focal species on their pairwise spatial associations as the 95% confidence intervals of coefficients intercepted 0 (Fig. 2). While for the remaining focal species, in general, more focal species’ spatial associations were negatively correlated to absolute trait distances than those were positively correlated (Fig. 2).