Spatial Regulation of Trait-Environment Interactions Shapes Plant Community Functional Patterns in a Tropical Forest.

Question: Inferring community assembly mechanisms requires an understanding of the responses of plant communities to environmental gradients and spatial processes. In a tropical forest, we asked: (1) What environmental gradients are associated with variation in community functional composition and diversity? (2) How does explicitly accounting for spatial structure influence inference about these environmental effects? and (3) To what extent can the observed trait-environment relationships be explained by environmentally structured species turnover (i.e., changes in species composition along environmental gradients)? Methods: We applied a spatially explicit hierarchical Bayesian modeling (INLA-SPDE) framework to data from 135,000 woody plants, measuring eight functional traits and 24 environmental variables covering topography, soil nutrients, and canopy structure. The model estimates environmental effects while accounting for spatial autocorrelation in trait distributions. We further used null model tests to evaluate question (3). Results: Spatially explicit models identified topographic water availability (TWI) as the main environmental gradient associated with community-level trait patterns, aligning functional composition along the trade-off between acquisitive and conservative resource-use strategies. Inferences for several environmental variables changed substantially after accounting for spatial auto-correlation. Null model analyses indicated that most trait-environment relationships in the Mo Singto forest plot were consistent with environmentally structured species turnover, with only limited evidence for additional trait-environment associations, as shown by tree height. Conclusions: Our results indicate that when traits are represented at the species level, smaller-scale community-weighted mean (CWM) patterns are primarily explained by species replacement along environmental gradients. Our study also demonstrates that high-resolution, fully stem-mapped forest plots provide strong power to detect environmental signals and reveal spatially structured patterns in community functional composition. By combining spatially explicit models with null model inference, we show that robust interpretation of CWM trait-environment relationships requires jointly accounting for spatial structure and species turnover, which provides a clear framework for understanding how environmental gradients shape community functional patterns.