Daniel Zuleta, Helene C. Muller-Landau, Alvaro Duque, Natali Caro, Dairon Cardenas, Nicolas Castaño, Juan Diego León-Peláez, Kenneth J. Feeley
This is a plain language summary of a Functional Ecology research article which can be found here.
Seen from above, the Amazon looks like a single homogeneous forest (Fig.). But down on the ground, variation in geology, soil fertility, climate and flooding frequency generate a mosaic of contrasting forest types at regional and local scales. In upland (terra firme) forests, erosion by streamlets determine different topographic structures, i.e. valleys and ridges, that exert a strong influence on the distribution of soil water and nutrients within what is generally considered a single forest type. These small-scale changes in topography are thought to influence tree species’ performance and distributions. In fact, recent studies in Amazon seasonal forests have demonstrated that tree species associated with valleys have a combination of traits that allow for more efficient water transport compared to species associated with ridges, which tend to have lower water availability. However, most studies have focused on very few individuals within particular taxa and have employed species’ mean trait values to assess these patterns, thereby ignoring the intraspecific variation that may allow species to maintain performance across different environmental conditions.
Using an unprecedented dataset including 1,077 trees of 72 dominant species, we quantified the interspecific (across species) and intraspecific (across trees within species) variation in 18 branch, leaf, and stomatal traits in relation to local variation in topography in an aseasonal Amazon terra firme forest: the 25-ha Amacayacu Forest Dynamics Plot. In contrast to seasonal forests, we found substantial trait variability that was not directly related to species’ topographic associations. Instead, most of the trait variability in this forest was exhibited at the intraspecific level and was associated with topographic elevation of trees for leaf traits, and with tree size for branch and stomatal traits. We also identified bivariate trait relationships at the tree-level (one point per tree) that generally mirrored relationships at the species-level (mean species value), but varied among trees within species, across habitat species’ association classes, and across tree size classes. Our results highlight the importance of intraspecific trait variability when testing trait-environment and bivariate relationships, and point to tree size as potential factor controlling trait variation in tropical forests.