Abstract: Trait-based ecology seeks to characterize plant functional responses to broad environmental gradients, but substantial environmental heterogeneity can occur around single organisms. An individual root system can encounter vastly different soil conditions, suggesting that local root trait variation may facilitate resource acquisition. Recent temperate studies have identified substantial morphological root trait variation at small spatial (centimeters) and biological scales (sub-individual). However, the driving forces behind this variation and the consequences for root functioning are unclear, and evidence from tropical systems is scarce.
To identify relevant spatial scales of morphological and physiological root trait variation and their edaphic drivers, we employed a spatially explicit, nested sampling scheme with paired soil sampling on a widely distributed tropical tree, Handroanthus ochraceus. We sampled three fine root samples each from ten trees at three sites (N = 90) along a fertility gradient in the seasonally dry tropical forests of northwestern Costa Rica. Root-adjacent soil cores were collected and pooled to the tree level (N = 30). Measured root traits included respiration, phosphomonoesterase activity, diameter, tissue density, specific root length, specific root area, mycorrhizal colonization, and chemistry. Soil variables included moisture, Olsen-P, ammonium-N, nitrate-N, %C, %N, Ca, Mg, K, bulk density, and pH.
Despite large differences in fertility across sites, most root traits were most variable between roots of an individual tree, followed by trees in a site. Phosphomonoesterase activity was the only variable that largely varied between sites and was moderately positively correlated with Olsen-P (r = 0.37, p < 0.001). Random effects variance partitioning demonstrated that soil variables varied most at the between-site scale, except for nitrate-N and ammonium-N which were most variable between trees in a site. A PCA of soil variables determined Ca, Olsen-P, nitrate-N, and pH opposed %N; the second component opposed C% and N% against bulk density and ammonium-N (62% variance explained). Across root samples, specific root length was strongly positively correlated with respiration (Pearson r = 0.72, p < 0.001). For the root trait PCA, the first component opposed diameter against specific root length, specific root area, and respiration; the second component opposed tissue density and diameter; phosphomonoesterase activity strongly contributed to a third component (92.8% variance explained).
These preliminary results suggest that tropical trees can adjust root functioning at the sub-individual level, and that relevant adjustments may align with spatial scales of environmental heterogeneity. Trees may thus fine-tune their root functioning to coordinate the trade-offs of resource acquisition.