Abstract
Background
To improve our understanding of microbial systems, it is essential to refine the conceptual frameworks that connect microorganisms to their ecological functions. While trait-based approaches can provide nuanced perspectives on how microorganisms influence ecosystem processes, there is ongoing debate over the link between microbial taxonomic classifications and life history traits. Here, we integrate genomic, metagenomic, amplicon sequencing, and experimental (stable isotope probing) data to investigate the scaling of bacterial growth traits from individual taxa to complex assemblages and to identify specific taxonomic groups of soil bacteria that can be used as indicators of community-scale microbial growth.
Results
Our results revealed broadly different distributions of growth rates among bacterial phyla, including significantly different mean and median rates. This, in turn, manifested in strong relationships between relative abundances of some phyla and community-scale growth rates in soil. Specifically, we calculated community weighted mean growth rates using measured growth rates of constituent taxa and found that the fast-growing taxa that had sufficient abundance and ubiquity across samples to contribute to variation in community-average growth were mostly lineages of Proteobacteria (e.g., Sphingomonas). As a result, the relative abundance of phylum Proteobacteria was the single strongest taxonomic predictor of community-average growth, explaining up to ~ 60% of the variation in growth rates across communities. In contrast, Verrucomicrobia were consistent indicators of slower community-average growth. These patterns were especially strong when using taxon-level growth rates measured following carbon and nitrogen additions to soil.
Conclusions
Our results demonstrate that phylum relative abundances can be strong indicators of community-level bacterial growth despite the wide variation in growth rates observed within phyla. The stronger phylum-growth relationships for whole assemblages than are apparent for individual taxa are due to relative abundance-weighted trait averaging in complex assemblages, i.e., at the community scale, broad differences in growth traits among phyla become more important than variation within phyla. Overall, our results provide clarity regarding the use of bacterial taxonomic information for inferring traits, demonstrating that high taxonomic ranks can be valid indicators of microbial traits in soil provided that inferences are drawn at the appropriate scale.