Why are some microbes more ubiquitous than others? Predicting the habitat breadth of soil bacteria

TLDR

Identifying traits that shape spatial distributions is difficult for bacteria due to limited phenotypic data, but genomic information offers a way to infer ecological attributes. The study aimed to assess whether phylogeny and genomic traits could explain differences in habitat breadth of soil bacteria. The authors mapped the spatial distributions of ~124,000 soil bacterial taxa across a 3.41 km² area and analyzed phylogenetic and genomic data to predict habitat breadth. The analysis revealed that bacterial occupancy strongly correlated with environmental range, and that genomic traits—especially larger genomes and greater metabolic versatility—were more effective than phylogeny alone at explaining habitat breadth, demonstrating that genomic data can predict microbial biogeographic patterns.

Abstract

Abstract Identifying the traits that determine spatial distributions can be challenging when studying organisms, like bacteria, for which phenotypic information is limited or non‐existent. However, genomic data provide another means to infer traits and determine the ecological attributes that account for differences in distributions. We determined the spatial distributions of ~124 000 soil bacterial taxa across a 3.41 km 2 area to determine whether we could use phylogeny and/or genomic traits to explain differences in habitat breadth. We found that occupancy was strongly correlated with environmental range; taxa that were more ubiquitous were found across a broader range of soil conditions. Across the ~500 taxa for which genomic information was available, genomic traits were more useful than phylogeny alone in explaining the variation in habitat breadth; bacteria with larger genomes and more metabolic versatility were more likely to have larger environmental and geographical distributions. Just as trait‐based approaches have proven to be so useful for understanding the distributions of animals and plants, we demonstrate that we can use genomic information to infer microbial traits that are difficult to measure directly and build trait‐based predictions of the biogeographical patterns exhibited by microbes.