Microbial functional diversity and redundancy: moving forward

Abstract

Microbial functional ecology is expanding as we can now measure the traits of wild microbes that affect ecosystem functioning. Here, we review techniques and advances that could be the bedrock for a unified framework to study microbial functions. These include our newfound access to environmental microbial genomes, collections of microbial traits, but also our ability to study microbes' distribution and expression. We then explore the technical, ecological, and evolutionary processes that could explain environmental patterns of microbial functional diversity and redundancy. Next, we suggest reconciling microbiology with biodiversity-ecosystem functioning studies by experimentally testing the significance of microbial functional diversity and redundancy for the efficiency, resistance, and resilience of ecosystem processes. Such advances will aid in identifying state shifts and tipping points in microbiomes, enhancing our understanding of how and where will microbes guide Earth's biomes in the context of a changing planet.

Keywords: ecosystem functioning; functional redundancy; microbial functional ecology; resilience; resistance.

Figure 1.

Microbial functional redundancy varies with trait definitions. Trait accumulation curves based on 957 MAGs, from the surface ocean (Delmont et al. 2018). Gene prediction was performed with Prokka (Seemann 2014), and MMseqs2 (Steinegger and Söding 2017) was used to generate the MAGs genes catalog (two steps: 1/dereplication of the predicted ORFs across all MAGs, and 2/grouping the ORFs at 90% homology into a catalog. KEGG Orthologs present in each MAG were inferred with enrichM (https://github.com/geronimp/enrichM). METABOLIC traits (Zhou et al. 2022) represent the ability to perform broad-scale biogeochemical processes. Accumulation curves were built using the R package preseqR (Daley and Smith 2013). They represent the number of traits (Y-axis) retrieved when randomly sampling n MAGs (X-axis). The number of traits per MAG is represented in the violin plots on the right side of the accumulation curves.

Figure 2.

Redundancy varies across microbial functions. Bacterial and Archaeal phylogenetic trees of 957 MAGs, from the surface ocean (Delmont et al. 2018), and the presence of various metabolic steps (carbon fixation, methane, nitrogen, and sulfur metabolisms) in each genome. The taxonomy of each MAG was inferred with GTDB-Tk (Chaumeil et al. 2022). The bacterial and archaeal phylogenetic trees were built with FastTree (Price et al. 2010) using the alignment of marker genes of each MAG constructed during the computation of GTDB-Tk. FastTree was run using the generalized time-reversible model and branch lengths were rescaled with a Gamma20-based likelihood (see scale). The tips of the tree’s branches are colored by phyla (see color legend). We note mismatches between phylogeny and taxonomy due to low MAG completeness and/or contamination (Delmont et al. 2018). The functional annotation of each MAG was performed with the METABOLIC pipeline (Zhou et al. 2022). Each concentric line around the tree represents a KEGG module (an ensemble of KEGG orthologs required to perform a reaction). We focused on modules involved in carbon fixation (15 modules), methane (11), nitrogen (6), or sulfur cycling (3, see color code). As modules may require various sets of KOs for the function to be performed, module completeness was computed as the percentage of KOs (involved in a module) observed in each MAGs compared to the total number of KOs required per module.

Figure 3.

Effects of methods, scales, biology, and ecology on the functional redundancy of microbiomes. (A) Coarse trait resolution will result in higher redundancy because these traits can be shared by many species. (B) Large spatial and temporal scales will group species and ecotypes with similar effect traits that do not cooccur at finer resolutions, resulting in higher redundancy. In turn, a finer phylogenetic resolution will detect microdiversity and lead to the delineation of various taxa with similar effect traits, resulting in higher redundancy. (C) The effect of genome size on redundancy is yet unknown. Larger genomes should increase the number of traits harbored by species. Thus, it could increase the number of traits shared within a microbiome but also decrease genome similarity, with opposite effects on redundancy. (D) Higher species coexistence in environments with low competitive exclusion will likely increase the occurrence of taxa with similar traits, thus increasing redundancy. (E) HGTs (conjugation, transduction, vesiduction, or transformation, from left to right) are transfers of traits from one taxon to another, suggesting that the traits harbored by a taxon could vary in space and time, likely affecting the overall functional diversity and redundancy. (F) Symbiosis or cross-feeding allows the acquisition or production of new traits.