Trait-trait relationships and tradeoffs vary with genome size in prokaryotes

Abstract

We report genomic traits that have been associated with the life history of prokaryotes and highlight conflicting findings concerning earlier observed trait correlations and tradeoffs. In order to address possible explanations for these contradictions we examined trait–trait variations of 11 genomic traits from ~18,000 sequenced genomes. The studied trait–trait variations suggested: (i) the predominance of two resistance and resilience-related orthogonal axes and (ii) at least in free living species with large effective population sizes whose evolution is little affected by genetic drift an overlap between a resilience axis and an oligotrophic-copiotrophic axis. These findings imply that resistance associated traits of prokaryotes are globally decoupled from resilience related traits and in the case of free-living communities also from traits associated with resource availability. However, further inspection of pairwise scatterplots showed that resistance and resilience traits tended to be positively related for genomes up to roughly five million base pairs and negatively for larger genomes. Genome size distributions differ across habitats and our findings therefore point to habitat dependent tradeoffs between resistance and resilience. This in turn may preclude a globally consistent assignment of prokaryote genomic traits to the competitor - stress-tolerator - ruderal (CSR) schema that sorts species depending on their location along disturbance and productivity gradients into three ecological strategies and may serve as an explanation for conflicting findings from earlier studies. All reviewed genomic traits featured significant phylogenetic signals and we propose that our trait table can be applied to extrapolate genomic traits from taxonomic marker genes. This will enable to empirically evaluate the assembly of these genomic traits in prokaryotic communities from different habitats and under different productivity and disturbance scenarios as predicted via the resistance-resilience framework formulated here.