Desert surface soils devoid of plant cover are populated by a variety of microorganisms, many with yet unresolved physiologies and lifestyles. Nevertheless, a common feature vital for these microorganisms inhabiting arid soils is their ability to survive long drought periods and reactivate rapidly in rare incidents of rain. Chemolithotrophic processes such as oxidation of atmospheric hydrogen and carbon monoxide are suggested to be a widespread energy source to support dormancy and resuscitation in desert soil microorganisms. Here, we assessed the distribution of chemolithotrophic, phototrophic, and desiccation-related metabolic potential among microbial populations in arid biological soil crusts (BSCs) from the Negev Desert, Israel, via population-resolved metagenomic analysis. While the potential to utilize light and atmospheric hydrogen as additional energy sources was widespread, carbon monoxide oxidation was less common than expected. The ability to utilize continuously available energy sources might decrease the dependency of mixotrophic populations on organic storage compounds and carbon provided by the BSC-founding cyanobacteria. Several populations from five different phyla besides the cyanobacteria encoded CO2 fixation potential, indicating further potential independence from photoautotrophs. However, we also found population genomes with a strictly heterotrophic genetic repertoire. The highly abundant Rubrobacteraceae (Actinobacteriota) genomes showed particular specialization for this extreme habitat, different from their closest cultured relatives. Besides the ability to use light and hydrogen as energy sources, they encoded extensive O2 stress protection and unique DNA repair potential. The uncovered differences in metabolic potential between individual, co-occurring microbial populations enable predictions of their ecological niches and generation of hypotheses on the dynamics and interactions among them.
IMPORTANCE This study represents a comprehensive community-wide genome-centered metagenome analysis of biological soil crust (BSC) communities in arid environments, providing insights into the distribution of genes encoding different energy generation mechanisms, as well as survival strategies, among populations in an arid soil ecosystem. It reveals the metabolic potential of several uncultured and previously unsequenced microbial genera, families, and orders, as well as differences in the metabolic potential between the most abundant BSC populations and their cultured relatives, highlighting once more the danger of inferring function on the basis of taxonomy. Assigning functional potential to individual populations allows for the generation of hypotheses on trophic interactions and activity patterns in arid soil microbial communities and represents the basis for future resuscitation and activity studies of the system, e.g., involving metatranscriptomics.