AbstractWhile metagenomics can provide insight into microbial community metabolic potential, understanding factors that influence gene abundance is necessary to maximize the information gained from this analysis. Gene abundances are influenced by chemical or physical conditions along with other factors, such as copy number variation between taxa, methodological biases, or issues associated with identification and classification. Here, we identify major drivers of spatiotemporal shifts in microbial gene relative abundance from multiple months, sites, and depths within Chesapeake Bay in 2017 using shotgun metagenomics. We compared changes in relative abundance of key genes for bacterial photosynthesis, nitrogen, and sulfur metabolism with each other and measured environmental variables. Major drivers of differences in key metabolic gene abundances are associated with environmental variables that largely change with depth and season (e.g., temperature, oxygen, phosphate). For sulfur oxidation, bacterial photosynthesis, and denitrification, genes within each process are generally significantly correlated with each other and with several environmental variables. For other processes, such as nitrification, nitrogen fixation, and dissimilatory nitrate reduction to ammonium, genes that encode enzymes within the same pathway are not well correlated. The lack of correlation typically results from differences in identified taxa carrying these genes, suggesting modular pathway structure, methodological errors, or discrepancies in gene copy number between taxonomic groups. To be suitable indicators of biogeochemical processes for models, genes or pathways should be strongly correlated with environmental variables and specific to and inclusive of all taxa mediating the associated process.
Read full abstract