Ocean biogeochemistry modeled with emergent trait-based genomics

V. J. Coles, University of Maryland Center for Environmental Science
M. R. Stukel, University of North Florida
M. T. Brooks, University of Maryland Center for Environmental Science
A. Burd, University of Georgia
B. C. Crump, Oregon State University
M. A. Moran, University of Georgia
J. H. Paul, University of South Florida St. Petersburg
B. M. Satinsky, University of Georgia
P. L. Yager, University of Georgia
B. L. Zielinski, Oregon State University
R. R. Hood, University of Maryland Center for Environmental Science


Marine ecosystem models have advanced to incorporate metabolic pathways discovered with genomic sequencing, but direct comparisons between models and “omics” data are lacking. We developed a model that directly simulates metagenomes and metatranscriptomes for comparison with observations. Model microbes were randomly assigned genes for specialized functions, and communities of 68 species were simulated in the Atlantic Ocean. Unfit organisms were replaced, and the model self-organized to develop community genomes and transcriptomes. Emergent communities from simulations that were initialized with different cohorts of randomly generated microbes all produced realistic vertical and horizontal ocean nutrient, genome, and transcriptome gradients. Thus, the library of gene functions available to the community, rather than the distribution of functions among specific organisms, drove community assembly and biogeochemical gradients in the model ocean.