Microbial Functional Genes Improve Development and Validation of Ecosystem Model

Knowledge of microbial mechanisms is critical to understand Earth's biogeochemical cycles under climate and environmental changes. However, large uncertainties remain in model simulations and predictions due to the lack of explicit parameterization of microbial processes. In addition, it remains challenging how microbially-enabled models can be linked with the massive, rich taxonomic/phylogenetic and functional data for microbial communities that are being generating from new sequencing technologies. We have developed the coupled Carbon-Nitrogen (C-N) Microbial-ENzyme Decomposition (MEND) model to represent soil organic matter (SOM) decomposition as well as N mineralization-immobilization by microbes. Here we further improved our MEND model by explicitly representing (i) multiple enzyme functional groups that catalyze nitrification, stepwise denitrification, and nitrogen fixation processes; (ii) plant-microbial competition for mineral N; and (iii) nitrate leaching and N gas emission from soil to the atmosphere. We tested the new MEND model with data from one of the few ecologically realistic, well-replicated long-term (12 years) CO₂×N experiments, BioCON, a grassland experiment in Central Minnesota, USA, which helped elucidate how microbe-mediated feedbacks underlying the long-term impact of elevated CO₂ (eCO₂, +180 ppm) stimulation on soil respiration are affected by N addition (+4 g N m^-2 yr^-1). Our model calibration and validation showed good agreement between simulated and long-term observations of soil CO₂ fluxes and soil concentrations of ammonium and nitrate. Compared to empirical data, our modeling results also showed reasonable rates of net N mineralization, nitrification, N fixation, and plant N uptake. In addition, the new MEND model with multiple functional enzymes indicated significantly higher positive effects of eCO₂ on the C-N enzyme concentrations under ambient N than under enriched N treatment, which were consistent with the GeoChip-detected responses in functional gene abundances. Our study provides strong evidence of microbial control over soil C-N processes. We also advocate the use of microbial functional gene information to improve development and validation of ecosystem models for better understanding soil C and nutrient cycling in a changing world.