Tue, Aug 16, 2022: 8:15 AM-8:30 AM
513D
Background/Question/MethodsSoil carbon (C) and nitrogen (N) cycles and their complex responses to environmental changes have received increasing attention. However, large uncertainties in model predictions remain, partially due to the lack of explicit representation and parameterization of multiple differential microbial groups, particularly related to the inorganic N cycle. This impedes a comprehensive validation of complex C-N processes and their interactions as have been done for classical terrestrial C-N coupled models. Therefore, the introduction of mechanistic inorganic N cycling into microbial ecological models may provide new opportunities to pose and validate further hypotheses about coupled C-N cycling in response to environmental perturbations, especially elevated atmospheric CO2 concentration (eCO2) and enhanced N deposition. Here, using soil enzymes as indicators of soil function, we developed a competitive dynamic enzyme allocation scheme and detailed enzyme-mediated soil inorganic N processes in the Microbial-ENzyme Decomposition (MEND) model. We conducted a rigorous calibration and validation of MEND with diverse soil C-N fluxes, microbial C:N ratios, and functional gene abundances from a 12-year CO2×N grassland experiment (BioCON) in Minnesota, USA. We focused on four CO2×N treatments in BioCON: ambient atmospheric CO2 & ambient N supply (aCO2-aN), eCO2-aN, aCO2 & enriched N supply (aCO2-eN), and eCO2-eN.
Results/ConclusionsOur model calibration with aCO2-aN data achieved good agreement between simulated and observed soil respiration (coefficient of determination (R2) = 0.60), so did the model validation of soil respiration in the other three treatments (R2 = 0.56–0.61). The simulated mean soil NH4+ and (NO3–+NO2–) concentrations also agreed well with the observations. Modeled biological N fixation and plant N uptake rates were generally in accordance with literature-reported data. In addition, the model correctly predicted microbial C:N ratios and their negative response to enriched N supply. Model validation further showed that, compared to the changes in simulated enzyme concentrations and decomposition rates, the changes in simulated activities of eight C-N associated enzymes were better explained by the measured gene abundances in responses to elevated atmospheric CO2 concentration. Our results demonstrated that using enzymes as indicators of soil function and validating model predictions with functional gene abundances in ecosystem modeling can provide a basis for testing hypotheses about microbially-mediated biogeochemical processes in response to environmental changes. Further development and applications of the modeling framework presented here will enable microbial ecologists to address ecosystem-level questions beyond empirical observations, toward more predictive understanding, an ultimate goal of microbial ecology.
Results/ConclusionsOur model calibration with aCO2-aN data achieved good agreement between simulated and observed soil respiration (coefficient of determination (R2) = 0.60), so did the model validation of soil respiration in the other three treatments (R2 = 0.56–0.61). The simulated mean soil NH4+ and (NO3–+NO2–) concentrations also agreed well with the observations. Modeled biological N fixation and plant N uptake rates were generally in accordance with literature-reported data. In addition, the model correctly predicted microbial C:N ratios and their negative response to enriched N supply. Model validation further showed that, compared to the changes in simulated enzyme concentrations and decomposition rates, the changes in simulated activities of eight C-N associated enzymes were better explained by the measured gene abundances in responses to elevated atmospheric CO2 concentration. Our results demonstrated that using enzymes as indicators of soil function and validating model predictions with functional gene abundances in ecosystem modeling can provide a basis for testing hypotheses about microbially-mediated biogeochemical processes in response to environmental changes. Further development and applications of the modeling framework presented here will enable microbial ecologists to address ecosystem-level questions beyond empirical observations, toward more predictive understanding, an ultimate goal of microbial ecology.