Transient Microsite Models of Denitrification: III. Comparison of Experimental and Model Results

Abstract

AbstractExperiments were designed to investigate the validity of proposed reaction‐diffusion models of denitrification. Brass respirometers were used to measure fluxes of oxygen (O2), carbon dioxide (CO2), and nitrous oxide (N2O) from a 5 × 10−3‐m layer of saturated soil incubated at 25°C. Gaseous fluxes were measured as a function of time after wetting, with and without 0.42 mol m−3 of acetylene (C2H2) present in the headspace. The presence of C2H2 in the headspace doubled O2 fluxes 35 to 60 h after incubations commenced. When C2H2 was not present, N2O evolution was negligible after 25 h of incubation. Nitrous oxide fluxes predicted by the models were compared to those observed experimentally. When C2H2 was present, the models predicted N2O fluxes reasonably well (r2 = 0.90 and 0.78) for initial nitrate (NO−3) N levels of 7.1 and 15.9 mg N kg−1 soil, respectively. When C2H2 was not present, the models predicted the experimentally observed initial flux of N2O if an 8 to 10 h lag in N2O reductase activity was included in the models. The best model for describing N2O fluxes in these experiments used differential rates of Michaelis‐Menten reduction without NO−3 and nitrite (NO−2) inhibition of N2O reduction. Experimental techniques and parameter characterization, however, need considerable refinement. Independent measurements are needed to quantify relationships between reaction order and kinetic parameters. Still, given the general agreement between observed and predicted fluxes, there can be little doubt that the reaction‐diffusion models are essential to a better understanding of denitrification.