Simulating microbial denitrification with EPIC: Model description and evaluation

Journal Article · Fri Sep 01 00:00:00 EDT 2017 · Ecological Modelling

DOI:https://doi.org/10.1016/j.ecolmodel.2017.06.007· OSTI ID:1686163

Izaurralde, R. César; McGill, William B.; Williams, Jimmy R.; Jones, Curtis D.; Link, Robert P.; Manowitz, David H.; Schwab, D. Elisabeth; Zhang, Xuesong; Robertson, G. Philip; Millar, Neville

Here, microbial denitrification occurs in anaerobic soil microsites and aquatic environments leading to production of N₂O and N₂ gases, which eventually escape to the atmosphere. Atmospheric concentrations of N₂O have been on the rise since the beginning of the industrial revolution due to large-scale manipulations of the N cycle in managed ecosystems, especially the use of synthetic nitrogenous fertilizer. Here we document and test a microbial denitrification model identified as IMWJ and implemented as a submodel in the EPIC terrestrial ecosystem model. The IMWJ model is resolved on an hourly time step using the concept that C oxidation releases electrons that drive a demand for electron acceptors such as O₂ and oxides of N (NO₃^-, NO₂^-, and N₂O). A spherical diffusion approach is used to describe O₂ transport to microbial surfaces while a cylindrical diffusion method is employed to depict O₂ transport to root surfaces. Oxygen uptake by microbes and roots is described with Michaelis-Menten kinetic equations. If insufficient O₂ is present to accept all electrons generated, the deficit for electron acceptors may be met by oxides of nitrogen, if available. The movement of O₂, CO₂ and N₂O through the soil profile is modeled using the gas transport equation solved on hourly or sub-hourly time steps. Bubbling equations also move N₂O and N₂ through the liquid phase to the soil surface under highly anaerobic conditions. We used results from a 2-yr field experiment conducted in 2007 and 2008 at a field site in southwest Michigan to test the ability of EPIC, with the IMWJ option, to capture the non-linear response of N₂O fluxes as a function of increasing rates of N application to maize. Nitrous oxide flux, soil inorganic N, and ancillary data from 2007 were used for EPIC calibration while 2008 data were used for independent model validation. Overall, EPIC reproduced well the timing and magnitude of N₂O fluxes and NO₃^- mass in surficial soil layers after N fertilization. Although similar in magnitude, daily and cumulative simulated N₂O fluxes followed a linear trend instead of the observed exponential trend. Further model testing of EPIC+IMWJ, alone or in ensembles with other models, using data from comprehensive experiments will be essential to discover areas of model improvement and increase the accuracy of N₂O predictions under a wide range of environmental conditions.

• Research Organization: Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
• Sponsoring Organization: USDOE
• Grant/Contract Number: AC05-76RL01830
• OSTI ID: 1686163
• Alternate ID(s): OSTI ID: 1393740
• Report Number(s): PNNL-SA-126008; S0304380017303265; PII: S0304380017303265
• Journal Information: Ecological Modelling, Journal Name: Ecological Modelling Vol. 359 Journal Issue: C; ISSN 0304-3800
• Publisher: Elsevier
• Country of Publication: Netherlands
• Language: English

Related Subjects

54 ENVIRONMENTAL SCIENCES
microbial respiration
nitrous oxide
gas transport equation
fertilizer nitrogen
Michaelis-Menten kinetics
Environmental Policy Integrated Climate