Kemanian, Armen R. (Washington State University, Biological Systems Engineering Dep., Smith Hall, Pullman, WA, 99164-6120; Phone: 509-335-3661; Fax: 509-335-2722; Email: armen@wsunix.wsu.edu)

Modeling Carbon and Nitrogen Cycling and Greenhouse Gas Emissions from Agricultural Systems

C.O. Stockle, A.R. Kemanian *, D.R. Huggins, H.P. Collins

A model of carbon and nitrogen cycling (CNC) was developed and integrated into the CropSyst cropping systems model. The main reason for developing our own model was to build upon the strength of CropSyst in simulating the soil water balance and temperature, and to take advantage of the already validated robustness of CropSyst in simulating crop biomass production and therefore the input of C to the soil under a variety of climates and management scenarios. The criteria to develop the CNC model were functionality and that the definition of pools and transfer rates among pools were measurable or inferable from field or laboratory measurements. The structure of the soil carbon subroutine has a resemblance with the model presented by Verberne et al. (1990), but includes several modifications. The CNC model runs on an hourly time step. The soil is divided in a user-defined number of layers of varying thickness. Inputs of carbon are from crop residues, senescent leaves, and dead roots. No explicit allowance was made for root C deposition, but in practice it can be integrated into the root turnover. Each time there is an input of residue or dead roots in a given soil layer, a new carbon residue pool is generated. Similarly, tillage events result in new incorporated pools. Carbon inputs such as crop residues, roots, and manure are divided in three fractions: lignin, structural fraction, and non-structural fraction, the latter having the smallest turnover rate. Soil carbon is divided in four pools of decreasing turnover rates: microbial biomass, labile, metastable, and stable pools. Carbon decomposition from the residue and the soil C pools is mediated by the biomass pool, except for lignin. Each microbial attack renders carbon dioxide and microbial biomass. The products of microbial decomposition are allocated to the different soil C pools following rules dependent on the clay content, similar to those presented by Parton et al. (1994). The microbial biomass pool plays a functional role and no separation between active and dormant microbial biomass was attempted. All decomposition processes are simulated using first order kinetics. Rates of microbial-mediated processes are adjusted by temperature and moisture factors operating multiplicatively. Soil temperature is modeled following an energy balance approach, and soil moisture using a finite difference method. Nitrogen cycling is linked to the flow of carbon, except for the ammonium and nitrate pools that are independently modeled. Total denitrification and the fraction of N lost as nitrous oxide are modeled as a function of nitrate concentration, soil respiration rate, and soil moisture. Tillage effects on the decomposition rate are incorporated into the model. Each tillage event determines a transfer of carbon and nitrogen from the stable and metastable pools to the labile pool. The transferred carbon slowly returns to the metastable pool unless there are more tillage events. These transfers represent in a simple manner aggregate break-up due to mechanical disturbance and the subsequent aggregation. Parameterization of the CNC model was performed by interpretation of information in the literature and by interpretation of our own data on soil carbon dynamics. Simulations for different cropping systems in the Pacific Northwest and a sensitivity analysis of some of the model parameters are presented.