Simulation of C and N mineralisation during crop residue decomposition: A simple dynamic model based on the C:N ratio of the residues
Résumé
C and N mineralisation kinetics obtained in laboratory incubations during decomposition of crop residues under non-limiting nitrogen conditions were simulated using a simple dynamic model. This model includes three compartments: the residues, microbial biomass and humified organic matter. Seven parameters are used to describe the C and N fluxes. The decomposed C is either mineralised as CO2 or assimilated by the soil microflora, microbial decay producing both C humification and secondary C mineralisation. The N dynamics are governed by the C rates and the C:N ratio of the compartments which remain constant in the absence of nitrogen limitation. The model was parameterised using apparent C and N mineralisation kinetics obtained for 27 different residues (organs of oilseed rape plants) that exhibited very wide variations in chemical composition and nitrogen content. Except for the C:N ratio of the residues and the soil organic matter, the other five parameters of the model were obtained by non-linear fitting and by minimising the differences between observed and simulated values of CO2 and mineral N.
Three parameters, namely the decomposition rate constant of the residues, the biomass C:N ratio and humification rate, were strongly correlated with the residues C:N ratio. Hyperbolic relationships were established between these parameters and the residues C:N ratio. In contrast, the other two parameters, i.e. the decay rate of the microbial biomass and the assimilation yield of residue-C by the microbial biomass, were not correlated to the residues C:N ratio and were, therefore, fixed in themodel. Themodel thus parameterised against the residue C:N ratio as a unique criterion, was then evaluated on a set of 48 residues. An independent validation was obtained by taking into account 21 residues which had not been used for the parameterisation. The kinetics of apparent C and N mineralisation were reasonably well simulated by the model. The model tended to over-estimate carbon mineralisation which could limit its use for C predictions, but the kinetics of N immobilisation or mineralisation due to decomposition of residues in soil were well predicted. The model indicated that the C:N ratio of decomposers increased with the
residue C:N ratio. Higher humification was predicted for substrates with lower C:N ratios. This simple dynamic model effectively predicts N evolution during crop residue decomposition in soil.