The CENTURY Soil Organic Matter Model and its daily timestep derivative DAYCENT have been widely used to predict the effects of agro-ecosystem management practices and global change on soil C and N storage and dynamics.

A major problem with CENTURY, and other similar soil organic matter (SOM) models, is that the SOM pools are conceptual, being defined kinetically and cannot be directly estimated by laboratory methods, making it difficult to initialise the model. Typically a very long period of historical landuse must be simulated to reach steady state conditions.

The soil organic C and N submodel was analysed to derive an algorithm to calculate equilibrium levels of the C and N pools in relation to plant litter inputs, litter quality, soil moisture, temperature, pH, soil texture and surface area. Average annual values for these variables can be used to calculate approximate equilibrium pool sizes to assist in model initialisation.

The sensitivity of the SOM sub-model to model parameters can be demonstrated. With other factors remaining constant, SOM levels are predicted to be directly proportional to C inputs. With C inputs constant, SOM levels are higher at low temperature and moisture status, in anaerobic soil conditions, with high clay surface area or low sand contents, and with high litter lignin content. Litter N status has little effect on either soil organic C or N, while increasing fertiliser N will decrease the C:N ratio.

Predictions are discussed in relation to observed SOM levels in long-term field experiments.