PREDICTION OF FIELD DENITRIFICATION RATES - A BOUNDARY-LINE APPROACH

The high degree of spatial and temporal variability in denitrification and the complexity of the controls on the denitrification process make prediction of field denitrification rates difficult. We developed a denitrification model that requires only easily obtainable data to predict daily denitrification rates on Chernozemic soils. The model was developed using boundary-line analysis to establish relationships between soil properties and denitrification for some Saskatchewan soils. Predictions of denitrification are made from water-film thickness (grams of water per unit surface area), air porosity, organic C, NO3-, and maximum daily air temperature. The coefficient of determination (R2) for the relationship between predicted and measured denitrification rates was 0.47, with 50% of the predicted denitrification rates failing within the range of the replicates for the corresponding measured rates. When the model was tested using an independent data set, the correlation (R2) between predicted and actual denitrification rates was 0.50 indicating that the model can explain approximately one-half of the variability in denitrification for a range of soils. The model can be combined with a soil water budget to estimate daily denitrification rates from readily available soil and climatic data.