In a laboratory experiment, soil was treated with 15NH4- and 15NO3-N compounds at various times during the incubation. Several approaches in the calculation of gross immobilization and gross mineralization between two sampling dates were compared and a new method was developed. It is based on a dynamic simulation model designed to interpret isotopic data and an optimization procedure used to determine the best fit between model output and experimental data. This flexible method was used to examine the validity of the assumptions usually made when calculating the gross transformations.

The results indicate that seriously erroneous estimates of the gross transformations can follow if it is assumed that remineralization of immobilized N does not occur. Less serious errors can result from the assumption that both opposing processes occur at a constant rate during the interval between sampling dates.

The combined use of the model and the optimization procedure has several advantages over traditional methods and some of the gross transformation estimates reported would not have been obtainable using older methods.