EMPIRICAL MODELING OF SOIL NITROGEN MINERALIZATION1

Abstract

The goodness-of-fit of first-order and empirically determined equations to net nitrogen mineralization data obtained by various investigators in laboratory incubations and greenhouse and field experiments is compared. In the majority of instances the empirically determined equation has a lower root-mean-square than does the first-order model. Parabolic functions of the form Y = AXB provided a good fit to most of the data. All the greenhouse and field data were fitted closely by the parabolic model or by a linear model, which can be considered as a special case of the parabolic model where B = 1. The first-order model, based upon the concept of a single pool of mineralizable N of discrete size, does not conform to reality and is inferior to empirically determined models as a means of describing nitrogen mineralization.