Statistical considerations associated with the use of the heterotrophic activity method for estimating Vmax and K′ for aquatic environments

Abstract

This article deals with whether linear or nonlinear, ordinary least-squares regression analysis is superior in the estimation of the kinetic parameters Vmax and K′ from data sets generated by the Wright–Hobbie heterotrophic activity method. An analysis of variance showed that weighting of data was theoretically required for the optimal estimation of Vmax and K′ for both models. However, the magnitude of the parameter estimates was largely independent of the weighting function employed. Monte-Carlo analysis also indicated that the nature of the weighting function did not appreciably affect the parameter estimates, and more importantly, that there was not a strong statistical justification for the use of the nonlinear model over the linear model in the estimation of Vmax and K′.