Limitations in Estimating Phosphorus Sorption Capacity from Soil Properties.

An important component of all P loss models is how P cycling in soils is described. The P cycling routines in most models are based on the routines developed for the EPIC model over 30 years ago. EPIC was developed so that it could be parameterized with easily obtainable soil data and thus, by necessity, oversimplifies the processes governing P transformation in soil. The EPIC model assumes three inorganic P pools: Active, Labile, and Stable. At equilibrium, EPIC partitions P among the 3 pools using a P sorption capacity parameter (PSP). In most cases this parameter is not directly measured but rather estimated from soil properties. In this study we re-evaluate existing regression equations relating PSP to soil texture properties and note limitations with use of regression equations to estimate PSP. We also demonstrate limitations with the underlying conceptual model of the P cycling routines. Finally, we propose modifications to the P cycling routines to improve the accuracy of this model for predicting P concentrations of the different inorganic P pools.