U.S. flag

An official website of the United States government

Dot gov

Official websites use .gov
A .gov website belongs to an official government organization in the United States.


Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.


Main content area

Comparison of Two Methods for Calculating the P Sorption Capacity Parameter in Soils

Carl H. Bolster, Peter A. Vadas
Soil Science Society of America journal 2018 v.82 no.2 pp. 493-501
computer simulation, inorganic phosphorus, landscapes, models, soil, sorption
Core Ideas Two methods for calculating soil P sorption capacity parameter (PSP) were evaluated. PSP values from the two methods were not well correlated with each other. PSP values calculated by these two methods are likely not estimating the same soil property. An assumption in the EPIC model that the ratio of stable to active P is 4 may not be valid. Methods of PSP estimation need improvement, and existing P‐cycling models need further evaluation. Phosphorus cycling in soils is an important process affecting P movement through the landscape. The P cycling routines in many computer models are based on relationships developed for the EPIC model. An important parameter required for this model is the P sorption capacity parameter (PSP). Using previously published data, we compare two methods for estimating PSP values: (i) measurement of changes in labile inorganic P (Pᵢ) concentrations following 6‐mo soil incubation studies, and (ii) calculation of PSP from concentrations of total Pᵢ and labile Pᵢ estimated from commonly used soil test P extraction methods. Depending on how labile Pᵢ was estimated, we either found a very poor correlation or large (>50%) median differences in PSP between these two methods suggesting they are not estimating the same soil parameter. We also found that PSP values calculated from soil incubation studies significantly underpredicted total Pᵢ. It is not clear whether this underprediction is due to limitations in the experimental approach for measuring PSP using 6‐mo incubation studies, or whether it is a result of limitations with the model itself. Our results also challenge the validity of the assumption in EPIC that stable P is four‐fold the size of the active Pᵢ pool. While calibrating PSP from measured soil P data is less costly and time consuming than long‐term soil incubation studies, we show that the fitted values are dependent on how labile Pᵢ is estimated and thus are not representative of an independently measurable physically based parameter.