Predicting soil solution chemical attributes from more easily measured soil and soil solution parameters

Abstract

Abstract An understanding of how soil solution ionic strength (Is) and major cation activities influence crop growth is often limited by the extensive measurements required to characterize ionic composition and subsequent speciation exercises. Easily measured solution and soil attributes need to be identified that can predict these important solution parameters. Soil and soil solution chemical properties of four Ultisols in the Coastal Plain and Piedmont of North Carolina were used to develop models to predict ionic strength and solution cation attributes. GEOCHEM‐PC‐predicted Is was linearly related to electrical conductivity (EC) across soils (r2=0.92), confirming that Is for soil solutions with complex composition can be estimated from their electrical conductivity. Models of the form lnMs=a+blnEC+clnME, or modifications thereof, were developed for predicting solution aluminum (Al), calcium (Ca), magnesium (Mg), and postassium (K) levels (Ms) from a knowledge of EC and either soil exchangeable cation #OPME) or cation saturation (MSATE) attributes. For each cation, total and free solution concentration and activity in absolute and saturation terms were investigated. The best models explained, at most, 68% of the variability associated with total solution Al concentration (Als‐T) or 74% when Als ‐T was expressed as a percent of major solution cations. Greater than 85% of the variability associated with solution Ca and Mg could also be accounted for, but only 67% of the variability associated with solution K could be explained. Including soil pH and interaction terms (MExEC, MExpH, and ECxpH) in models improved the relationship for total Al concentration (R2=0.87) and solution Ca parameters (R2 ≥0.93), but not for solution Mg and K indices. None of the models could account for >30% of the variability associated with free concentration and activity of Al3+, suggesting that the prediction of these parameters for a particular Al species could not be made from a knowledge of soil pH, solution EC, and ME or MSATE data.