Soil Color as a Proxy to Estimate Phosphorus Sorption Index: Potential and Limitations in California Vineyards

Abstract

Assessing the extent of phosphorus (P) sorption capacity (PSC) in soil agroecosystems can enhance the effectiveness of P management while preventing water eutrophication. Phosphorus sorption index (PSI) is a common metric to inform PSC, however, PSI is not typically measured in most commercial laboratories. Pedotransfer functions can be used to estimate PSC, but they rely on laborious and costly methods for extractable iron (Fe) and aluminum (Al). Hence, we aimed to develop a model to estimate PSI from readily accessible soil properties, such as available P (Olsen), clay, organic carbon, and soil color parameters that serve as proxies for Fe and Al oxides. Using soils collected from wine grape vineyards in Napa Valley, California, we built a multiple linear regression model to predict PSI. The model performed reasonably well for all soils (R2=0.62, n=86), but performed better when Ultisols with very high PSC were omitted (R2=0.72, n=65). When stratified by soil taxa, the model fit was as follows: Mollisols (R2 = 0.75; n=24) >Entisols (R2 = 0.50, n=12) >Alfisols (R2 = 0.45; n=29) > Ultisols (R2 = 0.21; n=21). The model worked best for Mollisols, which had a more uniform distribution of clay content throughout the soil profile. It performed poorly for Ultisols which had significantly higher PSI values, and clay contents that increased with depth. These results suggest that standard measurements from field observations or soil survey could be used to guide regional patterns in P sorption.

Keywords: Phosphorus sorption index, USDA soil taxonomy, soil color, soil organic carbon, Iron oxides, Clay