Comparison of Methods for Determining Critical Concentrations of Soil Test Phosphorus for Corn

Abstract

AbstractCritical concentrations of soil‐test P (STP) are used to identify soils where response to P fertilization should be expected. There is, however, little agreement concerning the methods that should be used to identify critical STP concentrations. This study compares the efficacy of critical STP concentrations generated by using various methods. Twenty‐five P fertilization trials with corn (Zea mays L.) were established in Iowa. Available soil P at each site was estimated by the Bray‐P1, Mehlich‐3, and Olsen extractants. Corn yield response was expressed in both absolute and relative terms and then related to STP values by using various statistical models (Cate‐Nelson split, linearplateau and quadratic‐plateau segmented polynomials, the quadratic polynomial, an exponential Mitscherlich‐type equation, and a multivariate polynomial). The use of various combinations of the extractants, expressions of yield response, and models resulted in a wide variety of critical STP concentrations. Comparisons of the ability of each critical concentration to generate economic returns when used to guide fertilization across the 25 sites showed that selection of the model was much more important than selection of the extractant or the expression of yield response. The best model was the Cate‐Nelson, which identified critical concentrations of 13 rag kg−1 for the Bray‐P1, 12 mg kg−1 for the Mehlich‐3, and 5 mg kg−1 for the Olsen extractants. Overall, the results of this study demonstrate that selection of the most appropriate critical STP concentration can be a major factor affecting the profitability of fertilization in areas having an abundance of soils testing high in P.