Estimation of plant response parameters to phosphorus on the basis of some soil chemical properties in Ultisols and Andisols from the south of Chile

Maria Adriana Carrasco R.1 y José Domingo Opazo A.1

The application of phosphate to soils is usually based on the soil chemical behavior and the availability of phosphorus which is influenced by the history of the soil management and it is measured by a chemical method. The objective of the present study was to select soil chemical properties related to the plant phosphorus response parameters. Sixteen soil samples were taken in the IX and X Regions for the analysis of some chemical properties and two greenhouse experiments were done, each one with eight soils, using perennial ryegrass as indicator. A complete randomized blocks design was used with two treatments (with and without P). Relative yields (RY) and absorbed P (Pab by ryegrass were determined. The soil chemical properties most closely related to the P response parameters were selected in the first greenhouse experiment; by multiple regression analysis (stepwise) equations for estimating the Pab (when P was applied), and the RY were obtained as a function of pHKCI and Aloxalate in the first one, and as a function of pHKCI and total acidity in the second one, with determination coefficients R2 of 0.995 (P< 0.01) and 0.822 (P< 0.01), respectively. On validating the equation for Pab with the data of the second greenhouse experiment, the soils Frutillar, Freire, Ralún and Nueva Braunau did not fit the equation. A second multiple regression analysis considering the data of both experiments, excluding the problem soils (n = 12), gave a similar equation with higher coefficients and R2 = 0.945 (P<0.01); for the RY an equation in function of Pab for the treatment with P was found with R2 =0.963 (P<0.01). On the other hand, it was found that the delta of P absorption in the treatments with and without P responds to an equation in functlon of the pHKCI and Sioxalate with R2 = 0.992 (P

Keywords: P absorption, P response in Andisols and Ultisols, models to predict P response of plants.
1 Departamentode Ingenierla y Suelos, Facultad deCiencias Agrarias y Forestales, Universidad de Chile, Casilla 1004, Santiago, Chile.