The Effect of Soil Phosphorus Compounds on Soil Test Correlation

Abstract

AbstractTwenty soils representing the level upland and the depressional soils in the hydromorphic sequence of profiles of the upland and lake plain catenas were characterized by the determination of pH, organic matter, clay content, acid‐soluble phosphorus, water‐soluble phosphorus, alkali‐soluble phosphorus, organic phosphorus, and iron content.When the above variables were included in a multiple correlation analysis using yield of phosphorus as the dependent variable, a multiple correlation coefficient of 0.931 was obtained. The elimination of all variables except acid‐soluble and alkali‐soluble phosphorus gave a multiple correlation coefficient of 0.851. The simple correlation coefficient for acid‐soluble phosphorus was 0.557 and for alkali‐soluble phosphorus 0.675. The increase in correlation where both acid‐soluble and alkali‐soluble phosphorus were used over the correlation for either one alone was significant at the 1% level. Hence, the use of both acid‐soluble and alkali‐soluble phosphorus were required to explain the contribution of the different soil phosphorus compounds to yield.Solubility curves of the 20 soils indicated that the soils were heterogeneous with respect to soil phosphorus compounds. The depressional soils from the lake plain catena contained more calcium phosphate and less iron and aluminum phosphate compounds than their counterpart in the upland catena.Solubility curves for calcium phytate and other organic phosphorus compounds indicate that they are partially acid‐soluble, suggesting that an acid extraction contains phosphorus from both organic and inorganic sources. This was supported by the fact that a significant correlation was obtained between organic matter and acid‐soluble phosphorus.