Relationships Between Agronomic and Environmental Soil Test Phosphorus in Three Typical Cultivated Soils in China

A study was conducted to determine the relationships between agronomic soil test P and environmental soil test P in three soils predominately distributing in three typical agricultural production areas of China. Soil P was analyzed using Bray-1 （BP）, Olsen （OP）, and Mehlich-3 （MP） methods as agronomic tests, and using Fe-oxide impregnated filter paper （FeP）, anion-exchange resin membrane （RP）, and water （WP） as environmental tests. There were linear relationships between soil P extractable with all the tests evaluated. The regression coefficients, R^2, ranged from 0.8164 to 0.9409 between each two of the agronomic soil test P, and ranged from 0.4702 to 0.8990 between each two of the environmental soil test P, when the three soils were considered separately. When soil test P was analyzed across all the three soils, the highest regression R^2 was found between OP and MP （0.7940） amongst agronomic soil test P, and between FeP and RP amongst environmental soil test P （0.8842）. While all of the three agronomic soil test P was linearly related to each of the environmental soil test P across the three soils, strongest relationships were found between OP and environmental soil test P. Agronomic OP may be adopted as an analytical tool for environmental prediction of soil P.

A study was conducted to determine the relationships between agronomic soil test P and environmental soil test P in three soils predominately distributing in three typical agricultural production areas of China. Soil P was analyzed using Bray-1 (BP), Olsen (OP), and Mehlich-3 (MP) methods as agronomic tests, and using Fe-oxide impregnated filter paper (FeP), anion-exchange resin membrane (RP), and water (WP) as environmental tests. There were linear relationships between soil P extractable with all the tests evaluated. The regression coeffcients, R2, ranged from 0.8164 to 0.9409 between each two of the agronomic soil test P, and ranged from 0.4702 to 0.8990 between each two of the environmental soil test P, when the three soils were considered separately. When soil test P was analyzed across all the three soils, the highest regression R2 was found between OP and MP (0.7940) amongst agronomic soil test P, and between FeP and RP amongst environmental soil test P (0.8842). While all of the three agronomic soil test P was linearly related to each of the environmental soil test P across the three soils, strongest relationships were found between OP and environmental soil test P. Agronomic OP may be adopted as an analytical tool for environmental prediction of soil P.