Comparison of methods for determination of phosphorus retention

Abstract

Three methods for determination of the phosphate retention capacity were compared to evaluate their suitability for a quartz sand, two Typic Hapludults (pH 4.3 and 6.0), two allophanic Andosols (pH 5.6 and 6.2), and a nonallophanic Andosol (pH 6.1), of which three soils were lacking in Bray II extractable phosphorus (P). The Andosols demonstrated large phosphate sorption, above 8000 μg P g^-1 according to the Japanese method, or 95–99% P retention of the P applied in the New Zealand (NZ) method, but none of the methods could significantly separate the non-allophanic soil from the allophanic ones. Use of the Mehlich method allowed for the total sorption of P by both the virgin Ultisol and all the Andosols, but separated the virgin Ultisol from the limed and fertilized one, which had a P sorption close to that of sand. Both the Japanese and NZ methods separated the virgin Ultisol from the Andosols. The quantity of P added (80, 5000 and 11730 μg P g^-1 soil in the Mehlich, NZ and Japanese methods, respectively), solution pH, retention time and shaking intensity explain the method dependent differences. Soil dependent differences were attributed to allophane and Al- and Fe-humus complexes in the Andosols, and sesquioxides and goethite in the virgin Ultisol.