Development of an alternative to the Olsen bicarbonate-extraction test for determining plant-available phosphorus in basaltic soils

Abstract. Recent work has demonstrated that the Olsen test for phosphorus (P) is an unreliable predictor of plant-available P in soils derived from basalt parent material in Northern Ireland. The present study was conducted to develop a more reliable soil-P test for these soils by regressing P fractions removed from soil by various chemical extractants against herbage P indices calculated from plant tissue test data using a diagnosis and recommendation integrated system. The degree of P saturation of the soil P sorption capacity, based on ammonium oxalate extractable P, Al and Fe, provided a better prediction of P available to swards on basaltic soils than either the Olsen test or a number of other well-known soil-P test procedures. The superiority of the degree of P saturation test on basaltic soils was attributed to the fact that it simultaneously takes account of both P quantity and P buffering capacity factors in predicting P availability. The Olsen-P test, which accounts for the P quantity factor alone, was only reliable for non-basaltic soils. Re-classifying the P fertility status of basaltic soils according to the degree of P saturation test could result in considerably less P being recommended for these soils with possible consequential benefits to water quality.