Assessment of phosphorus availability in soils fertilized with Langebaan rock phosphate or superphosphate

Abstract

Problems currently exist in the interpretation of soil test values for soils fertilized with rock phosphate. This study was undertaken to compare the ability of four extradants (Ambic, Bray 1, Mehlich 3 and modified Truog) to assess plant available phosphorus in soils fertilized with superphosphate or Langebaan rock phosphate (Langfos). A Balmoral clay (Typic Haplorthox) and an Avalon sandy loam (Plinthic Paleudult), adjusted to three levels of acidity and fertilized with superphosphate or Langfos, were consecutively cropped three times with maize (Zea mays L.) in a glasshouse. Soil samples were taken after each harvest and analysed for available phosphorus using the four extradants. Inorganic phosphorus in samples, taken after the third crop, was fractionated into aluminium-, iron- and calcium-bound phosphorus. Ambic, Bray 1 and Mehlich 3 were more sensitive than Truog in reflecting the favourable effect of acidity and cropping on the dissolution of Langfos. In spite of large differences in the quantity of phosphorus extracted by the Ambic, Bray 1 and Mehlich 3 reagents, they were closely correlated with one another and with the aluminium - and iron-bound phosphorus, but not with the Truog extradant or calcium-bound phosphorus. On the Avalon soil, the relative plant availability of phosphorus from Langfos was underestimated by Ambic and Bray 1 and overestimated by Truog. The Mehlich 3 extradant provided the best index of relative phosphorus availability. All extradants with the exception of the Truog, which overestimated phosphorus availability from Langfos, provided a reasonably satisfactory index of relative phosphorus availability on the Balmoral soil.