The effectiveness of rock phosphate fertilisers in Australian agriculture: a review

Abstract

Plant responses to apatite rock phosphates and Calciphos, a calcined calcium iron-aluminium rock phosphate fertiliser, have been measured in many pot and field experiments in Australia, but there is no consistent view of the agronomic effectiveness of these fertilisers. Quantitative indices of the effectiveness of freshly applied rock phosphates relative to freshly applied superphosphate (relative effectiveness or RE values) have been calculated from the data for 164 Australian pot and field experiments on the basis of the substitution value of the rock phosphates for superphosphate. RE values range from <0.1 to 2.5, with the mean value for apatite rock phosphates being 0.26 compared with 0.42 for Calciphos. Statistical analysis of the data demonstrate that variations in RE values were primarily due to systematic differences in experimental design and fertiliser solubility, and not to differences in soil pH, plant species, the capacity of the soil to adsorb P and mean annual rainfall. All RE values >0.4 were obtained from experiments in which only 1 or 2 levels of fertiliser were applied to soils that were poorly responsive to applied P; thus it was not possible to define the complete response curves required to obtain precise values of RE. In most cases, RE values were < 0.4 for experiments in which several levels of fertiliser P were applied to highly P-responsive soils so that complete response curves were defined and precise values of RE estimated. The effectiveness of previously applied rock phosphate (i.e, residual value) remained low and approximately constant for several years after application, being 5-30% as effective as freshly applied superphosphate for the various experiments. The corresponding average value of the relative effectiveness of superphosphate declined by 40% in the first year after application, by a further 15% in the second year, and by a further 30% over the remaining 6 years. However, the magnitude of these declines in relative effectiveness differed substantially between individual sites. The residual value of both the superphosphate and rock phosphate fertilisers appears not to have been systematically influenced by soil type, soil pH, the capacity of the soil to adsorb P, mean annual rainfall, and whether the fertilisers were topdressed or were incorporated. However, there is a need for additional experiments to investigate the influence of these factors on fertiliser effectiveness. It is concluded that, on the basis of published data, rock phosphate fertilisers cannot be regarded as economic substitutes for fertilisers containing water-soluble P for most agricultural applications in Australia. This is because, relative to freshly applied superphosphate, the fertiliser effectiveness of rock phosphates is low in the year of application and it remains low in subsequent years so that uneconomical, very high rates of application of rock phosphate are required.