Published online by Cambridge University Press: 27 March 2009
Surface soils (0–15 cm) from a field experiment on a sandy podzol were used to evaluate the residues from basic slag, Gafsa rock phosphate and superphosphate (using ryegrass in pots), and to measure the solubility and buffer capacity during cropping, and the distribution of the residual phosphate in different particle-size fractions of the soil.
After applying P at about 3·6 g P/m2/year for 6 years, residues of the three P fertilizers were equivalent but after applying more P (9–11 g P/m2/year) for a further 8 years, basic slag and rock phosphate gave larger labile P values, were more effective for ryegrass and buffered soils more than residues of superphosphate. The percentage recoveries of P from the surface soils were approximately 7, 40 and 65 % of the total applied in 8 years as superphosphate, basic slag and rock phosphate respectively.
Residues from basic slag (mainly 75–20 μm particles) accumulated in the < 20 μm fraction of the soil. Rock phosphate (mainly 200–20 μm particles) dissolved more slowly and about one-half of the material > 20 μm remained in the sand fractions of the soil.
Residues from rock phosphate maintained values of ½pCa + pH2PO4 and pH–½pCa consistent with the solubility of fluorapatite, even when P was removed from the soil by exhaustive cropping with ryegrass. Residues from basic slag were more soluble than hydroxyapatite.