Department of Plant Sciences

Three soil tests: Mehlich-1, mixed cation-anion exchange resins with HCO3 as anion, and iron oxide impregnated filter paper (Pi strip), were used to assess available P in a highly weathered tropical Oxisol which received rates of 0, 2.5, 12.5, 25, 50 and 75 g kg-1 of phosphogypsum, natural gypsum (gipsite) or pure gypsum (reagent grade) and 0 and 100 mg kg-1 of P [(triple superphosphate (TSP) or phosphate rock (PR)]. The treated soil samples were incubated for 25 days at room temperature followed by P extraction with the three methods. The trend of Mehlich-P was similar to Pi-P in response to rates and types of gypsum applied. For resin-P, it was underestimated due to the reaction of resin-HCO3 and CaSO4 that resulted in the formation of CaCO3. Consequently, the strength of resin-HCO3 was weakened and the CaCO3 formed could also further adsorb/precipitate soluble P and thus resin-P was underestimated. Mehlich-1 overestimated available P from PR with respect to TSP because its strong acidity that extracted undissolved PR. Gypsum depressed PR solubility due to the Ca common-ion-effect and thus available P of PR was underestimated by the resin and Pi tests. It is concluded that as long as a significant amount of phosphogypsum or natural gypsum applied to tropical acid soils still remains in the soils, it should be cautious to select a proper soil P test and provide a scientific interpretation in order to make a correct recommendation for the rate of applied water-soluble or water-insoluble P fertilizers.