Available Phosphorus Evaluated by Three Soil Tests in a Brazilian Tropical Oxisol Treated With Gypsum

Mehlich-1, resin-HCO(3), and Pi tests were used to assess available P in an acid tropical Oxisol in Brazil treated with gypsum, which has been preferred over lime to ameliorate the Al toxicity in the subsoil. The soil was incubated in the laboratory at rates up to 75 g kg(-1) of phosphogypsum (PG) containing 0.3% total P, natural gypsum, or reagent-grade gypsum, and up to 100 mg P kg(-1) as triple superphosphate (TSP) or phosphate rock (PR). In the greenhouse, two consecutive maize crops were grown on the soil treated with 50 mg P kg(-1) of TSP and PG rates up to 75 g kg(-1). The results of the incubation study showed that Mehlich-P and Pi-P increased with increasing PG rate for the treatments of TSP, PR, and control. Resin-HCO(3) underestimated available P from TSP and PR because of the reaction between resin-HCO(3) and gypsum. Mehlich-1 overestimated available P from PR compared with TSP because of an excessive dissolution of PR by the strongly acidic Mehlich-1. Pi underestimated available P from PR in the treatments of natural and reagent-grade gypsum because of Ca common-ion effect from gypsum on depressing PR dissolution. The results in terms of the effect of PG on available P are similar in both incubation and greenhouse studies. Both Mehlich-P and Pi-P correlated well with P uptake by maize, whereas resin-P did not