Phosphorus fixation in tropical soils may decrease under no-till. In this case, P fertilizer could be surface-spread, which would improve farm operations by decreasing the time spend in reloading the planter with fertilizers. In the long term, less soluble P sources could be viable. In this experiment, the effect of surface-broadcast P fertilization with both soluble and reactive phosphates on soil P forms and availability to soybean was studied with or without fertilization with soluble P in the planting furrow in a long-term experiment in which soybean was grown in rotation with Ruzigrass (Brachiaria ruziziensis). No P or 80 kg ha-1 of P2O5 in the form of triple superphosphate or Arad reactive rock phosphate was applied on the surface of a soil with variable P fertilization history. Soil samples were taken to a depth of 60 cm and soil P was fractionated. Soybean was grown with 0, 30, and 60 kg ha-1 of P2O5 in the form of triple phosphate applied in the seed furrow. Both fertilizers applied increased available P in the uppermost soil layers and the moderately labile organic and inorganic forms of P in the soil profile, probably as result of root decay. Soybean responded to phosphates applied on the soil surface or in the seed furrow; however, application of soluble P in the seed furrow should not be discarded. In tropical soils with a history of P fertilization, soluble P sources may be substituted for natural reactive phosphates broadcast on the surface. The planting operation may be facilitated through reduction in the rate of P applied in the planting furrow in relation to the rates currently applied.
A leaky model of long-term soil phosphorus dynamics
