The efficient use of phosphate fertilizers in soils of the tropical region is still a challenge, mainly regarding long-term soil management. Based on the cumulative response of three successive maize cultivations, alternatives of phosphorus supply combining P sources and application methods were compared in a formerly fertilized Red Argisol. Treatments arranged in a 4 x 3 + 1 factorial scheme represented by four P sources (180 kg ha-1 P2O5) as triple superphosphate - TS, magnesium termophosphate - MT, Arad reactive rock phosphate - RP, and Araxá rock phosphate - AP, and three application methods (broadcasted in the first year; along the sowing furrow in the first year; and annually split along the sowing furrow) and a check plot (without P application) as additional treatment. The applied P ratio was based on the total P2O5 content of each fertilizer. Cumulative biomass production, grain yield, and P accumulation in shoot and grain were determined. Residual available soil P was quantified at the end of the experiment. Indices of agronomic and economic efficiency were calculated. More differences between treatments were detected in P uptake than in grain yield. Most treatments provided similar yield increments, which was partly attributed to the residual effect of previous fertilizations and the genetic efficiency of P use by the corn plants. The most soluble sources, TS and MT, presented comparable performances. Unlike the AP rock phosphate, the split application of RP in the sowing furrow led to higher yields. Annual split applications along the sowing furrow did not affect the grain yield, and increased the residual effect. The treatments with greatest agronomic efficiency (TS and MT) were not the same as those with greatest economic efficiency. The two rock phosphates (RP and AP) presented the most favorable cost/benefit ratios.
phosphate fertilization; phosphate efficiency; rock phosphate; reactive rock phosphate; thermophosphate; Zea mays L
