The aim of this study was to determine medium and long management of phosphorus fertilization with two application methods on the maize. An experiment was established in El Ejido, Panama, from 2004 to 2013 on Udic Haplustalf soil. Two factors were evaluated in a 5 x 4 factorial arrangement of split plot under a randomized block design with three replications. The main plot was the broadcast application of P in rates of 0, 25, 50, 75, and 100 kgP/ha (applied only the first year), while in sub-plots applied P in band, for ten years, together with N and K doses of 0, 10, 20 and 30 kgP/ha. The maximum grain yield was achieved by applying average 304 kg (average of 30.4 kgP/year). It was observed that the residual effect broadcast applications of 0 and 25 kgP/ha had a relative yield below 80%. Treatment of 50 kgP/ha remained at 82%, while 75 kgP/ha declined from the year 2007 and is at a value close to 80%. Treatment of 100 kgP/ha broadcast presents a relative yield close to 100%. The results indicate that the broadcast application over 50 kgP/ha did not differ from the response obtained with annual applications of 30 P/ha in the first five years. These results suggest that the residual effect of phosphorus broadcast and incorporated before planting, on the type of soil used, varies according doses applied initially, where dosages above 50 kg/ha should be repeated after six years to maintain good productivity, since, from this time the residual effect is drastically reduced.

Keywords: broadcast, residual effect, critical level, optimal physical.