Having the objective of evaluating the efficiency of three nitrogen aplication systems in the tillering stage of rice, there was an experiment installed in the Federal University of Santa Maria (RS), Brazil, during 1998/99 growing season. The experimental design was two-factorial 3 x 5 in complete randomized block design with 4 replication in a split plot arrangement of treatments. Three forms of N application at tillering stage were placed in the main plots: 1) N + flushing where the N was topdressed and a flush of water was applied one day after and then flooding fourteen days later; 2) N + flooding where N was topdressed the day before the definitive flooding; 3) flooding + N, with the N applied in the water one day after flooding. Four rates of N were applied in the plots (0, 20, 40 and 60 Kg of N ha-1) and an absolute witness and control without N. In every treatment which received N, there was an application of 10 Kg ha-1 in the beginning of internode elongation. It was concluded that nitrogen application from tillering to the internode elongation increased tiller number and total N uptake. This increase is dependent on the nitrogen application system being greater in treatments of N + flush and N + flooding in relation to the flooding + N treatments. This increment of N doses provides an increase in grain production, while the forms of urea application at tillering stage don’t affect grain production when using N in the panicle initiation.
flooded rice; nitrogen; urea
