Management of fertiliser nitrogen in dry-seeded, delayed-flood rice

Abstract

Approaches taken in management of fertiliser nitrogen (N) in dry-seeded, delayed-flood rice (Oryza sativa L.) in the United States of America vary considerably. A 2-year study was conducted in Louisiana to evaluate N fertiliser efficiency and accumulation when applied in single or multiple applications. The cultivars Lemont and Tebonnet, representing semi-dwarf and conventional tall plant types, respectively, were grown on a Crowley silt loam (Typic Albaqualfs soil). Urea N was applied using 1 of 3 strategies: in a single (1-way) preflood (PF) application of 134 kg Nlha; in a 2-way split consisting of 100 kg N/ha PF and 34 kg N/ha at panicle initiation (PI); or in a 3-way split consisting of 66 kg N/ha PF, 34 kg N/ha at PI, and 34 kg N/ha at panicle differentiation. Microplots were established within each whole plot and fertilised with 2 atom% 15N urea. Fertiliser N application time had no influence on plant accumulation of 15N urea or biomass production. Plant accumulation of 15N urea was significant across year and cultivar. Tebonnet accumulated more 15N urea than Lemont. Accumulation was also significantly higher in 1992. Method of application of fertiliser N significantly affected grain yields. The highest average grain yield of 7905 kg/ha occurred with a single PF application. Yields were reduced by 378 kg/ha with a 2-way split application and by 1074 kg/ha with a 3-way split application. Method of application of main crop fertiliser N had no effect on ratoon crop yields. Total grain yield (main + ratoon) was significantly reduced with the 3-way split. This study demonstrates the importance of supplying adequate N to the rice plant before permanent flood establishment. It also suggests that midseason N applied at specific growth stages is less efficiently utilised, especially when insufficient N is applied at PF.