抄録
The interaction between N application and water management (ordinary water (OW) and deep-water (DW)) in dry-seeded rice was studied in 1995 and 1996. Deep-water treatment increased the percentage of ripened grain and grain yield, and decreased panicle number per m2. The maximum number of tillers and the percentage of productive tillers was increased by increasing the amount of N applied (0, 70, 150, 220 kg ha-1) in both water treatments. More tillers was produced in OW than in DW. However, the percentage of productive tiller was higher in DW than in OW (78% in 1995 and 72% in 1996 for OW and 86% in 1995 and 83% in 1996 for DW). There were significant differences in yield components and yield between the amount of N applied, and between the two water treatments. The panicle number per m2 (r=0.904 in OW, r=0.921 in DW) was positively correlated with the amount of N applied. In DW, the number of panicles was lower than in OW, but the percentage of ripened grain was higher than in OW. The total nitrogen contents in the flag leaf was positively correlated with the amount of N applied and the concentrations of nitrogen in flag leaf was positively correlated with the grain yield. The partial factor productivity (PFP) of applied N was significantly higher in DW than in OW at all levels of nitrogen application. Agronomic nitrogen-use efficiency (ANUE) was not different between the two water treatments. These results imply that the suppression of tiller formation by elevated water level would improve the yield determinants and utilization efficiency of indigenous soil N, which consequently helps to increase yield production.
