Response of grain yield and milled-rice protein content to nitrogen (N) rates at various growth stages is critical for quantifying real-time and real-amount of applied N requirement for target grain yield and protein content. An experiment including 10 N rate treatments at transplanting, tillering and panicle initiation stages with four rice cultivars in 2003, 6 N treatments with two rice cultivars in 2004 and 2005 was conducted. Increase of N rates at PIS significantly increased both grain yield and milled-rice protein content but increase of N rates at tillering stage significantly increased grain yield but not milledrice protein content. Therefore, high grain yield and low milled-rice protein content would be difficult to obtain only by adjusting N rates at PIS. Internal N use efficiency (INUE) was 60.5 kg grain/kg N accumulation on an average over N treatments, cultivars, and experimental years, showing considerable reduction especially at high shoot N accumulation in the experimental year of low sunshine duration. Milled-rice protein content tended to increase almost linearly with increasing shoot N accumulation, but it revealed big variation even at the same shoot N accumulation at harvest. Milled-rice protein content decreased with increasing INUE. N accumulation in the milled rice increased at an almost constant proportion of 45.5 percent of the shoot N accumulated at harvest, showing slight decresing proportion with the increasing shoot N accumulation.

Response of grain yield and milled-rice protein content to nitrogen topdress (N) timing at panicle initiation stage (PIS) is critical for quantifying real-time N requirement for target grain yield and milled-rice protein content. Two split-split-plot experiments with three replications, one in 2004 and the other in 2005, were conducted in Experimental Farm, Seoul National University, Suwon, Korea. The experiments included three N rates at tillering stage (TS), three N timing treatments at panicle initiation stage (PIS) and two rice cultivars. The N rates at TS, N timing at PIS, and rice cultivars were randomly assigned to main plot, sub plot, and sub-sub plot, respectively. Results showed that the delayed N application at PIS reduced grain yield in 2004 and increased milled-rice protein content in both years significantly at 0.05 probability level. The calculated optimum N timing at PIS from pooled data by N rates and rice cultivars in two years was at 28 days before heading (DBH). However, real-time of N timing at PIS was dependent on plant growth and N status around PIS that in turn was dependent on applied N rates at TS. The optimum N timing at PIS was at 30 DBH for no N treatments at TS while at 27 DBH for 3.6 and 7.2 kg N/10a treatments and at 27 and 29 DBH for Hwaseongbyeo and Daeanbyeo, respectively. In general, earlier applied N at PIS resulted in lower milled-rice protein content but the highest grain yield was expected to be obtained when N topdress at PIS was applied at the time when shoot N concentration started to drop below about 23 mg/g due to dilution effect after transplanting. In conclusion, the results of our experiments imply that the currently recommended N topdress time (24DBH) at PIS in Korea should be reconsidered for the higher grain yield and the better quality of rice.

A split-plot designed experiment including four rice varieties and 10 nitrogen levels was conducted in 2003 at the Experimental Farm of Seoul National University, Suwon, Korea. Before heading, hyperspectral canopy reflectance (300-1100nm with 1.55nm step) and nine crop variables such as shoot fresh weight (SFW), leaf area index, leaf dry weight, shoot dry weight, leaf N concentration, shoot N concentration, leaf N density, shoot N density and N nutrition index were measured at 54 and 72 days after transplanting. Grain yield, total number of spikelets, number of filled spikelets and 1000-grain weight were measured at harvest. 14,635 narrow-band NDVIs as combinations of reflectances at wavelength ~lambdal~;and~;~lambda2 were correlated to the nine crop variables. One NDVI with the highest correlation coefficient with a given crop variable was selected as the NDVI of the best fit for this crop variable. As expected, models to predict crop variables before heading using the NDVI of the best fit had higher r2 (>10~%) than those using common broad- band NDVI red or NDVI green. The models with the narrow-band NDVI of the best fit overcame broad- band NDVI saturation at high LAI values as frequently reported. Models using NDVIs of the best fit at booting showed higher predictive capacity for yield and yield component than models using crop variables.