Grain moisture kinetics: a new and effective method to determine grain filling duration in rice

Grain-filling duration (GFD) is a crucial determinant of grain yield in rice. Genotypes with longer GFD would benefit from more accumulation of assimilates by current photosynthesis, thereby potentially improving yields. The current method of sampling panicles at regular intervals, separating the grains, and drying to determine GFD is tedious, slow and takes a long sample processing time to obtain results. The authors developed a simple, convenient, and more efficient method based on grain moisture content (GMC) kinetics. The objectives were to determine if GMC can be used to estimate GFD and if it can discriminate variation in GFD among genotypes and environments. This method was tested in the 2012 dry- (DS) and wet-season (WS) trials at the PRRI farm and in the DS at PhilRice, Maligaya (PRRI) using high-yielding varieties planted at two planting densities. Furthermore, 295 indica accessions were grown at PRP1. Using a portable micro-grain meter, GMC was measured from about 15 grains sampled on panicles at weekly intervals starting at 15 d after maximum flowering. By linear regression of GMC over time, GFD was estimated based on a target GMC (20 or 18%) at which grain filling was considered completed. Grain-filling duration varied across locations and season. In the DS at 20% MC, GFD at PRRI was longer (22-30 d) than that at PRRI (28-33 d). At 18% MC, the range was from 23 to 31 d at IRRI and from 31 to 34 d at PRRI. In the WS, the onset of grain filling was late and the range shorter, 26-29 d at targeted 20% MC and 27-30 d at 18% MC. Generally, varieties planted under high population density tended to have shorter GFD. Clear genotypic variation in GFD from 50% flowering up to 18% grain moisture among the 295 indica accessions was observed. Based on the knowledge that water content of grains decreases as they mature, grain moisture kinetics content of grains decreases as they mature, grain moisture kinetics may be used in estimating GFD in rice. Caution must be taken, however, to avoid errors from (1) rain, (2) sampling error due to the GMC small sample sizes, and (3) inaccurate observation of time of flowering.