Genotypic Variation In Morpho-physiological Traits And Its Effect On Rice Yield Potential

Yield potential is defined as the maximum yield of a variety when grown under optimal crop management conditions. Yield improvement remains the main target for physiologists and breeders. Better morphological architecture means higher radiation use efficiency through spatial configuration, which is essential to higher quality population. Under optimal crop management conditions, studying the effect of morpho-physiological traits on rice yield potential and comparing the effectiveness between trait-based selection and yield-based selection were significant for properly selecting rice seeds with highest potential for the breeders and improving the yield potential of rice. Based on the selection of genotypes in 2008 dry season(DS), the field experiments were conducted in the same field at International Rice Research Institute (IRRI) farm for three consecutive seasons (2009 DS,2009 wet season (WS), and 2010 DS). In 2008DS, based on 196 breeding genotypes (F6 to F7 generations) grown in a replicated yield trial in a breeder’s field at the IRRI farm,52 genotypes were selected visually based on plant morphological traits within one week before final harvest. These traits included moderately tall plants, erect top three leaves, panicles located inside the canopy, large and compact panicles with more spikelets per unit panicle length, and large grain size. Sixteen genotypes were further selected from the 52 genotypes based on aboveground total dry weight and grain-filling percentage. Three out of the 16 genotypes selected based on plant traits were also ranked in the top 16 based on grain yield. A group of three check varieties (IR72, NSICRcl58, and a hybrid) was also included in all field experiments. The objectives were to study the effect of morpho-physiological traits on rice yield potential, compare the effectiveness of trait-based and yield-based selection in increasing rice grain yield, and determine whether genotypes with ideal plant traits have the potential to express higher yield under optimal crop management conditions. The following results were obtained:(1) Plant heights at flowering stage and maturity were significantly higher in the traits-based selection and the overlap than those in yield-based selection and check. The difference in plant height measured to the tip of the highest leaf minus that measured to the tip of the panicle was used to judge the relative height of panicles within a canopy. Results showed that the differences in plant height were positive for all the genotypes, which meant all the panicles were below the flag leaf. The difference in plant height was significantly higher in trait-based selection than in yield-based selection, and the difference was smallest in yield based selection. Therefore, compared to yield-based selection, the relative position for trait-based selection was better for photosynthesis of the flag leaf.(2) The selection based on plant traits had the highest stem diameter, which was significantly thinner than that of check. At the flowering stage, the tiller number was significantly higher in yield-based selection than in trait-based selection. The tiller number dropped with the death of some non-productive tillering, and the number of total tiller at maturity was lower than that at flowering. However, the selection based on yield had higher tiller number and percentage of tiller than the selection based on plant traits. So, higher effective tiller number was the fundamental to higher yield in yield-based selection compared with that in trait-based selection.(3) There was no significant difference in LAI and CGR among the four genotypic groups at the panicle initiation stage. During the flowering stage, trait-based selection had higher LAI than yield-based selection. The order of LAI for the lines during different stages was:Flowering> maturity> panicle initiation. There was significantly higher total dry weight in trait-based selection than that of yield-based selection, and the total dry weight was lower in wet season than that in dry season.The contribution of dry matter before heading to grain for the 32 genotypes was 26.9%,20.2% and 26.9%, respectively, in 2009DS,2009WS and 2010DS. After heading, the contribution of dry matter to grain in the three consecutive seasons was73.1%,79.8% and 73.1%, respectively. The grain yield was highest in the check; the average yield of trait-based selection was not significantly higher than that of yield-based selection in all three seasons. In fact, yield-based selection produced significantly higher average yield than trait-based selection in 2010DS. Therefore, higher total dry weight during the flowering stage in trait-based selection did not lead to higher grain yield compared to yield-based selection, which might be due to lower harvest index and lodging at maturity.(4) During the early vegetative, the intercepted radiation percentage was increased faster with rice rapid growth. The intercepted radiation percentage was about 91% 40 days after transplanting. After that, it increased slowly and the intercepted radiation percentage was 96.9%,97.1% ,97.4%, and 96.6%, respectively, for the check, trait-based selection, yield-based selection and overlap. Among the four groups, the check had highest grain yield, and the correlation coefficients (R2) between the intercepted radiation percentage and grain yield were 0.8503 and 0.9265, respectively,41 and 54 days after transplanting. This showed high intercepted radiation percentage was fundamental to high grain yield during the early vegetative stage.(5) The SPAD values at flowering stage were 35.0,32.8,34.7, and 33.8 in check, trait-based selection, yield-based selection, and overlap, respectively. After topdressing at flowering, the SPAD value climbed to the peak, and then dropped to the minimum. There was negative correlationship between grain yield and the difference in maximum and minimum SPAD per day and its difference, and the coefficients were 0.6851 and 0.7114, respectively. Results showed that higher SPAD value in flag leaf during the postheading period was important for higher grain yield. Thus, nitrogen topdressing at flowering could increase chlorophyll content, delay the duration of chlorophyll degradation, and maintain green leaf for a long time, which were significant for improving grain yield.The content of nitrogen in leaf was higher than that in stem during the flowering stage, and yield-based selection had higher nitrogen in leaf than trait-based selection. During the grain filling, the content of nitrogen in stem and leaf transferred to panicle. The content of nitrogen in filled grain was highest in the four groups. The nitrogen accumulation of the 32 genotypes was 52.6、54.7% and 59.5%, respectively, in 2009DS, 2009WS and 2010DS. Higher nitrogen dry matter production efficiency in trait-based selection might be the reason why there was high dry matter weight during the flowering stage. The nitrogen grain production efficiency in yield-based selection was higher than that in trait-based selection, and there was significant difference between the two groups in 2010DS. Higher nitrogen grain production efficiency in yield-based selection might be the one reason why it had higher grain yield.(6) Both spikelets/leaf (m2) and filled spikelets/leaf (m2) in yield-based selection were higher than those in trait-based selection. In 2009WS and 2010DS, the filled spikelets/leaf (m2) in yield-based selection was significantly higher than that in trait-based selection. Because spikelets/leaf (m2) and filled spikelets/leaf (m2) could reflect the sink size per unit area of leaf and photosynthesis after heading. Results showed that sink size in yield-based selection was higher than that in trait-based selection, and yield-based selection had better photosynthesis after heading. So the yield-based selection could build high quality population and improved photosynthesis, and its grain yield and yield potential were higher compared with the trait-based selection.There was positive relationship among spikelets/leaf (m2), filled spikelets/leaf (m2), grain weight (mg)/leaf (m2) and yield, which indicated improved spikelets/leaf (m2) could not decrease grain filling percentage and grain weight, on the contrary, it could improve photosynthesis intensity and accelerate transporting photosynthesis production to grain.In conclusion, trait-based selection resulted in higher plant height, leaf area index, total dry weight at flowering, and grain weight than yield-based selection. In genotypes selected on the traits of plant traits, more leaf area was above panicles as reflected by a greater difference between plant heights from the base to the tip of the flag leaf and to the tip of the panicle. However, panicle size and grain-filling percentage were not improved by trait-based selection. Trait-based selection did not increase grain yield compared with yield-based selection across three seasons with a large differences in climatic yield potential. In fact, grain yield was significantly lower in trait-based selection than in yield-based selection in 2010DS. Among all 29 tested genotypes, maximum yield was produced by yield-based selection and minimum yield came from trait-based selection in all three seasons. Therefore, genotypes with ideal plant traits did not express higher yield under the optimal crop management conditions of this study. Yield-based selection was more effective in increasing grain yield than trait-based selection in breeding.