The Research On Growth Dynamics And Yield Trait Of Hybrid Rice Under Different Sowing Dates And Genotypic Variation

This research aimed to study the adaptability to drought-evading direct seeding techniques and direct-seeded after rape(wheat) harvest of hybrid rice in northeast hills of Sichuan. Twenty representative hybrid rice varieties were chosen as trial materials and directly seeded in eight times from April 10 to May 31(7 days per period). We studied the growth characteristics, grain yield and yield components(including effective panicle, spikelets per panicle, filled spikelets per panicle, seed setting rate and yield per plant) of these rice varieties, then the adaptability to delayed seeding of these varieties were evaluated preliminary in accordance with their performance in above terms. Results are as follows:1.Compared with traditional seeding date(April 10, T1), rice plants seeded from April 10 to May 17(T1 to T6) got shorter and shorter durations from seeding to heading with sharper delays, in contrast, longer and longer grain filling stages were observed simultaneously. For the whole growth stage, rice plants seeded from T1 to T6 went through different degree of shortening. Instead, the ones seeded at T7 and T8(May 24 and May 31) got longer growth periods. As far as effective accumulative temperature concerned, most of these varieties obtained less ones, while very few varieties got more and more effective accumulative temperature with sharper delayed seeding date.2. With seeding date delayed, the effective panicle of these rice varieties happened with some regularity. Rice plants seeded during T1 and T6 produced similar amount of effective panicles, however, the ones seeded at T7 and T8 obtained significantly more effective panicles on average, which indicated that delayed seeding showed no significantly negative effective on effective panicles.3. Spikelets per panicle and filled spikelets per panicle showed no obvious regularity when seeded during T1 to T6, while they showed a significant reduction trend among a few varieties when seeded at T7 and T8. Compared with panicle size, seed setting rate was affected less by seeding date. All rice varieties gained relative stable seed setting rates when seeded during T1 to T5(May 10), very few varieties got significantly decreased seed setting rates when seeding date were delayed to May 16 and later.4. Be similar with panicle size, 1000-grain weight of these trial varieties showed no obvious regularity, and the average 1000-grain weight among all varieties of each seeding date showed no significant differences. Except for a few ones, most varieties seeded during T1 to T5 obtained stable 1000-grain weight. When seeded at T6 and later, 1000-grain weight of the some varieties decreased significantly. As for T7 and T8, most varieties possessed smaller 1000-grain weights comparing to the ones seeded during T1 to T5. In conclusion, delayed seeding could deteriorate grain filling greatly.5. The average grain yield of single rice plant seeded during T1 to T7 maintained similar, only the ones seeded at T8 produced significantly less grain yield per plant. With seeding date delayed. The average harvest index(HI) among all varieties of each seeding date decreased gradually, varieties under T4 gained significantly larger(smaller) HI than the other seeding dates.According to the grain yield difference per plant among each seeding date, all rice varieties were divided into three types: seeding date sensitive type, seeding date neutral type and seeding date stable type. Among the three types, rice varieties of seeding date neutral type could be seeded late in northeast hills of Sichuan to avoid drought stress. Basing on the grain yield comparison between T8 and other seeding dates, seven rice varieties which could produce almost the same amount of grains with super late seeding date(May 31) were chosen as the optimal ones to satisfy the demand of delayed seeding resulting from preceding crops.6. Correlation analysis of grain yield and its components between T1 and T8 showed that effective panicles, spikelets per panicle, filled spikelets per panicle and panicle weight all had positive correlation with grain yield per plant, though these relationships failed to be significant. With delayed seeding date, the above relationships got increasing close. Based on the variation extent of effective panicles, filled spikelets per panicle and 1000-grain weight, the 20 rice varieties could be classified into four types in accordance with their responses to seeding date.The results could also be regarded as an evaluation criteria for screening rice varieties suitable for late seeding and ruling matching cultivation techniques. in northeast hills of Sichuan.