Abstract: 【Objective】The grain filling rate and grain filling duration are major determinants of grain yield. Examination of the contributions of growth period and grain filling related traits to kernel weight, and study of quantitative trait loci (QTL) for grain filling rate and duration to dissect the genetic basis of grain filling are helpful to the practice for the breeding of high yield in maize.【Method】To identify QTL for grain filling rate and duration, a recombinant inbred line (RIL) population including 172 families was developed from the cross between Huangzaosi (HZS) and Lv28 which are foundation parents used in maize breeding of China. Firstly, the differences between the Logistic and the Richards models for the fitting of maize grain filling process were compared and the parameters of the grain filling were calculated. Secondly, the differences between HZS and Lv28 in grain filling characteristic were compared. Correlation analysis and regression analysis were applied to elucidate the relationship among growth period and grain filling related traits and the contribution to hundred-kernel weight (HKW). Thirdly, the approach of genotyping-by-sequencing (GBS) was used to detect polymorphic SNP markers between the parents and among the RILs. Finally, the inclusive composite interval mapping (ICIM) was used to identify QTL of growth period and grain filling related traits.【Result】The genetic map was constructed with 1 471 filtered SNP markers, the total length was 1 471 cM with the average length 1 cM. The grain filling process tended to be a slow-fast-slow pattern and could be divided into three phases: the lag phase, the effective grain filling phase and the maturation drying phase. The use of model fitting of the grain filling process could reflect the dynamic changes of grain filling, but the Richards model and the Logistic model were different in simulating the grain filling dynamic process. The r² measure of predicted values obtained by the Richards model with phenotypic values was higher than that by the Logistic model. The grain filling characteristics showed significant differences between the two parents, since the grain filling rate of HZS was 1.28 times higher than that of Lv28, while the grain filling duration of Lv28 was 1.07 times longer than that of HZS. Correlation analysis showed that the grain filling related traits and HKW reached a significant positive correlation except for the grain filling duration in the lag phase. Regression analysis showed that the grain filling rate and the grain filling duration could explain 57.50% and 30.00% of the phenotypic variation of HKW, respectively. The effective grain filling phase played an important role in the formation of HKW. A total of 26 QTL were detected for grain filling related traits, 3 QTL were detected for HKW and 14 QTL were detected for growth period related traits under the single environment, which were distributed on the chromosomes 1, 2, 3, 4, 5, 6, 8, 9 and 10. The LOD values ranged from 3.27 to 9.05, and the range of phenotypic variation explained was 5.97%-21.16%. Under the joint environments, the QTL controlling the grain filling related traits were located at the same or similar positions of chromosomes and formed some QTL clustering regions on bin 1.05, bin 2.03, bin 4.05, bin 4.06, bin 7.04 and bin 9.04. Importantly, some QTL for grain filling rate in different grain filling phases were detected to be clustered on the region of 48.24 cM-135.72 cM on chromosome 4 and the region of 110.10 cM-114.73 cM on chromosome 9. 【Conclusion】 Maize grain filling dynamic process could be simulated well with the Richards model. The grain filling rate and duration of the effective grain filling phase played an important role in the formation of HKW. The inheritance of grain filling related traits were strongly influenced by the environments, so some QTL could be detected only in one environment. Under the joint environments, some QTL related to grain filling rate were detected on bin 4.05 and bin 9.04.

Key words: maize (Zea mays L.), grain filling rate, grain filling duration, QTL