Modeling Flowering Characteristics And Kernel Number And Optimizing Irrigation Scheduling For Improving Water Use Efficiency And Regulating Seed Quality In Hybrid Maize Seed Production

Hybrid maize seed production is an important method to produce high yield and stress tolerant maize seed in which both kernel quality and grain yield should be concerned.In arid and semiarid areas,irrigation is one of the most important factors affecting hybrid maize seed production.Thus optimizing irrigation scheduling can regulate the grain yield and kernel quality of maize inbreds for hybrid seed production.In this study,deficit irrigation experiments were conducted from 2014 to 2015,and deficit irrigation and pollination treatments were conducted in 2016 on maize inbreds for hybrid seed production in Shiyanghe experimental station of China Agricultural University.The reponses of evapotranspiration,flowering characteristics,biomass,yield and its components to water deficit of different growth stages were analyzed based on the experiments.Empirical statistical models were developed to simulate the water-flowering and water-biomass gain relations for maize inbreds,and a kernel number predicting model and a kernel weight predicting model were also established.Based on these models,optimal irrigation decision making method was evaluated.The main results of this study are as follows:(1)Water deficit at vegetative or flowering stage decreased silking rate of female population,silking rate per ear,total number of exposed silks per ear,seed-set capacity of the female inbreds,total pollen density at ear height,and delayed silking time of female population.Only water deficit at vegetative stage decreased Startshed rate,Maxshed rate and Endshed rate of Malel and Male2 populations for pollen shed,and delayed Startshed time,Maxshed time and Endshed time of Malel and Male2 populations for pollen shed.Only water deficit at flowering stage increased anthesis-silking interval.Water deficit decreased evapotranspiration of maize inbreds and decreased above-ground biomass and biomass gain post flowering of female inbreds.Water deficit at vegetative or flowering stage decreased kernel number and water deficit at grain-filling stage decreased kernel weight.The response of yield to water deficit was similar to kernel number.(2)The empirical statistical models describing the water-flowering characteristics and water-biomass gain post flowering relations were evaluated with experimental data of 2014 and 2015.Jensen model was recommended to simulate silking rate of female population,silking time of female population,silking rate per ear,total number of exposed silks per ear,seed-set capacity of the female inbreds and total pollen density at ear height by evapotranspiration at vegetative and flowering stage.The Startshed rate,Maxshed rate and Endshed rate,and Startshed time,Maxshed time and Endshed time of Malel and Male2 populations for pollen shed were simulated by evapotranspiration at vegetative stage using the single phase crop water production function.Jensen and Sterwart model were recommended to simulate biomass gain post flowering by evapotranspiration at vegetative,flowering,grain-filling and mature stage.(3)Flowering model,which can simulate kernel number of maize from flowering characteristics,was improved and its parameter was calibrated based on the 2 years' deficit irrigation experimental data.The relative root mean square errors of simulated kernel number and measured data fell to 0.1923 from 1.2812.Water-Flowering model,which coupled the water-flowering chacteristics with Flowering model,was developed.It performed well on the predicting of kernel number by evapotranspiration at vegetative and flowering stage,and the average predicting relative root mean square errors of kernel number was 0.2216.(4)The relationships between the upper and lower bound of kernel weight and source-sink ratio was established based on the exerimental data of 2016.Coupled the relationships with water-biomass gain post flowering model and Water-Flowering model,ULKW model was developed to predicte kernel weight upper and lower bound by evapotranspiration at vegetative,flowering,grain-filling and mature stages.The measured kernel weights of different irrigation treatments in 2014 and 2015 were in the range of predicting upper and lower bound of kernel weight.(5)An optimal irrigation decision making method for maize inbreds was developed based on Water-Flowering model and ULKW model to maximize kernel number and minimize irrigation amount under the kernel weight constraints.Optimal irrigation schedulings were achieved by solving the model under different cases,providing scientific guidances for the water-saving irrigation management in the production of high quality hybrid seed.