玉米穗粒数形成过程模型研究

Study on the Formative Process Model of Ear Kernel Number in Maize

采用均匀设计方法,通过播期、密度、水肥管理等措施创造不同的玉米生长状态,在动态调查玉米关键生育时期生长及穗粒形成的基础上,建立玉米穗粒数与干物质积累、出苗-吐丝持续时间、比茎重(单位茎长的干物质重)的关系模型,模型模拟的相对误差(RE)为0.08%,相关系数(R)为0.857,模拟值与观测值吻合度高,拟合效果显著。模型预测效果的t测验表明,预测值与观测值之间没有显著性差异,方程可以很好地预测玉米穗粒数。通过对玉米穗粒数形成过程的分析表明,模型变量基本反映了玉米穗粒数决定过程涉及的因素,从统计学及生物学方面均能较好地解释玉米穗粒数的形成。本研究也建立了玉米比茎重与种植密度的关系模型,两者在9叶期、12叶期和吐丝期均呈负相关关系,模型的拟合及预测效果较好。

A multi-factor experiment including sowing date,plant density,irrigation,and nitrogen,phosphors and potassium fertilizer by uniform design method was conducted in 2008 and 2009,the data of maize growth was investigated,and the EKN formation model was established.The relative error（RE） of the model was 0.08%,correlation coefficient（R） was 0.857,the fitting result between simulated value and observed value was significant.Test for the predicting effect showed no significant difference between the predict value and the observed value,the equation could be used to predict the process of maize EKN formation.The model included the basic factors involved in the process of maize EKN formation,such as dry matter accumulation,increment of dry matter,duration of seedling to silking,and specific stem weight,could explain this process in the aspect of statistics and biology.A model between specific stem weight and plant density also was established.Negative correlations were observed between the two factors at maize 9-leaf expanded,12-leaf expanded and silking stages.Fitting effect and predicting effect of the model were both well.