于2013年和2014年在天津、安徽、海南、黑龙江、新疆和内蒙古试验,应用RBF网络模型、逐步回归等多元统计方法,从区域的角度阐明甜高粱产量及品质相关性状与环境因子之间的关系,明确影响目标性状的关键环境因子,量化关键环境因子的作用程度,剖析高产优质甜高粱对环境因子的要求并构建产量预测模型。结果表明,环境(E)效应对甜高粱性状的影响很大,大于基因型(G)和G×E互作效应;甜高粱产量及品质相关性状对降水量、昼夜温差、土壤pH值、有机质、全氮和有效钾反应度较大;土壤pH值、有机质和全氮为关键影响因子;鲜重产量和茎秆相对总含糖量预测模型均通过显著性检验,决定系数R2值较高,为可用模型。研究结果为中国甜高粱生态划区、栽培调控提供了理论依据,使高产、优质得以充分的发挥。
The present study aims to elucidate the effects of environmental factors on the traits related to yield and quality in sweet sorghum, to quantify the responsive degree of sweet sorghum traits to the key factors, to analyze the requirements of envi-ronmental factors for high yield and quality, and to construct prediction models of yield. Based on the data of five traits and envi-ronmental factors obtained from six provinces of China in 2013 and 2014, we studied the responsiveness of sweet sorghum in five traits to environmental factors by the methods of RBF analysis and stepwise regression analysis. The results demonstrated that the environment was the dominant factor affecting the sweet sorghum traits, with greater effects than genotype and G×E interaction. Soil factors exerted more significant effects on sweet sorghum traits than meteorological factors. Precipitation, diurnal tempera-ture difference, soil pH, organic matter, total nitrogen, and available potassium were important environmental factors determining fresh yield, sugar brix, juice yield, sugar yield and stem diameter in sweet sorghum. Organic matter, pH value and total nitrogen content of soil were the three most important environmental factors. The prediction models for fresh weight yield and total sugar yield were verified by significant test with high determination coefficientR2. The paper provides an accurate and reliable theo-retical basis for ecological region division, planting management, high yield and good quality in sweet sorghum.