摘要
以云南特色香蕉品种红研一号、红研二号、红研三号、红研五号和滇蕉一号5个品种为供试材料,探讨了叶长、叶宽与实测叶面积之间的关系,运用植物图像分析仪测量一造蕉、二造蕉叶面积,分析了叶长、宽、及长宽乘积与叶面积的相关性,并建立一元线性、一元二次、一元幂曲线方程模型来估测香蕉叶面积。结果表明:一造蕉、二造蕉的一元线性方程和一元幂曲线方程的系数极显著(P<0.01),拟合数据最好。进一步检验并比较决定系数(R^2)、回归标准误差(Se)和模型系数表明,二造蕉叶面积回归模型中以长宽乘积拟合的模型可靠性较以叶长叶宽单独拟合的模型可靠性高。鉴于计算的简便性,在科研和生产的运用中建议使用一元线性回归模型来估算云南特色香蕉品种的叶面积,即一造蕉叶面积模型为A=0.805 P-171.33,二造蕉叶面积模型为A=0.836 P-1073.421。
In this study,the yunnan characteristic bananas including Hongyan 1,Hongyan 2,Hongyan 3,Hongyan 5 and Dianjiao 1 were used as experimental materials,and the relationship among leaf length,leaf width and measured leaf area was explored,By plant image analyzer measurement of leaf area,the correlation among leaf length,width,the product of length width and leaf area was analyzed,and the equation of linear,quadratic and power curve of one variable was established to estimate the leaf area of bananas.The results showed that coefficients of linear equation and power curve equation of the first and second generations of the banana’s leaf area were extremely significant(P<0.01).Fitting data was the best.Further test and comparison of determination coefficient R^2,regression standard error Se and model coefficient showed that the model of length and width product fitting was more reliable than that fitted separately with leaf length and width.In view of calculation convenience,it was suggested to use a linear regression model to estimate the Yunnan characteristic banana’s leaf area in the application of scientific research and production.In other words,the model of the first generation banana’s leaf area was A=0.805 P-171.33,and the model of the second generation banana’s leaf area was A=0.836 P-1073.421.
作者
只佳增
杜浩
周劲松
李宗锴
杨绍琼
张光勇
陈伟强
ZHI Jia-zeng;DU Hao;ZHOU Jin-song;LI Zong-kai;YANG Shao-qiong;ZHANG Guang-yong;CHEN Wei-qiang(Honghe Research Institute of Tropical Agriculture in Yunnan Province,Hekou 661300,China)
出处
《江西农业学报》
CAS
2020年第10期57-61,共5页
Acta Agriculturae Jiangxi
基金
云南省科研院所技术开发专项“云南特色热带香蕉新品种选育研究与应用”(NO.2019DC011)。
关键词
云南特色
香蕉
叶面积
回归模型
Yunnan characteristic
Banana
Leaf area
Regression model
