橡胶草叶面积及鲜重无损估计模型及它们的取样数量

he aim of this study was to determine nondestructive models for estimating leaf area and fresh weight of Russian dandelion plants. Regression analyses were performed between leaf area, fresh weight, leaf length, and leaf width in two hundred and fifty leaf samples collected during different growth stages of Russian dandelion plants. Data from another fifty leaves were used for validating the proposed models. Regression analyses were performed among ten data groups with different numbers of data randomly selected from the total three hundred leaves data set to determine the smallest sampling number for applying the final models correctly. The model for estimating leaf area(LA)is: LA=6226.424+26.31L+545.334W－313.993L0.5－3138.047W0.5－0.009L2－3.86W2+0.057LW, with R2 and RMSE values of 0.818 and 168.29 respectively. The model for estimating leaf fresh weight(FW) is: FW=1125.572－24.857L+233.070W+0.055LW+276.956L0.5－1264.466W0.5+0.067L2－1.964W2, with R2 and RMSE values of 0.735 and 87.84 respectively. At least ten leaf samples are required when applying the two models. Determining transformed forms of leaf dimensions that are linearly related to leaf area and fresh weight, and integrating all of them into one equation maybe a better solution for establishing models to estimate leaf area and fresh weight of plant species, particularly those with higher variation among individual leaves.In this paper.