Prediction of Leaf Area Using Montgomery Models in Ramie

Abstract

The Montgomery model has been proved to be suitable for the leaf area estimation, that is, A = cm1 ×L ×W. However, the same plant affected by different genes causes the variation of leaf shape. The effects of allometric growth and variation of ramie leaf on the parameter (cm1) of Montgomery model still needs further exploration. In this study, a total of 3020 leaves were taken from 151 varieties in the ramie germplasm resource nursery (20 leaves/variety). Based on the root mean square error (RMSE) and the parameter (cm1) variation coefficient, six mathematical models of leaf area were compared. The results show that the Montgomery model is the optimum support model. It is also found that the parameter (cm1) of Montgomery model of ramie leaf with different genes varies greatly, which ranges from 0.5633~0.6621; the morphological variation of ramie leaf can be described by the change in oval parameters (a, b, c) and the length of leaf opex, and partially explain the change in the parameters of Montgomery mod-el. Therefore, in order to improve the accuracy of leaf area estimation of Montgomery model, the allometric growth and variation of ramie leaf with different genes should be considered.