科研之友
微信
新浪微博
Facebook
分享链接
摘要
Research in fruit tree physiology and breeding often requires accurate and non-destructive methods for estimating leaf area (LA). The development of unbiased allometric model from linear measurements [leaf length (L) and/or width MO] to predict individual LA of apricot irrespective of cultivars is still lacking. The models were built using LA, L, and W data measured in 3,040 leaves collected on trees of nineteen apricot cultivars (calibration experiment). Model(s) were validated on 520 apricot leaves collected from the trees of two additional cultivars (validation experiment). LA prediction models based only on L measurements (L or L-2) were not suitable for estimating LA of apricot. A significant improvement in LA prediction was observed when the model including W-2 as an independent variable was adopted. However, the coefficients of one dimension LA model (W2) were affected by leaf shape (L:W ratio) and consequently were excluded. To develop an accurate LA model for apricot, independent of leaf shape groups, the product L x W was used as an independent variable. The linear model LA = 1.193 + 0.668 (L x W) exhibited the highest R-2, the smallest mean square error (MSE) and predicted residual error sum of squares (PRESS). In the model validation, correlation coefficients showed that there was a highly reliable relationship between the predicted and the observed LA values, giving an underestimation of 2.9% in the prediction. The LA model using LW as independent variable can be successfully adopted in research on apricot, since it provides an accurate, simple and non-destructive estimation of LA across apricot cultivars without the use of any expensive device.
- 出版日期2017-4
