A simple linear model for leaf area estimation in Persian walnut (Juglans regia L.)

Abstract

Accurate non-destructive leaf area estimation is a useful subject of study for the fields of applied plant science, physiology and plant genetic engineering. The relationship between leaf area and fruit is very important in fruit crops. It determines the nut size and nut filling potential. The aim of this paper is to produce a statistical model based on linear measurements such as leaflet length and width in combination with other simple parameters. Measurements of leaf length, width and surface area of the leaflets were achieved using digital photography. Digital images of Persian walnut (Juglans regia L.) leaves were prepared in August 2011 for 14 genotypes and in 2012 for one genotype under open field conditions. Results of regression analysis indicated several models are suitable to estimate leaf area of Persian walnut. The linear model with LW as an independent variable (LA=1.11+0.69LW) with accurate estimation (maximum R2=0.99 and lowest MSE=10.09) was the best model. To determine the leaf area, the earliest measurements to take would be the length and width of leaflets. Total area of the leaflets is then reported as leaf area. In conclusion, in walnut, number of leaflets and leaflet area should be involved in the estimation of leaf area.