Leaf area is important in determining the growth and development of agricultural crops. The aim of this study was to compare the methods of leaf discs and digital photos in estimating leaf area of Crambe abyssinica, and model leaf area according to length (C), width (L) and/ or the product of length width (CxL) for different sizes of leaves. For this, in 308 leaves it was determined the leaf area, length, width and the product of length width using the methods of leaf discs and digital photos. Then the methods were compared using the linear correlation coefficient between the leaf areas. Then, for each method, leaf area (Y) depending on the C, of L and the CxL, was modeled through these models: linear, simple linear, quadratic, geometric and exponential. The Pearson and Spearman linear correlation coefficient between the methods of leaf area and leaf discs of digital photos were 0.9917 and 0.9889, respectively, which shows concordant methods. In both methods, the geometric and quadratic models showed good coefficients of determination of leaf area depending on the length and width of leaves. The leaf width is the variable that best estimates the leaf area. The method of digital photos can be used to estimate Crambe's leaf area.
modeling; non-destructive method; regression
