Leaf area estimation of canola by leaf dimensions

The objective of this work was to model and identify the best models to estimate the leaf area determined by digital photos, of three canola hybrid, with the length or width and / or the product length width of the leaf. Three uniformity trials were carried with the culture of canola (Brassica napus L.). In each trial was valued one of the following hybrids: Hyola 61, Hyola 76, Hyola 433. In each hybrid were collected 125 leaves at 77, 84, 91, 97 days after sowing, totaling 1,500 leaves. In these 1,500 leaves were measured length (C) and width (L) and calculated the length width (C×L) of the leaf. Was determined the leaf area of each leaf, by the method of digital photos (Y). After, for each hybrid were separated, randomly, 80% of the leaves (100 leaves by collects × 4 collects by hybrid = 400 leaves per hybrid) to build models of quadratic, potency and linear type for Y function of the C, L and/or C×L. The remaining 20% of the leaves (100 leaves by hybrid), separately, were used to validate the models. In canola, the potency model for hybrid Hyola 61 (Ŷ = 1.3000x^1.9336, R² = 0.9531), Hyola 76 (Ŷ = 1.0579x^2.0383, R² = 0.9733) and Hyola 433 (Ŷ = 1.4154x^1.9096, R² = 0.9613), are adequate for estimation of leaf area determined by digital photos (Y) as function of the width of the leaf (x).

Brassica napus L.; digital photos; modeling; non destructive method