The objective of the present study was to develop a reliable equation which enables a fast and non-destructive calculation of leaf area using linear leaf measures from different Passiflora species (P. alata, P. coccinea, P. gibertii, P. ligularis, P. misera, P. mucronata, P. nitida, P. setacea). Three hundred leaves with different sizes were collected from each species. It was performed the analysis of regression of leaf area versus lenght of the main vein, greatest leaf width and their product. Leaf area was measured with a device for automatic measurement and linear measures were determined using a ruler. The linear model chosen for all species, was the one which uses independent variables product because it presents the greatest coefficient of determination (R²) and significative coefficient of regression. The models showed adjusted coefficient of determination higher than 0.927 and optimal performance index, according to the classification of Camargo and Sentelhas. In the process of model validation, it was shown that the correlation between measured leaf area and observed leaf area was very high. The results of this study demostrate that leaf area of Passiflora species can be predicted, using the measurements of the length and width of leaves.
passion fruit; linear measures; Passifloracea; model validation
