The objective of this study was to adjust and to test simple, precise, and accurate mathematical models for estimating the leaf area of 'Royal Gala' and 'Fuji Suprema' apple trees grown under coverage of black hail net and in open sky. The experiment was carried out in a commercial apple orchard, in Vacaria, Brazil, during the vegetative cycles of 2008/2009 and 2009/2010. Linear and quadratic regression equations were adjusted for estimating the leaf area in function of leaf dimensions. The selection for best models was based on mean square error, determination coefficient, standard error of estimative, and significance of equation coefficients. The leaf area estimated by the best models was compared with data obtained by electronic planimeter and estimated by digital photography. The prove of models was based on the significance of the angular coefficient, on the 1:1 line, and on the root mean square error. Linear and quadratic models that consider the length and width of leaves estimated accurately and precisely the leaf area of 'Royal Gala' and 'Fuji Suprema' apple trees for both cropping conditions, open sky or under hail net. The apple trees leaf area estimated by regression models using leaf length and width is similar to those obtained by planimetry or digital photography.
Malus domestica; regression models; leaf; length; width
