The objective of this study was to determine an empirical mathematic model for estimating leaf area of different soybean cultivars from leaf dimensions, with emphasis to modern and currently used soybean cultivars used in Rio Grande do Sul. A field experiment conducted at Santa Maria during three growing seasons (2010/2011, 2012/2013 e 2013/2014) and 13 soybean cultivars (NS 4823 RR, BMX Energia RR, Igra RA 518 RR, BMX Turbo RR, NA 5909 RG, TMG 7161 RR Inox, TEC 5936IPRO, IAS 5, BMX Potência RR, Fepagro 36 RR, BRS 246 RR, Bragg e CD 219 RR) was used. Twenty leaves of each cultivar were used for calibrating the model and another 50 leaves were used for testing the predictive capacity of the model as independent data. The model is based on the length (L) and the greatest width (W) of the central foliole of trifoliolate leaf. The relationship between leaf area and the linear dimensions of the central foliole was fitted to a linear model for each cultivar as well as to a generalized model for all cultivars. The predictive capacity of the equations was evaluated with the statistics root mean square error, BIAS, modified index of agreement and coefficient r. The results indicate that the non destructive method, based on linear dimensions of the central foliole is appropriate for estimating the area of individual leaves in soybean and that the general equation LA = 2,0185.(L.W) can be used for several modern soybean cultivars.
Glycine max; leaf area index; mathematics models; linear dimensions
