Validation of an empirical model for grapevine leaf area estimation with data from the varieties ‘Cannonau’ and ‘Vermentino’ grown in Sardinia

• Main Author: Oliveira, A.F.
• Publication Date: 2009
• Other Authors: Nieddu, G., Lopes, C.M.
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10400.5/23488
• topic: grapevine; leaf area; statistical model; Tempranillo; Cannonau; Vermentino

The performance of two mathematical models for non-destructive estimation of primary and lateral leaf area per shoot of Tempranillo grapevines was tested in Sardinia using independent datasets from two main traditional varieties. One collected on Cannonau grapevines from Nurra wine region, and another from Vermentino grapevines grown in Gallura wine region. The models presented good fit between observed and estimated values with high modeling efficiency. For primary leaf area estimation the mean absolute percent error for both varieties was lower than 10%. Both linear regressions between observed and estimated primary leaf area had high and significant R2 but while Vermentino fitted line presented a slope not significantly different from 1, Cannonau fitted line showed a slope significantly < 1, indicating that the model overestimated the primary leaf area per shoot. The validation of the model for lateral leaf area presented lower goodness of fit as that reported for primary leaf area. Linear regressions had a very high and significant R2 but the slopes were significantly <1 indicating that the model overestimated lateral leaf area per shoot. The positive validation shows that these models can accurately predict leaf area per shoot independently of ecological conditions, variety, year, growth stage and training system. Low goodness of fit for lateral leaf area model may be avoided building the model on each variety data. The generalized use of this type of model represents a powerful tool for grapevine research, for consultants and advanced growers, allowing the evaluate vine leaf area more frequently and with low cost