Elliptic Fourier Analysis (EFA) and Artificial Neural Networks (ANNs) for the identification of grapevine (<i>Vitis vinifera</i> L.) genotypes

  • S. Mancuso
Keywords: ampelography, backpropagation, cultivar identification, EFA, Vitis vinifera


The potential application of the Elliptic Fourier Analysis (EFA) for the objective quantitative description of leaf morphology, combined with the use of a Back-propagation Neural Network (BPNN) for data modelling, was evaluated to characterize and identifiy 12 Sangiovese-related accessions (Vitis vinifera L.). The results enable us to distinguish, with considerable certainty, between 10 accessions. Cluster analysis revealed the existence of a uniform group for the Prugnolo (acerbo, medio and dolce) ecotypes showing a high degree of relatedness. Among all accessions only the so-called Casentino ecotype significantly diverged from all the others, indicating probably a different origin. The application of EFA coupled with the use of artificial neural networks opens interesting prospects for the characterization of varieties, allowing to study differences and/or relationships which can not be detected by standard ampelographic systems.