It is possible to predict Sangiovese wine quality through a limited number of variables measured on the vines

Abstract

<p style="text-align: justify;"><strong>Aims</strong>: The research work aimed at creating and testing a method to evaluate vine performance of Sangiovese (VPS), in particular, a method able to predict the potential oenological result through a limited number of variables measured on the vines.</p><p style="text-align: justify;"><strong>Methods and results</strong>: A matching table was created on the basis of literature and the experience acquired over twenty years of research activity on Sangiovese vine and wine quality in Tuscany, which allowed the selection of eight viticultural parameters and three VPS classes. In order to validate the matching table, a specific experiment was conducted during the years 2002 and 2003 in 10 vineyards (selected from 7 farms) representative of the main soils and climates of the vine cultivation areas of the Province of Siena (Italy). The experimental results validated the proposed matching table through a non parametric statistical analysis. A multivariate regression analysis between wine sensory evaluation (score) and viticultural parameters significantly predicted wine quality even with only 4 grape parameters (<em>P</em> < 0.05).</p><p style="text-align: justify;"><strong>Conclusion</strong>: It was possible to predict VPS by means of a matching table based upon eight simple viticultural parameters. The reliability of the wine quality prediction increased proportionally according to the number of viticultural parameters, but remained rather high (R<sup>2</sup> = 0.606) when taking into account only sugar content, sugar accumulation rate, mean berry weight, and extractable polyphenol index (EPI).</p><p style="text-align: justify;"><strong>Significance and impact of the study</strong>: It is now possible to predict the quality of Sangiovese wines with a few selected grape parameters. Because of the wide variability in soil and climatic condition of the viticultural areas of the Province of Siena, where the method was developed, and the strong climatic contrast between the years when the method was validated, the use of both matching table and multiple regression is recommended for VPS prediction in Mediterranean environments.</p>