1H NMR and LC-MS-based metabolomic approach for evaluation of the seasonality and viticultural practices in wines from São Francisco River Valley, a Brazilian semi-arid region

Abstract

São Francisco River Valley (SFRV) is a wine-producing semi-arid region in Brazil. Therefore, we used a 1H NMR and UPLC-MS-based metabolomic approach coupled to chemometrics to evaluate the variability in Chenin Blanc and Syrah wines for two harvest seasons, two vine training system and six rootstocks. Overall, the secondary metabolites were influenced by the three factors studied, whereas the primary metabolites were only by the seasonality. Chenin Blanc wines made in December presented higher content of an unidentified carbohydrate. In Syrah wines, glycerol, tartaric acid, succinic acid and 2,3-butanediol were greater in December, while proline and lactic acid were more abundant in July. For training system, caffeic acid derivatives were increased in wines produced from espalier. Lyre system increased phenolic compounds, organic acids and apocarotenoids. The effect of the rootstocks was less pronounced, affecting basically caffeic acid derivatives. Thus, we expect that our results may assist the winemakers to improve the SFRV wine quality.