AROMA COMPOUNDS PREVISION USING ARTIFICIAL NEURAL NETWORKS: INFLUENCE OF NEWLY INDIGENOUS SACCHAROMYCES SPP. IN WHITE WINE PRODUCED WITH VITIS VINIFERA CV SIRIA

摘要

Commercial yeasts strains of Saccharomyces cerevisae are frequently used in white wine production as starters in fermentation process, however, these strains can affect the wine characteristics. The aim of this study was to evaluate the effect of three strains of Saccharomyces spp. (var. 1, 2 and 3) on wine aroma compounds produced in microvinification assays. Microvinification assays were carried out with Vitis vinifera cv Siria grapes using the strains in study as starters. Aroma compounds were identified and quantified by GC-FID and GC-MS. At the end of fermentation process and during the first three months of maturation some aroma compounds were detected, namely propanol, isobutanol, isoamyl acetate, isoamylic alcohol, ethyl hexanoate, ethyl lactate, hexanol, ethyl octanoate, 3-ethylhydroxibutirate, benzaldehyde, 3--methyl-2-butanol, 2,3-butanediol, γ-butyrolactone, ethyl decanoate, diethyl succinate, methionol, 4-hydroxi-2--butyrolactone, heptanoic acid, phenylethyl acetate , ethyl dodecanoate, phenylethanol, octanoic acid, 2-methoxy-4--vinylphenol and decanoic acid. Artificial Neural Networks (ANNs) were used to predict the concentration of twelve wine aroma compounds from the phenyl ethanol, propanol, isobutanol, hexanol, heptanoic acid, octanoic acid, and decanoic acid concentrations. Results showed that, either, maturation time and Saccharomyces strain used as starter influence the aroma compounds produced. Wines produced with S. cerevisae var. 1 and 5. cerevisae var. 2 showed a similar composition in aroma compounds, relatively to the wines produced with the strain S. cerevisae var. 3. However, for S. cerevisae var. 1 and S. cerevisae var. 2 the time of maturation influence the aroma composition of wines. From a technological approach, the choice of yeast strain and maturation time has decisive influence on the aroma compounds produced.