Evaluation of vineyards, fruit and wine affected by wild fire smoke

Abstract

Context and purpose of study ‐ Wineries may randomly reject fruit from vineyards near wild fires exposed to smoke. It is difficult to determine if fruit has been compromised in quality when exposed to smoke, and whether or not smoke taint flavors will result when fruit is fermented into wine. Phenolic smoke compounds bind with sugars in the fruit with enzymes (glycosyltransferases) and are then hydrolyzed during maturation, wine making and even in a taster’s mouth. Testing the fruit for volatile phenols and glycosides is both expensive and not completely predictive as standards are not well defined for damage based on smoke chemical content. Micro‐vinification even with partially ripened fruit is an inexpensive and fairly accurate method to quickly determine if fruit has a potential smoke taint problem. Wines can then be tasted for the presence of off flavors. Developing standards based on volatile phenolic and glycocide concentrations to predict whether fruit is affected by smoke and how wine will taste when vinified would be very helpful for accepting or rejecting fruit from affected areas.

Materials and methods ‐ Following wild fire smoke exposure, fruit was sampled and micro‐vinified during veraison and again 2 weeks before harvest from 13 Cabernet sauvignon vineyards in a transect 25 km across Lake County, California. A control vineyard unexposed to wildfire smoke was sampled outside of the area. Sub samples from each vineyard were analyzed immediately for guaiacol and 4‐methyl guaiacol. 19 liter wine lots were then microvinified, stabilized and bottled for each vineyard for both sampling dates. The wine was analyzed for volatile phenols and glycoside compounds (guaiacol and 4‐methyl guaiacol, methyl cresol, 4‐methyl syringol, o‐cresol, p‐cresol, syringol, syringol gentiobioside, methyl syringol gentiobioside, phenol rutinoside, cresol rutinoside, guiaocol rutinoside and methyl guaiacol rutinoside). A 14 member tasting panel evaluated the wines for smoke flavors. Panel members were able to detect off flavors in both sample sets, and tainted wines were highly correlated with elevated concentrations of volatile phenols and glycosides. GIS data of vineyard proximity to the fire, elevation, temperature and wind direction and speed were used to conduct multivariate analysis of factors affecting wine smoke compound chemicals and flavor impacts on wine.

Results ‐ Not all wines were affected; in this study, 6µ/l guaiacol was the threshold of detection for off flavors in wine by most tasters. Off flavors were much stronger in the wines made from riper fruit, as were the concentration of smoke compounds, by as much as six fold compared to unfermented fruit. Wind direction and speed, proximity to active fires, and temperature are the factors that are most highly correlated to smoke damage to fruit near wildfires. The control wine sample had no off flavors and no volatile phenols were detected. By contrast, some sites close to the edge of fires and immediately downwind were very heavily affected, and contained high levels of smoke taint compounds. This study will help to better understand when vineyards are most at risk to wild fire smoke damage, and how micro‐vinification may be a reliable and quick way to predict fermentation outcomes before harvest in vineyards affected by wildfire smoke.