Arran Rumbaugh,* Andrey Khlystov, Dave Campbell, Yeongkwon Son, Yan Wen, Chen Liang, Hao-Lin Fang, Christopher Wallis, and Anthony Wexler
*USDA-ARS, One Shields Avenue, Davis, CA, 95616 (arran.rumbaugh@usda.gov)

Smoke generated from wildfires can detrimentally affect grape and wine composition through increased concentrations of volatile phenols (VPs) and their glycoconjugates (bound), a phenomenon termed “smoke taint”. This study evaluated passive samplers as a method for monitoring the levels of VPs in ambient wildfire smoke that are known to contribute to smoke taint in wines. Cabernet Sauvignon grapes and passive samplers were exposed to smoke under controlled conditions to evaluate the efficacy of passive samplers across various smoke intensities. Smoke was generated from oak chips using a custom-built smoking chamber that consisted of: an infrared heating burner system, a polycarbonate chamber internally lined with Teflon sheets, and a ventilation system to operate the system under negative pressure and eliminate emissions. Grapes and samplers were exposed to four different smoke levels: control exposure (CE, no smoke), low exposure (LE, 0.5 hrs, ~20 g oak wood), medium exposure (ME, two hrs, ~60 g oak wood), and high exposure (HE, four hrs, ~180 g oak wood). Afterward, triplicate microfermentations (0.5 L) were performed per treatment. Passive samplers were analyzed using thermal desorption (TD) coupled to gas chromatography-mass spectrometry (GC-MS). Grapes and wines were analyzed for free and bound VPs using GC-MS/MS (free) and LC-MS/MS (bound). Free and bound VPs in grapes and wines correlated positively with VP levels in passive samplers. Linear regression analysis of free VPs indicated a good linear fit between passive samplers and grape samples for all compounds, with R2 values between 0.9999 and 0.9985 for guaiacol, methyl guaiacol, 4-ethyl guaiacol, and o-, p- and m-cresol. This study was the first step to determine the efficacy of using passive samplers as an early warning system in vineyards. Future experiments will correlate passive sampler VP levels to grape and wine composition across several smoke intensities, exposure durations, and repeat exposures to build a predictive model.

Funding Support: USDA-ARS funded research