Detection and prediction of Botrytis cinerea infection levels in wine grapes using volatile analysis

•Predictive models developed for B. cinerea infection levels from volatile profiles.•Predicted infection levels assessed by measuring ergosterol and antigen detection.•Key markers of infection (trans-2-octen-1-ol, 3-octanol, 1,5-dimethyltetralin)•Volatile metabolites tracked during the early stages of B. cinerea infection. Infection of grape berries (Vitis vinifera) by the fungus Botrytis cinerea (grey mould) frequently occurs in vineyards, resulting in off-flavours and other odours in wine and potential yield losses. In this study, volatile profiles of four naturally infected grape cultivars, and laboratory-infected grapes were analysed to identify potential markers for B. cinerea infection. Selected volatile organic compounds (VOCs) were highly correlated with two independent measures of B. cinerea infection levels, demonstrating that ergosterol measurements provide accurate quantification of lab-inoculated samples, while B. cinerea antigen detection is more suitable for naturally infected grapes. Excellent predictive models of infection level were confirmed (Q2Y of 0.784–0.959) using selected VOCs. A time course experiment confirmed that selected VOCs 1,5-dimethyltetralin, 1,5-dimethylnaphthalene, phenylethyl alcohol and 3-octanol are good markers for B. cinerea quantification and 2-octen-1-ol could be considered as an early marker of the infection.