Data-driven Determination of Disease Markers’ Threshold Values in Rot-affected Wine Grapes

Abstract

Grapevine bunch rot is detrimental to grape and wine quality. Traditionally, detecting and quantifyingthe severity of rot infection is executed visually. This study aimed at defining local, area-specific thresholdvalues of rot-associated disease markers. This is a first step towards making informed decisions aboutthe quality of grapes delivered at winery intake. Viticulturists visually assessed on-vine rot infection innine white wine grape cultivars. Results showed that severity ratings were consistent between assessors.Chemical analyses of the grape must from these assessed samples were done. Multiple factor analysis (MFA)showed that rot severity was positively correlated with glycerol, alcohol, gluconic acid and acetic acidconcentrations. As severity increased, gluconic acid, glycerol, alcohol, Brix, acetic acid and total titratableacidity (TA) concentrations also increased. Following the probability chosen for sensitivity and specificity,grape rot indicators’ threshold values in white grape must are as follow: Alcohol ≥ 0.10 %v/v; acetic acid≥ 0.17 g/L; glycerol ≥ 0.79 g/L; gluconic acid ≥ 0.99 g/L; TA ≥ 8.86 g/L. Statistical determined thresholdvalues differentiating between rot-affected and healthy grape must, would eliminate the subjectivity andbias associated with visual assessments.