Logistic modeling of summer expression of esca symptoms in tolerant and susceptible cultivars in Bordeaux vineyards

Abstract

The seasonal dynamics of esca leaf symptom development were monitored and modelled over 10 years (from 2004 to 2006, 2012 to 2014, and 2018 to 2021) in eleven vineyards near Bordeaux (France) and on five cultivars, including three susceptible and two tolerant Field observations performed once or twice a week from the end of May to mid-September confirmed i) the evolution over time of esca leaf symptoms, ii) the presence under the bark of a discolored xylem longitudinal stripe with nonfunctional vessels, and iii) a gradual increase in the number of symptomatic plants within each vineyard. Of the three models tested, nonlinear logistic regression was the best fitting curve, showing a clear and systematic progressive sigmoidal pattern of cumulative esca leaf symptom observations regardless of ‘vineyard*year’ situation. Relationships with climatic data confirmed that all periods of symptom expression corresponded to the warmest and driest period of each vegetative season. Examinations of key dates corresponding to four threshold levels of cumulative incidence of leaf symptomatic vines [S1 (first observed symptoms), S10 %, S50 % and S90 %] showed that tolerant cultivars (Merlot noir and Malbec) generally developed leaf symptoms later than susceptible cultivars (Cabernet-Sauvignon, Cabernet franc, and Sauvignon blanc). A variance analysis and a principal component analysis (PCA) confirmed that compared to susceptible cultivars, tolerant cultivars were associated with increased temperature sums above 10 °C from 1st January, reaching the same symptom thresholds S1 and S10 % and with more cumulative rainfall at the S1 stage. Overall, this study reveals the key role of temperature as a triggering factor for esca symptom expression in relation to fungal activity. The results indicate that the S10 % stage can be used as a discriminant variable to separate cultivars according to their susceptibility. Finally, logistic modelling can be used as a descriptive and analytical tool to study the seasonal dynamics of esca.