BotRisk: simulating the annual bunch rot risk on grapevines (Vitis vinifera L. cv. Riesling) based on meteorological data

Abstract

The aim of the present investigations was to simulate the annual risk of bunch rot (Botrytis cinerea) on Vitis vinifera L. cv. Riesling grapes based on three long-term (n= 3 × 7 = 21 cases) assessment data sets originating from three Central European grape-growing regions. Periods when meteorological parameters were significantly (p< 0.01) correlated with the cumulative degree day (CDD7;18;24) reaching 5% disease severity were determined by Window Pane analysis. Analyses revealed five critical weather constellations ("events") influencing annual epidemics: relatively low temperatures after bud break, dry conditions during flowering, high temperatures after flowering, and low temperatures and high precipitation sums during/after veraison were all associated with thermal-temporal early epidemics. Meteorological data in each of the five events served as input for the bunch rot risk model "BotRisk." The multiple linear regression model resulted in an adjusted coefficient of determination (R2adj.) of 0.63. BotRisk enables (i) the simulation of the thermal-temporal position of the annual epidemic and, based on this, (ii) the classification of the annual bunch rot risk into three classes: low, medium, or high risk. According to leave-one-out cross-validation, 11 of 21 case studies were correctly classified. No systematic bias caused by location was observed, indicating that the transfer of the model into other locations with comparable climatic conditions could be possible. BotRisk (i) represents a novel viticultural decision support tool for crop cultural and chemical measures against bunch rot and (ii) enables an estimation of the bunch rot risk under changing environmental conditions.