Utilizing Disease Prediction Models to Time Fungicide Applications for Controlling Ripe Rot, Caused by Colletotrichum spp., in Maryland Vineyards.

ABSTRACT

Two previously published ripe rot prediction models, DF2-NN and GH2-DT, were evaluated for fungicide application timing efficacy in Maryland vineyards. Both models utilize leaf wetness duration (LWD), temperature, and grape cluster phenological stages as model parameters. These three parameters were tracked throughout the 2021 to 2023 seasons in three vineyards. The fungicide efficacy trials started at the veraison phenological stage and included a nontreated control, a 12-day interval treatment, and two model-triggered treatments when risk predicted by the models crossed a threshold. The severity of ripe rot on the clusters in each treatment was assessed when the fruit were mature. Ripe rot severity in the nontreated controls was higher during seasons with more LWD and more precipitation. Days in which the models were triggered by the environmental conditions primarily coincided with precipitation events and lengthy LWDs. The model-triggered treatments never had significantly higher ripe rot severity than the 12-day interval treatment but had significantly lower severities than the nontreated control in most trials which had high ripe rot pressure. Furthermore, the model-triggered treatments resulted in fewer fungicide applications than the 12-day interval treatment on average. The DF2-NN model appeared to be more accurate and useful for ripe rot prediction and treatment than the GH2-DT model because it triggered fewer fungicide applications while reducing ripe rot. This model may be useful for improving or maintaining ripe rot control with fewer fungicide inputs, which may be beneficial for the environment and the reduction of fungicide resistance selection.