Field prediction models combining the first peak for ascospore dispersal with the degree-day for latent period in apple blotch caused by <i>Diplocarpon coronariae</i> based on weather data

<p>Apple blotch (Marssonina blotch), caused by Diplocarpon coronariae (synonyms: D. mali; Marssonina coronaria) is a major disease of cultivated apples in Japan. Here we tested the usefulness of weather data to predict the start of primary infection and symptoms in relation to data for ascospore discharges and apple growth stages. We analyzed 18 cases of the 1st peak of ascospore dispersal (“1st peak”) among seven prefectures in Japan during 1998–2020. All the 1st peaks occurred on rainy days after bloom started for cv. Fuji, and on average, they coincided with the petal fall stage. A total 76 rain events that occurred from start to 1st peak were examined (1st peak: 18 cases; nonpeak: 58 cases). Logistic regression analyses revealed that, of three models examined, the wetness duration model was the best fit model based on Akaike information criterion. A final logistic regression model only comprised the interaction variable between average temperature (≧13.8°C) and duration (≧6 h) during wetness. In the final model, the rain events could be divided into 1st peak and nonpeak with 97% accuracy. The thermal time required for a latent period during 90% of total apple blotch lesions after conidial inoculation was estimated as 159 degree-days (base = 4.6°C). Both date of first infection and first symptom found in the field exposure and survey corresponded almost exactly to the predicted dates using both the 1st peak model and the latent period. These results suggest that weather data can be used to estimate both the start of primary infection and symptoms to analyze epidemic trends for apple blotch.</p>