S05 - Session O1 - Detection and prediction of Botrytis cinerea incidence and vase life of cut roses using thermal imaging and neural network techniques
Information
Authors: Byung-Chun In *, Yong-Tae Kim, Gyeong-Rok Yang, Suong Tuyet Thi Ha
Pathogenesis of Botrytis cinerea ( B. cinerea ) causing gray mold of cut roses resulting in physiology disorders within rose petals that decrease longevity and economic values of cut flowers. In this study, we examined the use of thermal imaging for estimating the changes in petal temperature in cut rose flowers during B. cinerea infection and developed the vase life prediction model via neural network (NN) technique. As early as 3 d after inoculation, infection with B. cinerea increased the average temperature of rose petals by almost 1.5 ºC, compared with non-inoculated flowers. One day before the appearance of necrotic symptoms on day 4, the temperature of petals in infected flowers was significantly higher than that of non-inoculated flowers. These data revealed that the changes in temperature of infected petals could allow the discrimination between infected and healthy flowers in the thermal images, even before visible necrotic symptoms of gray mold appeared. Additionally, a strong positive correlation ( r² = 0.91) was observed between disease symptom severity and petal temperature estimated with thermography. The multiple regression analysis results also indicated that the longevity of cut rose flowers was negatively affected by, and primarily depended on petal temperature difference. Thus, we developed the vase life prediction model using petal temperature difference parameters. The NN analysis showed that the prediction model had a relatively high predictability between observed and predicted vase life. These results show that thermography is an effective technique for evaluating changes in petal temperature and is a practical method for detecting early B. cinerea infection in the cut flower industry. In addition, the NN is also a superior technique to predict and guarantee the longevity of cut rose flowers.