圃場周辺の土地利用状況からアカスジカスミカメによる斑点米被害を予測するモデルの検証：宮城県の巡回調査データに対する適合性

抄録

We evaluated the potential availability of the predictive model of pecky rice damage by Stenotus rubrovittatus using land-use data (Tabuchi et al., 2022) in Miyagi Prefecture, considering the data from the Miyagi Prefectural Plant Protection Office. We also improved the model from the combined data of the current study of 38 rice paddies from 2017 to 2020 and from previous studies. We then applied the current study data to the predictive model and calculated the predicted value of pecky rice damage. Moreover, the model of the previous study was improved by adjusting the parameter coefficients of the fixed factors. The accuracy of the improved model was evaluated by comparing the observed and predicted values of the model. The relationship between predicted and observed values was positive and marginally significant in the model of the previous study with the data of the current study. At the same time, the relationship was statistically significant in the improved model. The coefficient of determination of the improved model increased by 0.269 compared with the model from the previous study. Our study demonstrated that the predictive model of pecky rice damage was available with scattered data on the prefectural scale and was possibly applicable in other areas.