Modeling Incidence of Bolting in Autumn Planting Onion in Snowy Region

Abstract

Using the bolting rate data in autumn planting onion in Tonami City, Toyama Prefecture, and Joetsu City, Niigata Prefecture, which are snow-covered areas in winter, this paper examines a bolting rate estimation model. The model uses two indexes, VD and VD_LN: VD is the value integrating the daily temperature response of vernalization, and VD_LN is the value integrating the product of the daily temperature response and number of leaves to include the effect of plant size on bolting. The coefficient of determination, R², in the nonlinear regression to the approximate curve with the bolting rate was investigated for VD and VD_LN by varying the periods of integration. When VD was used, R² was under 0.25 in all cases, and the rate of bolting could not be estimated accurately. On the other hand, using VD_LN, R² was around 0.75 at the maximum in the integration until January 20, or 80 days after planting when all the data of Tonami and Joetsu were used. R² was over 0.95 in the integration until February 20 or later, or until 120 days or more after planting when only the data of Tonami were used. Therefore, the effectiveness of the model using VD_LN was indicated. Practically, VD_LN > 490 at 140 days after planting was shown as a threshold of bolting incidence in Tonami.