Estimation of the occurrence of pecky grain damage by a sweeping census of Leptocorisa chinensis Dallas (Hemiptera: Alydidae) in rice fields

Abstract

To estimate the occurrence of pecky grain damage caused by Leptocorisa chinensis in rice fields, I censused the rice bug utilizing the sweeping method, and checked the occurrence of pecky grains among total hulled grains from 2000 to 2004. Using the data obtained from the field census, I constructed 30 logistic regression models. The explanatory variables of each model were different combinations of rice variety, date of initial heading stage, number of adults at early ripening stage, and numbers of adults, nymphs and their total at the mid-ripening stage. The dependent variable of all models was whether or not the occurrence of pecky grains among total hulled grains was over 0.1%. Akaike's information criteria indicated that a model whose explanatory variable was number of nymphs at the midripening stage was statistically better than the other models. These results suggest that a census of nymphs at the midripening stage, which is 200-300 day-degrees from the initial heading stage, is an appropriate way to estimate the occurrence of pecky grains.