Predicting harvest date of 'Niitaka' pear by using full bloom date and growing season weather.

The effect of full bloom date and growing season weather on harvesting date of 'Niitaka' pear (Pyrus pyrifolia) in Naju province and the model of multiple linear regression for predicting the fruit growing days was studied. Earlier year in full bloom date, the harvesting date tended earlier but fruit growing days tended longer. Mean and coefficient of variation of fruit growing degree days (GDD) accumulated daily mean and maximum temperature at the base of from full bloom date to harvesting date was 3,565, 2.9% and 4,463, 2.5%, respectively. Fruit growing days was not correlated with the fruit GDD accumulated daily mean and maximum temperature at the base of in each month but highly correlated with GDD accumulated daily meteorological factors at days after full bloom date. Especially, it was highly negatively correlated with GDD accumulated daily mean and maximum temperature at the base of from day after full bloom to day. The determination coefficient () of multiple linear regression model by full bloom date, GDD accumulated daily mean and maximum temperature from day after full bloom to day for predicting fruit growing days was 0.7212. As a result, the fruit growing days of 'Niitaka' pear in Naju province can predict with 72% accuracy by the model of multiple linear regression.