关键词: 水稻, 产量构成, 差异调查, 种植密度, 浙江

Abstract: Rice yield difference widely exists among different cultivars and paddy fields. At present, rice yield potential is hard to breakthrough, so decreasing yield difference and increasing yield of different paddy field is one of the effective ways to increase rice average yield. In this study, in order to analyze the cause of rice yield difference, we investigated the grain yield and 23 typical cultivars, which were planted in 236 different paddy fields in Zhejiang Province, including early-season indica conventional rice, single-season indica hybrid rice, single-season japonica hybrid rice, single-season japonica conventional rice and late- double season indica hybrid rice, and further analyzed influence of planting density on grain yield. The results showed that the coefficient of variation of grains per panicle and effective panicles per plant were larger than that of 1 000 grain weight, indicating they were the main causes of yield difference. Relation analysis indicated that the correlation coefficient of grains per panicle and yield was the largest in yield components and was significant in single-season japonica conventional rice and late-season indica hybrid rice, and rice cultivar with large panicle usually had higher yield than cultivar of little panicles. With the increase of planting density, rice grain yield first increased then decreased. In single-season indica hybrid rice, when rice planted in （1.00-1.20）×10⁴ per 667m², the frequency of yield larger than 650 kg·667m^-2 was higher, when planted in （1.20-1.40）×10⁴ per 667m², the frequency of yield between 500-650 kg·667m^-2 was higher, and fewer paddy field was lower than 500 kg·667m^-2, grain yield was relatively stable. In late-season indica hybrid rice, the reasonable planting density was （1.10-1.40）×10⁴ per 667m²．

Key words: rice, yield components, difference investigation, planting density, Zhejiang