水稲収量予察の作物学的研究 (予報) : IV. 生育時期別乾物重・その成長曲線並に相対成長率と豊凶との関係

抄録

The authors examined the value of measuring dry-weights of rice plants for forecasting yields of the crop, using the data (I) (1929-1935), (II) (1936-1944) of experiments on climate-adaptability of the plant at Konosu Agricultural Experiment Farm and the data (III) of artificial shading experiments carried out at the same place in 1950 and 1951 by the authors. And the results obtained are summerized as follows: (1) In the field where the growth was normal possitive correlation coefficients were almost always found between yields of kernels and dry weights of plants at different growth stages in the 9 varieties tested. Early varieties showed significantly high coefficients at every time shortly after transplanting, but most of other varieties did not show significant correlations at any times up to heading (Table 1. 2) (2) In the field where the growth of rice Plants became slender at the end of growing season, an early variety and a medium one showed significantly high coefficients at 12 days before heading and at heading time respectively, but late one showed very low values in all stages. (Table 3) (3) From the data (III) of shading experiments, in which the growing rice plants of a medium variety were covered with clothes in different stages, slightly significant coeffifients were obtained at the stage 5 days before and 14 days after heading. (Table 4) (4) Robertson's growth-curve equations of dry-weights for 3 varieties in 7 years were derived from the data (I), and the relations of crop yields to the epuations were investigated with special reference to their "K" which is termed the coefficient of growth. The results showed no close relation between them. (Table 5) (5) As a result of examining the correlations of yields with the value of relative weight-growth rates calculated from the data (III) at an interval of 6 days after transplanting, there were found mostly low coefficients between them (Table 6, average value of 3 medium vatieties). And also the correlations of yields with the values calculated at a long interval did not show significant coefficients, except in the case of early stage of early varieties. (Table7) (6) It was suggested from the results above mentioned, that tracing up the dry-weights through the growing period would not play any important role in forecasting yields in rice. In order to make the dry-weights of the plant valuable as the indexes to yield-forecast it should be born in mind that, firstly, methods of sampling plants to be weighed should be re-examined, and secondly, not merely the quantity of dry-matter but also the quality of it should be considered.