Crop scientifical studies on the yield-forecast of lowland rice. : Preliminary report : I. Crop scientifical forecast on the yield of lowland rice and the annual variations relating to the constitutional factors of its yield

• Published in: Japanese Journal of Crop Science Vol. 21; no. 3-4; pp. 219 - 220
• Main Authors: MATSUSHIMA, Shozo, YAMAGUCHI, Shunzi
• Format: Journal Article
• Language: English
• Published: CROP SCIENCE SOCIETY OF JAPAN 30.06.1953
• ISSN: 0011-1848; 1349-0990
• DOI: 10.1626/jcs.21.219

(1) The authors intended to forecast the yield by judging the growing process of rice plants. For this purpose they tried to analyze the factors constituting the yield of rice crop and to trace up their growing processes. (2) The following four factors were considered to constitute the yield of rice crop. (i) Number of panicles per unit area. (ii) Number of spikelets per panicle. (iii) Fruiting percentage. (iv) Weight of 1, 000 grains (brown rice). (3) First of all, the annual variations of these four factors were studied as fundamental data for forecasting the values of them. Using the data of the studies on climate-adaptability of the crop performed at Konosu Agricultural Experiment Farm, in which nine varieties - including early, intermediate and late maturing ones - were used as material, the annual variations of the factors were investigated by calculating coefficients of variation through nineteen years. (4) The results obtained were as follows : (i) Variation coefficients of the number of panicles ranged from 10.2 % t 15.6 %. The less was the number of panicles in varieties, the greater was the coefficient of variation ; the correlation coefficient between them had been so significant that it was r = 0.88. (ii) Coefficients of variation of the number of spikelets per panicle ranged from 6.5 % to 16.6 %, and those of intermediate maturing varieties were smaller than that of any other varieties in general. (iii) Coefficients of variation of the fruiting percentage and the weight of 1, 000 grains were 5.0∼9.1 %, 2.2∼4.4 % respectively, and only the slight differences among varieties were seen. (5) It was found in every variety above mentioned that both the number of panicles and the number of spikelets per panicle showed the largest variation, the fruiting percentage followed next, and the weight of 1, 000 grains was in final order. (6) These results obtained are very useful for the yield-forecast of rice crop because of the following reason : it is supposed that the order of these four factors determined in the growing period of rice plant may be such as the first, the number of panicles, the second, the number of spikelets, the third, the fruiting percentage and the fourth, the weight of 1, 000 grains ; this order agrees with the descending order of values in their annual variations. That is, the larger is the annual variation in the factors, the earlier it is determined in the stage of growth. (7) The annual variations of these four factors are available as the accuracy-index of the yield-forecast. For example, the number of panicles, the number of spikelets per panicle and the fruiting percentage, exactly being forecasted and only the weight of 1, 000 grains remaining unknown, we can decide that the yield forecast is, under about 70 % probability, within 3.3 % probable error which is the mean value of variation of 1, 000 grains weight.