Crop-Scientific Studies on the Yield-Forecast of Lowland Rice. (Preliminary report) : (XXVI) On the mechanism of determining the number of spikelets per panicle in rice plant.

摘要

MATSUSHIMA, one of the authors, Suggested the fact in the previous paper that the mumber of spikelets per panicle in rice plants is determined by the balance between the number of differentiating spikelets and that of degererating ones. The purpose of the present paper is to reaffirm the above mentioned fact. 1. Tracing up the increasing process of the number of spikelets on young panicles at an interval of two days, the authors found a maximum, in the number of spikelets at about the midst of reduction division stage. The spikelets which degenerated could be distinguished more and more clearly, and they increased in their number as the young panicles grew longer. (cf. Fig. 1) 2. From the results of thinning-out experiments, in which every other hill was thinned out in the rice-fields of 3 varieties at an interval of 10 days, and from the results of water-culture experiments, by which we could apply a sufficient fertilizer to rice-plants in a given 10-day perios through the whole growing season after transplanting. it was elucidated that the good conditiond in the 10-day period from the differentiating stage of 1st bract primordia to the beginning stage of differentiation of spikelet primordia were most effective on increasing the number of differentiating spikelets, and good conditions during the period after that were of no effect to increase it (cf. Fig. 2 apd 3) These facts led the authors to conclude that the top-dressing of fertilizer at the stage of young panicle formation which has been generally practised in Japan should be of little effect to increase the number of differentiating spikelets, but be of a great value for preventing the degeneration of spikelets so far as the number of spikelets per panicle is concerned. 3. The reason why the 10-day period from the differentiating stage of 1st bract primordia to the beginning stage of differentiation of spikelet primordia is most effective in increasing the number of differentiated spikelets, and the periods after that, are of no value in increasing it, can be explained by the fact that the number of spikelets per panicle is strongly correlated with the number of branches of 2nd order. (cf. Fig. 4) As thb 10-day period above mentioned contains the stages of differentiation of 1st and 2nd order, good conditions in the period are considered to have good effects upon the differentiation of branches of 2nd order, and consequently their number increases. 4. From the investigation of branches and spikelets, which degenerated it was made clear that spikelets which once differentiated might degenerate in those three ways as follows: the most common way is to degenerate with branches of 2nd order, the second way is to do with branches of 1st order, though the latter is not so common as the former, and the third is to do by spikelets alone, the percentage of degenerating spikelets induced by the last case being the least of those by all the three ways. (cf. Table 1) 5. Basing upon the principle mentiond above, the authors suggested that it would be a recommendable method for increasing the number of spikelets per panicle to fertilize the plant at two stages, one being the differentiating stage of 1st bract primordia, and the other being the stage just before the reduction division when spikelets are apt to degenerate most easily.