Abstract
Dry weight of rice plant is an important character for the vegetative diagnosis and growth forcast of this plant. However, it requires a long time and much labor to precisely estimate the dry weight, which makes the survey difficult. Accordingly, it is necessary to establish a new survey method effective in laborsaving while maintaining precision of a definite level or above. In this report, the sample size, precision and necessary time in a survey on dry weight by double sampling and ratio estimate were examined and analyzed so as to discuss the efficacy. When fresh weight was employed as an auxiliary variable, the time required for obtaining an estimation of precision comparable to the one obtained by using simple random sampling was shortened, which suggested that it was effective in laborsaving. When the number of tiller and the number of panicle obtained by stand observation were employed as auxiliary variables, it was advantageous that the number of hills to be taken from fields in practice could be decreased, compared with simple random sampling, though the degree of laborsaving was somewhat lowered. Further, the precision and necessary time in various dry weight survey methods, in which the fresh weight, the number of tiller and the number of panicle of 20 hills were employed as an auxiliary variable, were examined and analyzed to thereby clarify a survey method effective in laborsaving.
