874 Pb 232 Statistical Determination of Adequate Sample Size and Reduction of Variance in Tomato Tissue Culture

Abstract

<jats:p>To estimate the components of variance of a tissue culture experiment, hypocotyl sections of tomato cultivars (C) `Healani,' `Kewalo' and `Anahu' were cultured on modified Murashige and Skoog medium with 10 uM IAA and 10 uM BAP. Three explants were placed in each of 45 flasks per cultivar. This procedure was repeated three times (R) on different dates. Mean values of the number of shoots per explant were 7.0. 8.2 and 9.2 for `Healani', `Kewalo', and `Anahu', respectively. The data were transformed by adding 0.5 and then taking the square root of the sum. Components of variance of the transformed data were σ<jats:sup>2</jats:sup> = 0.01, σC<jats:sup>2</jats:sup> = -0.02, σRC<jats:sup>2</jats:sup> = 0.09, σFC<jats:sup>2</jats:sup> = 0.22 and σE<jats:sup>2</jats:sup> = 1.68 (error term). The components of variance were used to determine the theoretical variance of the cultivar (treatment) mean. Results indicated that the most efficient method of reducing the error is by increasing the number of explants per flask. Stein's two-stage sample equation determined the number of explants required for a 95% confidence level to be 1613, 807 and 380 to detect differences of 5% 7% and 10%. respectively, of the overall transformed mean.</jats:p>