Abstract

A uniformity trial was conducted using the variety SL95 4430 at the Sugarcane Research Institute, Uda Walawe Sri Lanka to determine the optimal plot sizes and sample sizes of sugarcane for their efficient assessment in different stages of the varietal selection program in Sri Lanka. The optimal plot sizes were determined for the characteristics; plot weight (WT), number of millable stalks (ST), hand refractometer brix (HB), laboratory brix (Brix), stalk length (SL), stalk diameter (SD), rind hardness (RD), pol in juice (Pol), purity (Pur), pure obtainable cane sugar (POCS) and fibre content (Fib) by determining the point of inflection of the coefficient of variation (CV) verses plot size graph through split line regression and maximum curvature methods. The optimal sample sizes were found for the prescribed margin of error of 10% of the plot mean for the above-mentioned variables except for the number of millable stalks. Three-metre long single-row plots and five-metre long two-row plots were found as the optimum for sugarcane variety evaluation in stages 2 and 3 of the selection program, respectively. For assessing the varieties for hand refractometer brix and rind hardness at stage 1, samples of three and four millable stalks, respectively from a clump was found as the optimum. The optimal sample sizes to appraise varieties in variety selection stage 2 for stalk length, field brix and rind hardness were six, three and four millable stalks; respectively Six millable stalks for stalk length and 3 millable stalks for laboratory brix, purity and fibre content were found to be the optimum sample sizes for assessing the varieties at stage 3. For the evaluation of varieties in Preliminary Yield Trials and Replicated Yield Trials, a sample of 12 stalks should be obtained and eight stalks should be used for extracting mixed juice for the analysis of brix, pol, and purity and the remaining four stalks for the analysis of fibre content.