241 Estimating Optimal Sample Size for Sweet Orange Fruit Quality Experiments

Abstract

Variability in fruit quality of citrus occurs among and within trees due to an interaction of several factors, e.g., fruit position, leaf: fruit ratio, and fruit size. By determining variability in fruit quality among i) fruit, ii) trees, iii) orchards, and iv) geographic locations where citrus is produced in Florida, optimal sample size for fruit quality experiments can be estimated. To estimate within-tree variability, five trees were randomly selected from each of three `Valencia' orange orchards in four geographic locations in Florida. Six fruit were harvested from each of two tree canopy positions, southwest top and northeast bottom; fruit were not selected or graded according to fruit size. °Brix and titratable acidity of juice samples were determined, and the °Brix: acid ratio was calculated. Statistical analysis of fruit quality variables was done using a crossed-nested design. The number of trees to sample and the number of fruit per sample were calculated. To estimate between-tree variability, 10 trees were randomly selected from each of three `Valencia' orange orchards from four geographic locations in Florida. Fifty-fruit composite samples were picked from around the tree canopy (0.9 to 1.8 m). Juice content, SSC, acid content, and ratio were determined. Using a nested design, the number of orchards and number of trees to sample were determined. There was greater variability in fruit quality among trees than within trees for a given canopy position; the optimal sample size when taking individual fruit samples from a given location and canopy position is four fruit from 20 trees. There was less variability in fruit quality when 50-fruit composite samples were used, resulting in an optimal sample size of five samples from three orchards within each location.