Número ideal de amostras para classificação comercial de cultivares de trigo no Brasil

The objective of this work was to estimate the optimal number of samples for commercial classification of wheat (Triticum aestivum) cultivars for gluten strength average, in the states of Rio Grande do Sul and Parana, the main wheat producing regions in Brazil. The alveography analysis results of grain samples, from competition assays of three wheat cultivars (up to 280 experiments in 11 years), were used. The optimal number of samples was estimated based on the properties of the mean sampling distribution by adaptation region, in the states of Rio Grande do Sul and Parana. The optimal number ranged from 13 to 22, 30 to 49, and 121 to 197 samples per state, respectively, for to error margin of 30, 20, and 10 gluten strength units (W), with minimum of 7, 16, and 63 samples per adaptation region and per state for error margin of 30, 20 and 10 W, respectively. The optimal number of samples for commercial classification of wheat cultivars, in the states of Rio Grande do Sul and Parana, is of at least 13 samples per state and of 7 samples per adaptation region and per state for error margin of 30 W, which are higher values than those used by breeders for classification.