Tamanho de amostra para caracteres morfológicos e produtivos de trigo mourisco

Abstract

This work aimed to determine the sample size for estimating the mean of the morphological and productive traits of the ‘IPR91-Baili’ and ‘IPR92-Altar’ buckwheat (Fagopyrum esculentum). Eighteen uniformity trials were carried out, consisting of two cultivars (‘IPR91- Baili’ and ‘IPR92-Altar’) in four sowing times in the 2017/2018 agricultural year and in five periods in the 2018/2019 agricultural year, with a size of 9 m × 17 m. On average, every 14 days, 100 or 110 plants were randomly collected, and their morphological traits (plant height and the number of leaves and nodes) and productive traits (leaf fresh mass, stem fresh mass, and shoot fresh mass) were measured. After drying in an oven, the leaf dry mass, stem dry mass, and shoot dry mass were determined. Plants were collected from emergence to flowering. The following statistics were observed: minimum, maximum, mean, median, variance, standard deviation, standard error, coefficient of variation, asymmetry, kurtosis, and p-value of the Kolmogorov-Smirnov normality test. The sample size of each trait was calculated for estimation errors equal to 2, 4, 6, ..., 20% of the mean. The results showed variability in sample size for morphological and productive traits and among buckwheat cultivars. For the ‘IPR91-Baili’ cultivar, considering the morphological and productive traits, larger sample sizes are needed than for ‘IPR92-Altar’. The 80 and 70 plants are needed, respectively, for the ‘IPR91-Baili’ and ‘IPR92-Altar’ cultivars to maintain the mean of the morphological and productive traits with a maximum estimated error of 20%.