Variability of the agronomic indicators of Pennisetum purpureum cv. Cuba CT-115 with the sampling distance

Resumen

An experiment was conducted during two years to establish the variability of the agronomic indicators of Pennisetum purpureum cv. Cuba CT-115. Samples were taken at different sampling distances (1, 2, 3, 4 and 5 m length). The data of an experiment with random block design and four replicates were used. The lowest variance, standard deviation and coefficient of variation were obtained with sampling distance of five meters for the height (4.3, 2.1 and 1.88, and 1.1, 1.0 and 0.99 for the rainy and dry season, respectively). Similar performance had the per cent of leaves and dry matter of the whole plant. The sampling distance of five meters propitiated lower yield variability (18.04 and 12.35 % in the rainy season and dry season, respectively). Similar performance had the number of bunches/m (8.34 and 12.27, in that same order). In all the indicators, the lowest width of the values range was obtained, in general, in the distance of five meters. When sampling the population, five bunches/m recorded the lowest variability compared with 11 bunches/m. It is concluded that for the best estimation of the agronomic indicators, the samplings should be conducted in five linear meters. No less than five bunches as sample units should be taken. Similar studies in grasslands with different growing habits are recommended.
Key words: sample, yield, grass.