| Resumo: |
For 100 sunflower head, maximum and minimum diameters, thickness, height of the ground, and mass of all seeds on sunflower head (SH) for five conventional varieties were measured. To examine the correlation between maximum SH diameter and minimum SH diameter, SH thickness, SH height from the ground and SH seeds mass, linear regression method was used. Also SH diameter and SH thickness of four sunflower hybrids varieties were measured. Log-normal, Weibull and Generalized extreme value (G.E.V) distribution were used in order to model SH diameter and SH thickness distributions of all varieties and hybrids; also, Kolmogorov-Smirnov methods were used for comparison of all probability density. Results indicated that value of average of dimensions, geometric and arithmetic mean diameter, head height from the ground, and mass of the seeds on each head of Dorsefid and Sirena varieties were more and less than the other varieties, respectively. For all varieties, there was linear correlation between maximum SH diameter and minimum SH diameters, SH thickness and mass of seeds; while there wasn’t any correlation between SH height and maximum SH diameter. Results showed that whenever skewness and kurtosis had positive values, Log-normal and G.E.V distribution had good performance and whenever skewness and kurtosis had negative values, Weibull and G.E.V distribution had good performance based on Kolomogrov-Smirnov methods.
|
