Principal components analysis in sunflower (Helianthus annuus L.)
1998
Miklic, V. | Marinkovic, R. | Crnobarac, J. (Institut za ratarstvo i povrtarstvo, Novi Sad (Yugoslavia))
Field experiment was carried out with 20 NS-hybrids of sunflower. Plant height, leaf number, head diameter, seed yield, 1000-seed mass, test weight, contents of oil, seed, husk and protein were analysed. Seed yield is a complex trait. According to Pearson's coefficient correlations it was significantly dependent on plant height and test weight. Using PCA, that finds linear combinations between original variables, provides possibility of analysis relationships among more traits. The first four principal components explained 76.1% of total variability of the origin variables, and creating a simpler system of four new complex variables (principal component). The clear membership of origin variables to principal components provided by varimax rotation of principal components. The first PC explained 27.26% of total variance and consists contents of oil, husk and kernel, i.e. traits, which present chemical composition of seed. The second PC explained 19.03% of total variance and consist of: content of protein, test weight and head diameter and in less of plant height, i.e. traits which present physical properties of seeds. The third PC explained 15.76% of total variance and consists of: leaf number and plant height, and less head diameter, i.e. traits which present vegetative growth of plant. The fourth PC explained 11.06% of total variance and consists of: seed yield and 1000-seed mass, and less plant height, i.e. traits, which present quantitative properties of yield.
显示更多 [+] 显示较少 [-]