Vol: 34 Page: 223

Authors:
Leszek Sieczko
Wiesław Mądry
Andrzej Zieliński
Jakub Paderewski
Krystyna Urbaś-Szwed

Title:
 Application of principal component analysis (PCA) to multivariate characterization of genetic diversity in a durum wheat (Tritricum durum L.) germplasm collection

Language: Polish

Keywords:
principal component analysis (PCA)
cluster analysis
germplasm collection
genetic diversity
argonomic and phenological characters
durum wheat

Summary:
The aim of the paper was 1) to give a sketch of the principal component analysis (PCA) theory, 2) to present a synthesized interpretation of the correlation matrix of the three-year averages of quantitative agronomic and phenological characters for 89 accessions (genotypes) in a durum wheat germplasm collection and obtain, through the PCA, information on how each character discriminated the genotypes, 3) to apply the PCA in reducing dimensionality in study of multivariate diversity of genotypes in the durum wheat collection and of phenotypic characterization of the collection genetic diversity using a small number of dimensions. Field experimental data obtained in 2000 through 2002 years have been used. In order to divide the accessions into groups, homogeneous with respect to the seven characters, the hierarchical cluster analysis has been applied (with both the farthest neighbour and the Euclidean distance methods). The accessions were divided into seven homogeneous groups. The PCA appeared to be an effective method in simplified but sufficiently exact presentation of multivariate diversity of durum wheat genotypes by means of the first two principal components explaining more than 75% of the Euclidean distance variation between accessions for the seven characters. Thus, both methods, cluster and PC analyses, complemented each other in providing useful information for more efficient evaluation, development and using this collection.