Evaluation of the phenotypic variation in a caper (Capparis spinosa L.) population growing in south of Tehran using multivariate analysis
Fatemeh Bina and Abdolamir Bostani
The aim of this study was to investigate the phenotypic variation among 100 caper (Capparis spinose L.) plants growing naturally in south of Tehran. Plant samples were taken randomly from three different regions of southern Tehran. The number of 29 phenotypic traits were studied on selected plants. Data were subsequently subjected to multivariate analysis. Results showed that fruit yield was significantly correlated with the number of fruits per plant, fruit weight, fruit length, fruit diameter, plant canopy, the number of thorns in every 50 cm of stem, thorn length and the number of flowers per plant (NFP). Also, there was a negative correlation between fruit length and the number of branches per each main stem. Stepwise multiple linear regression revealed that the number of main stems per plant, leaf width, petiole length, peduncle length and NFP were added to the model and hence had the greatest impact on fruit yield. Principle component analysis showed that the first two components accounted for 41.1% of the total variation. The first component was related to the fruit yield and its related traits, while the second component was related to the vegetative growth and showed competition between reproductive and vegetative functions. Cluster analysis of genotypes using Ward method and squared Euclidian distance criteria classified the samples into ten different groups. The results of this study suggested that crossing between samples 17 and 81 may produce useful recombinants.
Key words: Capparis spinosa, cluster analysis, graphical correlation coefficients, principal component analysis, stepwise multiple regression