Cultivars Clustering by Multivariate Analysis on Horticultural Characteristics of Yardlong Bean

Abstract

The success of plant phenotypic selection depends upon the range of genetic diversity available in the population. The objective of this research was to cluster 23 yardlong bean genotypes based on multivariate analysis of their horticultural characteristics. These genotypes consisted of 20 commercial cultivars, namely Naka, Greenarrow 692, Saifon, Nuethong 9, Saitara, Tarnthong, Lamnamphong 2, Suvarnabhum, Saisawan, Petkajee, Sornsawan 5, Moneygreen, Rangsit 888, Lamnamchee, Yodpetkasem, Airgreen 99, Munggornyok 9, Choiphiphop, Chiataisen, and Nigrodok, and 3 selected lines namely Green 33, Bangpra 2, and Purple 33. They were laid out in a randomized complete block design with 3 blocks under 2 seasons. Data were recorded for 12 horticultural traits and the mean values ​​of traits from 2 growing seasons were used for multivariate statistical analysis by Mahalanobis distance (D²) method to study genetic divergence from those 12 traits, and the genotypes were grouped into clusters following Tocher’s method.  The results revealed that the genotypes were classified into 5 clusters. Cluster 1 had the maximum (15) and cluster 5 had the minimum (1) number of genotypes. Cluster 4 had the highest intra-cluster distance (28.34) and the lowest in cluster 5 (0). The inter-cluster D² values of five cluster revealed that the highest inter-cluster generalized distance (106.42) was between cluster 5 and cluster 1, while the lowest (33.38) was between cluster 3 and cluster 2. The percent contribution of each character towards divergence was as followings; pod length with the maximum contribution of 21.74%, followed by pod weight (18.58%), plant length (12.65%), number of pods per plant (11.46%), leaf width (5.53%) and pod width (8.70%), respectively.