Genetic diversity in stem amaranth based on principal component, cluster and biplot analyses

Abstract

Genetic diversity is the base line of any breeding program determining the variability of germplasm for crop improvement strategies. Approaches like principal component and cluster analysis measure the amount of genetic diversity in different characters and also assess the relative contribution of different traits to the total variation. This research was carried out at the vegetable research field of Olericulture Division of Horticulture Research Centre (HRC), Bangladesh Agricultural Research Institute (BARI), Gazipur, Bangladesh to study the genetic diversity and select the high yielding stem amaranth genotypes using principal component, cluster and biplot analyses. A total of 27 stem amaranth germplasms were evaluated in RCBD with two replications in consecutive two years 2020 and 2021 in summer. Principal component (PC) analysis depicted that first four PCs with Eigen-values higher than unity contributing 81.012% of total variability for different traits among 27 stem amaranth germplasm. The PC1 accounted for 37.833% variation, showed the positive loading effect of stem fresh weight (yield/plant) (0.864), leaf fresh weight (0.728), shoot fresh weight (yield) (0.968), stem dry weight (0.829), leaf dry weight (0.760) and shoot dry weight (0.980). The PC2 accounted for 24.556% of the variation and the major contributing traits were plant height (0.796), relative water content (RWC) (0.723), SPAD value (0.759) and F_v/F_m (0.694). The PC3 accounted 10.95% of the variation, and traits with high positive loading effects were observed for number of leaves/plant (0.595) and leaf area/plant (0.727). Regarding the first two PCs, germplasms were differentiated into five diverse groups. On biplot, two germplasm i.e. AM65 and AM72, were in positive side and positioned far away from the origin and were considered as diverse from the others. Cluster analysis grouped all germplasms into five clusters. The germplasms in clusters-IV and V exhibited higher stem diameter, stem fresh weight, leaf fresh weight, shoot fresh weight as well as their corresponding dry weight. The four characters viz., stem diameter, stem fresh weight, stem dry weight and shoot dry weight showed maximum contribution to total divergence among 13 characters studied. The D² statistics confirmed the maximum inter-cluster distance between cluster-III and V followed by cluster-III and-IV, whereas the smallest inter cluster distance between clusters - I and II. Crossing between the germplasm of Cluster- III with Cluster -V and Cluster-III with Cluster-IV could result better genetic recombinant and might exhibit better heterosis in F₁ for hybrid development.  Selection of superior germplasm from different clusters could be used in future crop improvement program.

Bangladesh J. Agril. Res. 51(1): 583-603, March 2026