MULTIVARIATE ANALYSIS OF PEARL MILLET DATA TO DELINEATE
GENETIC VARIATION
SUNITA CHAUDHARY, PREM SAGAR, B. K. HOODA AND R. K. ARYA
Department of Plant Breeding
CCS Haryana Agricultural University,
Hisar-125 004 (Haryana), India
(Received : 22 November 2014; Accepted : 15 January 2015)
SUMMARY
Loss of biodiversity not only disturbs the process of plant development aimed at genetic amelioration but also disrupts the fundamental services that ecosystems provided to humanity. Assessment of variability is a multidimensional problem. The multivariate statistics can help in a comparative assessment of genetic variability. A set of 66 lines of pearl millet was analyzed for cluster and principal component analysis (PCA). PCA identified six principal components which explained 77.7 per cent of total variability among the 66 genotypes. The PC1 characters –main ear weight, dry fodder weight, total ear weight, grain yield, growth rate and plant height, the major characters of plant biomass and the basis for grain yield contributed maximum 35.94 per cent variability among the lines. The remaining PCs accounted for progressively lesser and lesser amount of variability. The lowest contribution 5.27 per cent was recorded by PC 6, the characters grain starch, starch recovery and ear girth. Only grain starch contributed positively to all the six components. The genotypes 50 (77/371), 3 (IPC-115), 41 (204/2 MP), 12 (IPC-1462), 37 (TCH-37-1), 22 (TCH-10-1), 61 (1307), 14 (862-P2), 20 (TCH-3-2), 40 (204-2-3) were found to be better performers and diverse on the basis of principal factor scores with regard to grain yield and yield contributing characters. Hierarchical cluster analysis grouped 66 genotypes into six clusters, cluster 1 included maximum number of 21 genotypes and clusters 3 and 6 had the lowest number of 6 genotypes. The results on hierarchical cluster analysis almost mimicked the PCA. The grouping pattern of genotypes obtained by cluster analysis and PCA plots was almost similar. A wide range of diversity for most of the traits observed would enable to pick lines with suitable traits to be used in a breeding programme. Genetic diversity was not essentially associated with geographic diversity.
Key words: Pearl millet, germplasm lines, growth rate, quality-starch, protein, cluster, principal factor, 3-dimensional plot