Grouping of flue-cured tobacco genotypes based on multivariate statistical analysis

Forty nine tobacco genotypes, selected from Tobacco Research Center of Rasht, Iran, were evaluated for genetic diversity on the basis of 14 characters (No. of leaf, leaf length, leaf width, leaf area coefficient, leaf shape index, flowering duration, days to flowering, plant height, stem diameter, fresh leaf yield, dry leaf yield, revenue, price of one kilogram dry leaf and SPAD value) using simple lattice design (7×7) with two replications in 2010. Cluster analysis using UPGMA method assigned the 49 genotype in five different groups consisting of 6, 1, 25, 4 and 13 genotypes, respectively. Grouping results were confirmed by canonical discriminate function analysis and Fisher Linear method (87.8%). C258×MC944 hybrid had the highest values for most economical traits. The fifth group included genotypes that had low value for most economical characteristics. Factor analysis results based on principal component analysis after Varimax rotation showed that four factors; economic yield, phonology, price and leaf characteristics determined 79.19% of variation from the total observed variation. Fourth region was detected as desirable region on the basis of the first and second factors evaluation. Results of the present experiment could be used for selection of suitable parents in crossing programs to increase dry leaf yield in tobacco.