GENETIC VARIABILITY, CORRELATION COEFFICIENT, PATH COEFFICIENT AND PRINCIPAL COMPONENT ANALYSIS IN TOMATO (Solanum lycopersicum L.) GENOTYPES

Tomato is one of the most popular and major vegetable crops in Bangladesh as well as in the world. Genetic variability, phenotypic and genotypic correlation coefficient analysis, genetic advance, heritability, path coefficient analysis, principal component analysis and cluster analysis were done using fifteen tomato genotypes based on yield and yield contributing characters for Bangladesh condition. The objective of this experiment was to select the best genotypes as parents for specific traits for future hybridization programs. Phenotypic coefficient of variations for all the traits showed higher compared to genotypic coefficient of variation that revealed the environmental influence controlling the expression of genes responsible for that traits. Clusters per plant showed the highest difference between phenotypic (60.80) and genotypic (51.52) coefficient of variation. High heritability with high genetic advance in percent of mean were found in case of fruits per cluster (92.25 and 104.76), fruits per plant (98.33 and 203.65), fruit weight (95.87 and 103.87) and yield per plant (93.19 and 123.86) suggested that these traits would be considered for genotype selection. Correlation studies revealed that yield per plant was significantly positive correlation with clusters per plant (0.4976) and branches per plant (0.4949) at genotype level and cluster per plant (0.4851) at phenotypic level which indicated the good traits for parent selection. Path coefficient analysis showed direct positive effect of days to first flowering (0.0054), clusters per plant (0.0092), fruits per plant (0.0024) and yield per plant (1.0063) on yield per hectare. Cluster analysis revealed

the highest inter cluster distance between clusters I and III which indicated that genotypes G1, G2, G8, G9, G10, G11 and G14 can be selected for high heterosis in hybridization program.