Comparison of plant breeding designs for genetic analysis

Different experimental designs have been proposed for the estimation of genetic variances. However, limited critical information regarding the comparison of the estimates of genetic variances and its reliability is available. Hence, an attempt was made to conduct an field experiment involving North Carolina designs I, II and III along with Diallel (Griffing Method I) mating design with bidi tabacco (Nicotiana tabacum L.) as test crop. Additive and dominance genetic variances estimated in different designs indicated that highest percentage of significant, positive variances with lower standard error of estimate were observed in diallel and North Carolina design I. These variances were ranked (1 to 4) considering their significance, signs of variances and their magnitude of standard error. According to the overall ranks score of the design,the design with highest score value was considered more reliable. Diallel design having the highest overall score vaiue was found to be more reliable as compared to N C designs. Among three, N C designs, NCD-I had slightly higher score value while II and III had the same score values. Negative variances were observed in all the designs.