A complete randomized block design was used for all experiments in which each treatment was replicated three times for callus and ten times for shoots. The standard error of the means and LSDs (least significant differences) were calculated using JMP statistical discovery software.