Statistical analysis. For all the above experiments, at least
10 explants were used for each treatment and each experiment
was repeated three times. The embryogenic response
was calculated as the percentage of cultures in each treatment
producing at least one somatic embryo. The approximate
number of embryos formed per explant was calculated as
the mean number of somatic embryos produced per
responding explant in a particular treatment. The number of
somatic embryos of different stages formed from each
responding explant was counted under a stereo microscope
(Leica, Germany). The mean, standard error and one-way
ANOVA were calculated using SPSS (version 13) software.
The mean separations were carried out using Duncan's multiple
range tests (Duncan 1955) and significance was determined
at P≤.05.