LOCAL PHENOTYPES
By the principal component analysis, we clarified populations
distributed each area show the characteristics which different
phenotypes each area.
On the other hand, we conducted one-way ANOVA to understand the
phenotypes in each group because they were classified into different
phenotypic groups within the region by cluster analysis.
The mean values of some variable between groups and the results of
one-way ANOVA are shown in Table 2.
There were significant differences among the groups in 43 variables,
except for stamen length and petiole length.
In group A, the plant height was lower and the flower color was darker.
In group B, flowering period was shorter, plant height was higher, and
leaf and flower colors tended to be darker.
In Group C, the flowering period was long, the plant height was low, and
the number of florets per cluster was small, but the clusters were dense
and the flower color tended to be light.
In group D, flowering was slightly delayed and the number of
inflorescences was small, while the number of lateral shoots was large
and the flower color was light.
In Group E, flowering was delayed, the inflorescence and flower cluster
was large, the leaves tended to be broad but less serrated.
As a result of correlation analysis, The number of florets showed a
positive correlation with the longitudinal diameter of the inflorescence
and the transverse diameter of the inflorescence (0.72, 0.61),
indicating that the inflorescence size increased in proportion to the
number of flowers.
The days to flowering were not correlated with the number of
inflorescences and the distance from the length/threshold of the
inflorescence to the first inflorescence, but correlated positively with
the first flowering nodal position (0.68).
As there were significantly different variables among the regions, we
performed correlation analysis among the populations within each region,
but no correlation coefficient of 0.5 or more was shown.