Table 1.
Characteristics of twenty SSR loci used in the present study detected by Popgene software.
Table 2.
The standard dimensionless matrix of the eleven parameters related to the polymorphism and discriminability of the 20 SSR loci evaluated by the TOPSIS method.
Figure 1.
Part of the SSR fingerprints and the corresponding molecular identities of 480 Chinese traditional chrysanthemum cultivars established by five core loci.
For the SSR fingerprints, the white blocks represent the presence of amplified fragments, which were transformed to “1” in the binary identities; the abscissa represents the amplified 63 alleles that ranged from 105 to 389 bp, which were reordered according to the fragment lengths of the top five loci evaluated by the TOPSIS method; the vertical axis represents the sample code of the selected cultivars. The denary identities were transformed from the binary identities.
Figure 2.
Comparison between the cluster analysis based on Nei's genetic distances and the appropriate subpopulation number (K = 10) detected using STRUCTURE 2.3.4.
Each gray scale represents one subpopulation and was assigned separately for the estimated appropriate value for K = 10. As many as ten groups were clustered at the Nei's genetic distance of 0.836. Group 1: brown and dark-red cultivars; Group 2 and 3: cultivars with tubular petal type; Group 4 and 5: modern cultivars; Group 6: old cultivars; Group 7: cultivars with anemone petal type; Group 8: red cultivars; Group 9: orange cultivars; Group 10: pink cultivars.
Table 3.
Analysis of Nei's gene diversity in subdivided populations according to Nei [45].
Figure 3.
Estimation for the best subpopulation numbers based on the appropriate K values.
(A) lnP(D) values revealed using STRUCTURE 2.3.4 by 20 separate runs with K values between 2 to 14. The separate 20 runs of the software revealed 10, 11 or 12 subpopulations, which showed a stable increase with small variations. (B) The mean ΔK values of 20 separate runs with K values between 2 to 14 based on lnP(D) values. The mean values of ΔK among the 20 runs reached a maximum at K = 10.
Figure 4.
The first two axes of a principal component analysis representing the microsatellite data obtained from 480 Chinese traditional chrysanthemum cultivars.
The brown and dark-red, red, orange, pink, tubular, anemone, modern and old cultivars were selected as phenotypic indicators to justify the space distributions of each group, which were determined from the cluster analysis based on Nei's genetic distances of the 204 polymorphic alleles.