RESUMO
Morphological and molecular characterization of germplasm is essential for the improvement of cultivated plants efforts. This study investigated the genetic diversity of 32 Cyclamen genotypes comprising 16 C. persicum varieties and 16 Cyclamen species using multivariate analysis for 36 morphological traits (19 quantitative and 17 qualitative) and molecular characterization (SRAP and SCoT markers). The color CIELab parameters were collected via PCE-CSM7 that separately measured the leaves dark and silvery patterns and the flower's slip (region of the petal top), eye (the region of the petal base) and sepal. Genetic diversity was also evaluated using Shannon Wiener (H') and Simpson's (λ) Indices, and Pilou evenness (J) using the library vegan from R software. According to the principal component analysis, the variables that contributed the most were leaf pattern color, leaf abaxial surface color, pedicel coiling, leaf and stem number. The color indicators of Cyclamen leaves showed decreased L* values in darker colored genotypes, whereas increased a* values were noticed in flower eye and lower in slip. Molecular characterization was based on 26 SRAP and 12 SCoT markers that produced clearly repeatable DNA bands and exhibited significant levels of polymorphism. Based on the morphological traits and molecular markers data, the UPGMA method for hierarchical clustering technique was used to generate the dendrograms, and their entanglement was obtained using the Tanglegram algorithm from the dendextend package with the R software. Entanglement analysis (0.30) between dendrograms obtained from the morphological and genetic analysis using SRAP markers showed a high association. Comparison between color measurements of flowers (entanglement=0.45) and leaves (entanglement=0.47) with SCoT analysis revealed differences at species level, discriminating between similar genotypes. Combined phenotypic and molecular analysis improved the comprehensive estimation of real diversity in the investigated Cyclamen genotypes. The findings of the present study are useful for quantifying diversity and genetic variability in Cyclamen breeding and genetic investigations.