VARIABILITY AND INTRASPECIFIC CLASSIFICATION OF INDONESIAN EDIBLE CANNA (Canna indica L.) BASED ON RAPD MARKER ANALYSIS

Edible canna (Canna indica L.) is herbaceous plant originated from South America which its rhizome commonly used and its utillization not optimize in Indonesia. Molecular characterization are needed to determine characters of C. indica for genetic variability to improve edible canna varieties. For that reason, the purpose of this research was to determine diversity and phenetic relationship of C. indica based on molecular characters using RAPD analysis. Samples were collected from Indonesian Center for Agricultural Biotechnology Research and Development (BBBiogen). Data was analysed by RAPD analysis using OPA-01, OPA-02, OPA-10, OPD-02, and OPN-05 primers. Similarity index was counted by Gower General Similarity Coefficient formula based on molecular scoring. Cluster analysis was conducted by Unweighted Pair Group Methods using Arithmetic averages (UPGMA) method to create a dendogram. Principal component analysis (PCA) was also performed to defined role of each molecular character in grouping of accessions with Multivariate Statistical Program (MVSP) v. 3.1 software. The results showed that C. indica has divided into two main clusters. There were the green and red cultivar groups with 56% similarity. The green cultivar group also divided into green and green purple based on colour of sheats, tip of bud, rachis inflorescence, petals, brachtea, and colour pattern of staminodia with 73 % similarity. Red cultivar divided into red and red purplish based on colour of sheats, rachis inflorescence, and petals with similarity index 74%.