This study aimed to compare multiallelic distances to quantify genetic diversity in papaya. We evaluated forty-three individuals in the S2 generation, from the backcross between F1 (Cariflora x SS783) and Cariflora, and four accessions from the UENF/Caliman Germplasm Bank. Genetic distances used were Smouse and Peakall (1999), Kosman and Leonard (2005) and weighted index. Clustering among genotypes was performed using the hierarchical unweighted pair-group method with arithmetic mean analysis (UPGMA) and projection of the distance on the bidimensional plan. A high correlation between genetic distances was observed; however, through UPGMA group analysis, the distance determined by the weighted index provided the complete separation of 52BC1S2-08, 52BC1S2-29 and 52BC1S2-34 inbred lines. Through projection of distances in the plan, Kosman and Leonard (2005) coefficients and weighted allowed the differentiation of individuals in the S2 generation (52BC1S2-08, 52BC1S2-29 and 52BC1S2-34), the progenitor ('Cariflora' and 'SS783'), and the four germplasm bank accessions in a different manner than the Smouse and Peakall (1999) index, which did not provide this discrimination among the accessed genotypes. We conclude that the Kosman and Leonard (2005) coefficient and weighted index are more efficient than the Smouse and Peakall (1999) algorithm on the disposition of the accessed genotypes in dendrograms and in the Cartesian axis displaying genetic similarity.
Carica papaya L.; microsatellite markers; cultivar characterization; cluster analysis
