Comparison of methods for estimates of molecular genetic diversity in genus Croton: influence of coefficients, clustering strategies and data projection.

We investigated 10 similarity (and disimilarity) coefficients in a set of 40 wild genotypes of Croton linearifolius subjected to analyses using hierarchical grouping methods, grouping methods by optimization and data projection in two-dimensional space. Genotypes were characterized by analyzing DNA polymorphism with the use of 15 ISSR and 12 RAPD markers. The distance measurements were compared by the Spearman correlation test, projection in two-dimensional space and grouping efficiency evaluation. The Spearman correlation coefficients between the 10 coefficients evaluated were significant (P < 0.001) and indicated significant changes in genotype ranking due to type of coefficient used (0.76 <= r(s) <= 1). Wide variation was also observed in the efficiency of clustering methods, where the unweighted pair group method with arithmetic mean was the most suitable (0.3 <= D <= 1.5; 0.41 <= r(c) <= 0.77; 5.99 <= S <= 12.61). Projection efficiencies in two-dimensional space showed high-stress values (65 < S < 89%). Similar to the results observed for hierarchical clustering methods and for projection in two-dimensional space, the formation of groups with grouping methods by optimization showed variations when using different coefficients. We believe that the results confirm the influence of coefficients in studies of genetic diversity, showing the need to use criteria and standards for selecting appropriate methods for genetic studies of the genus Croton.