The knowledge of genetic distances among individuals or populations in plant breeding programs can be used to complement phenotypic information and to help the selection of individuals to breed towards segregating populations with high variability. The objectives of this study were to identify primers that produce polymorphisms among cotton (Gossypium hirsutum L.) genotypes, to compare the effect of primer length in the generation of polymorphisms, to determine genetic distances among 24 cotton

more »

... genotypes from eight countries using the RAPD technique and compare them with results obtained through phenetic analysis of the genotypes based on 52 qualitative and quantitative characteristics using numerical taxonomy. The 24 genotypes were evaluated by the RAPD technique utilizing short primers and combinations of long primers. RAPDs were scored as presence (1) or absence (0) of prominent bands reproducible in duplicated PCR reactions. These data were subjected to analyses to produce a dissimilarity matrix based on Simple, Nei and Li's and Jaccard's Distances. UPGMA was utilized to create dendrograms based on the genetic distance matrices. Short primers produced polymorphic bands that were efficient to discriminate among genotypes, while long primers couldn't detect differences among the genotypes. Comparing the three methods used for estimation of genetic distances, the method that better dealt with distances obtained by using phenotypic character analyses was the Simple Distance Method. The correlation among the results of this method with Jaccard's and Nei and Li's Distances were 93 % and 91 %, respectively. The maximum simple distance found between pairs of genotypes based in the polymorphisms produced by short primers was 18%, which possibly indicates a narrower genetic basis than in the phenotypic analysis. The correlation among phenetic analysis and RAPD based on short primers was moderate and negative, tough significant (r=-0.43).