Comparison of AFLP and SSR markers for genetic diversity studies in maize populations

Using a set of 18 maize populations belonging to six different agro-ecological groups we report on a comparison of AFLP and SSR techniques to evaluate their applicability for the study of genetic diversity. Also, a DNA-pooling strategy (bulked samples) that proved to be effective in RFLP analysis of maize populations was employed in order to certify if it could be applied with the same success with SSR and AFLP markers. Twenty-five SSR markers generated a total of 224 polymorphic bands, while three AFLP primer combinations generated 188 bands with 79.2% being polymorphic. The average polymorphism content (PIC) was higher for SSRs, i.e. 0.79 for SSRs versus 0.28 for AFLPs. Higher genetic similarity (GS) values were estimated for AFLPs. The Pearson's correlation coefficient calculated for SSR and AFLP genetic similarities was low (0.149) indicating a poor agreement between the two marker systems. The ranking of the agroecological groups according to their mean GS values (the average GS calculated for the three populations representing each group) was congruent for both marker systems and the Pearson's correlation coefficient calculated using these values was 0.853. The cluster analysis did not group populations precisely with expectations based upon their agro-ecological groups, although a much better agreement was accomplished with AFLP analysis. Based on the results in this work it could be concluded that, when using a DNA-pooling strategy, AFLPs are more reliable in assessing genetic diversity of maize populations