In order to compare their relative efficiencies as markers and to find the most suitable marker for maize diversity
studies we evaluated 18 tropical maize inbred lines using a number of different loci as markers. The loci used were:
774 amplified fragment length polymorphisms (AFLPs); 262 random amplified polymorphic DNAs (RAPDs); 185
restriction fragment length polymorphisms (RFLPs); and 68 simple sequence repeats (SSRs). For estimating genetic
distance the AFLP and RFLP markers gave the most correlated results, with a correlation coefficient of r = 0.87.
Bootstrap analysis were used to evaluate the number of loci for the markers and the coefficients of variation (CV)
revealed a skewed distribution. The dominant markers (AFLP and RAPD) had small CV values indicating a skewed
distribution while the codominant markers gave high CV values. The use of maximum values of genetic distance CVs
within each sample size was efficient in determining the number of loci needed to obtain a maximum CV of 10%. The
number of RFLP and AFLP loci used was enough to give CV values of below 5%, while the SSRs and RAPD loci
gave higher CV values. Except for RAPD, for all the markers genetic distance correlated with single cross
performance and heterosis which showed that they could be useful in predicting single cross performance and
heterosis in intrapopulation crosses for broad-based populations. Our results indicate that AFLP seemed to be the
best-suited molecular assay for fingerprinting and assessing genetic relationships among tropical maize inbred lines
with high accuracy.