Further characterization of AFLP® data as a tool in genetic diversity assessments among maize (Zea mays L.) inbred lines

AFLP® markers generated by CNG methylation sensitive (PstI/MseI) and CNG methylation insensitive (EcoRI/MseI) enzyme combinations and AFLP markers collected from hypomethylated (PstI/MseI) and hypermethylated (mPstI/MseI) regions were compared for their polymorphism information content, sampling variance and patterns of genetic diversity in a representative sample of 33 inbred lines of maize (Zea mays L.). We demonstrate that the mean polymorphism information content generated by sets of PstI/MseI and mPstI/MseI markers (0.38) is significantly higher than by sets ofEcoRI/MseI markers (0.33). Also the sampling variance highlighted the distinctive nature of the (m)PstI/MseI markers: to achieve a mean standard deviation of 5% in the estimation of genetic distance among the 33 inbreds, the PstI/MseI and mPstI/MseI marker sets (135 and 129 markers, respectively) are clearly smaller than the EcoRI/MseI marker set (173 markers). A further minimizing of the sampling variance of AFLP data in the estimation of genetic similarities was obtained by reducing marker information redundancy by selecting markers evenly distributed over each chromosome: a set of only 106 AFLP markers, sampled conditionally on their genetic map position, was required for a mean standard deviation of 5% in the estimation of genetic distance among the 33 inbreds.