Direct evidence that restriction endonucleases may under estimate the degree of divergence between molecules.

We studied two polymorphic forms of mtDNA extracted from A. lixula eggs. In order to compare and to quantitate the variability, we sequenced specific regions of the two molecules. In this way, we obtained a precise measurement of the variability within two haplotypes. We also obtained a direct demonstration that some differences in nucleotide sequence can escape detection when restriction endonuclease analysis is used. Our results underline the unreliability of the use of restriction mapping to estimate divergence between relatively short and closely related DNA sequences.