Abstract
A major drawback of Amplified Fragment Length Polymorphisms (AFLP) as genetic makers for phylogeographic studies is their lack of a temporal dimension. In a recent publication in Molecular Ecology, Kropf et al. (2009) proposed a molecular clock for AFLP. In this comment we evaluate the proposed approach both theoretically and empirically. A linear increase with time is a prerequisite to use a genetic distance as molecular clock. Testing the relationship between genetic distance and time in the data of Kropf et al. (2009) for linearity revealed that the relationship was in fact not linear for their pooled data, as well as for one of the three species analyzed. Also, the relationship was not linear in two new species, where divergence times could be inferred from macrofossils. When applying the proposed molecular clock to data from eight species, dates obtained were plausible in some cases, but very improbable in others. The suggested genetic distance was also influenced by intrapopulation genetic diversity, leading to a potential bias. In the future, investigations of AFLP mutation rates combined with phylogeographic modelling may contribute to adding a time scale to the understanding of AFLP data.