Complex Signals for Population Expansions in Europe and Beyond

François Jacob, in his brilliant 'The possible and the actual' (Jacob 1982), reminds us that 'scientific investigation begins by inventing a possible world, or a small piece of a possible world'. One may add that the space allowed for the possible is likely to be in strong positive correlation with the level of our ignorance. What chance, then, when discussing language / farming / gene dispersals, do we have to identify the actual from a plethora of possible scenarios? Since the presentation of 'African Eve' (Cann et al. 1987; Vigilant et al. 1991), the last decade has demonstrated an increasingly better understanding of the phylogeny and phylogeography of mtDNA and of the Y chromosome. Here, the first influential achievement was a series of papers from Emory (reviewed in Wallace 1995) where, inter alia, it became obvious that human maternal lineages world-wide are very clearly structured geographically. This knowledge came thanks to phylogenetic analysis of the coding part of the mtDNA genome. Secondly, as Richards et al. (1996) have shown, the mtDNA hypervariable 1 (HVR 1) region offers an increased resolution of a phylogenetic tree, in particular as far as Europeans are concerned. Although mtDNA hypervariable region sequences started to accumulate in quantities (thanks largely to forensics), it soon became obvious that the results coming from RFLP analysis or the HVR sequence(s) alone were not informative enough to go further. Quite the opposite; it became clear that trees, based on HVR 1 sequence alone, were often phylogenetically wrong. However, a synthesis of what is known about polymorphisms in the coding region (extensive RFLP as a tool) and HVR (direct sequencing) removes most of the ambiguities and leads to a much better understanding of the details of the topology of the phylogenetic tree of mtDNA (e.g. Macaulay et al. 1999). This analysis owes much to the use of median networks as an approach (Bandelt et al. 1995). In this contribution we demonstrate that coalescence age calculation of the monophyletic branches of the mtDNA phylogenetic tree, applied together with a detailed phylogeographic knowledge, is an instrument which provides new insight into demographic processes of the past and, in particular, allows to see informative differences there, where mere haplogroup frequency calculations are able only to register flat landscapes.