>>14412041(1) The haplotypes (gene variants) mentioned are all on one special chromosome, the human Y-chromosome (this is the sex chromosome that determines gender). The Y-chromosome doesn't recombine. So variants on it don't give independent sources of evidence. Just like 'mitochondrial eve', the ancestral type is just the most recent ancestor that had that chromosome. Though it may tell us something about the 'most recent common ancestral' human being that had that particular chromosome, the particular ancestry can and does vary for the many other bits of the human genome. For a clear picture of human ancestry, you therefore need to examine many independent genetic linkage groups.
(2) The authors do not clearly state what method they use (the PHYLIP package includes many methods, that are mostly concerned with inferring phylogeny, rather than population ancestry). More modern coalescence methods would probably be more appropriate.
(3) It is not clear if the tree presented here is rooted correctly. This can have a big effect on interpreting what evolved from what.
(4) The authors really overextend what can be said from tree shape -- there is no clear logic behind the splitting of haplotypes that they discuss, and the biogeographic history they suggest for the descendent lineages. Their geographic reconstructions are simply not supported by their data.
But even if their methodology was OK, and their interpretation of the genealogy and biogeography correct (which is unlikely), this would still be a gene tree for only one linkage group, the Y-chromosome.
So, don't be misled by the reference to European and African haplotypes. Haplotypes are not people, they are gene fragments, and the story they tell can bear little relationship to the ancestry of the people they belong to.
