The figure to the left comes from a short paper in The American Journal of Human Genetics, A Revised Root for the Human Y Chromosomal Phylogenetic Tree: The Origin of Patrilineal Diversity in Africa. The paper is interesting because of two factors: 1) they sequenced more of the Y chromosome 2) their African data set of individuals was very large, in excess of 2,000. The weird thing about the results is that it upends one of the truisms in human phylogenetics: that African Pygmies and Khoisan are basal to other human lineages. By this, I mean that they “split off” first (this is why people say these are the “oldest” human populations). This is not what these researchers found. Rather, the basal Y chromosomal haplogroups were concentrated in Central West Africa and Northwest Africa! The map below shows the distributions of the two most divergent Y chromosomal lineages, red being the outgroup, and green being the second most divergent:
A second major point in this paper is that they recalculated the coalescence back to the last common ancestor of these Y chromosomal lineages, and doubled that value, to ~140,000 years. I don’t know the technical details of the calibration well here, last I checked the dates for Y chromosomal lineages were a total mess. So I’m not going to express confidence in this specific value, but, it does align well with a suspicion among many people that the idea of modern humans all tracing back to ancestors on the order of ~50,000 years B.P. just isn’t tenable any longer (this excludes the issue of possible admixture with other lineages such as Neandertals).
Dienekes has a lot more to say. I’m kind of left scratching my head. Remember that this is the male lineage of humanity. Who knows what peculiarities there are here?
Here’s their conclusion:
In conclusion, we present here a Y chromosome phylogenetic tree deeply revised in its root and earliest branches. Our data do not uphold previous models of Y chromosomal emergence…and demand a reevaluation of some fundamental ideas concerning the early history of the human genetic diversity we find today….Our phylogeny shows a root in central-northwest Africa. Although this point requires further attention, we think that it offers a new prospect from which to view the initial development of our species in Africa.
Razib Khan’s degrees are in biochemistry and biology. He has blogged about genetics since 2002, previously worked in software development, is an Unz Foundation Junior Fellow and lives in the western US. He loves habaneros.
May 21st, 2011 at 6:48 am
I do not think that the Berber presence is of great meaning regarding “Adam”: both divergent basal lineages A1b and A1a are obviously much more common south of the Sahara and the remainder lineage pulls the second branch (A1a-T) towards the East and South (with quite a lot of strength).
I can think therefore on a Chad Lake (so to say) origin of the Y-DNA of modern humankind but more generically a Sahel (and northern jungle) corridor, between Ethiopia and as far West as Senegal, where the Fulani may well be some of the retainers of such deep roots.
What the presence of these lineages in NW Africa emphasizes is that there are surely very old roots also there, from a time when those lineages were still much more frequent South of the Sahara. There has been a tendency to dismiss mtDNA L(xM,N) in North Africa (c. 25%) as recent arrivals (typically slave trade, same with Arabia Peninsula) but I strongly suspect that they are at least in part very old lineages from around the time of the Out of Africa migration (maybe arrived with Aterian culture, maybe with Irhoud, who knows?) c. 125 Ka (or 90 Ka) ago. Similarly, I think that Y-DNA E1b1b1b1-M81 is probably very old in NW Africa (and that would explain that is seldom found in other places and its presence in relatively high frequencies in probably “undisturbed” populations of Paleolithic roots in NW Iberia, notably Cantabrians and Asturians). E1b1b1b1 must have been in NW Africa in the time of the Oranian (Iberomaurusian) genesis and its feedback into the Iberian Gravetto-Solutrean (winged, back-tipped points, of ultimately Aterian roots).
May 21st, 2011 at 11:10 am
i’m going “agnostic” on this issue until we get even better population coverage in africa. who knows what rare haplogroups we’re missing?
May 21st, 2011 at 11:41 am
[...] may have been an important process in the evolution of modern humans. Dienekes (here and here) and Razib Khan [...]
May 22nd, 2011 at 10:54 am
My concern is the reliabiltiy of rooting trees in chimp sequences, especially since chimp Y-DNA has been mutating rather actively (http://www.nature.com/nature/journal/v463/n7280/full/nature08700.html). The authors already admitted than of the loci orthologous with chimps that was deemed stable is likely homoplastic (“By comparing a human sequence of 500 kb MSY flanking the M91 site to the orthologous sequence in the chimp, we have observed that less than one-third of the 9-Ts homopolymeric stretches have retained the same length in the two species (15/56 human versus chimp, 15/53 chimp versus human, see Table S4), which is consistent with the hypothesis that this marker might be evolutionarily unstable.”).
It’s rather odd that the human loci orthologous with chimps’ would wait for millions of years to mutate and would only do so with modern humans leaving Africa (or diversifying within Africa). It would make more sense if there were several recurrent mutations accompanying the branching of hominids in and out of Africa. We need ancient DNA from African AMH and archaic specimens to ascertain that those loci were indeed passed down to living humans in their pristine chimp-like condition. Outside of Africa we know that human populations sometimes share mutations with hominids (Neanderthals and Denisovans) but not with chimps.
May 23rd, 2011 at 3:08 pm
“it does align well with a suspicion among many people that the idea of modern humans all tracing back to ancestors on the order of ~50,000 years B.P. just isn’t tenable any longer”
This study makes a case a much younger age for the Y-DNA haplogroups of the vast majority of people relative to prior studies by a factor of approximately two. It puts a split of CT at something on the order of 39,000 years rather than 70,000 years in prior estimations. This makes the age estimates for the Y-DNA hgs found in Papuans and Australian Aborgines inconsistent with the hard evidence to support that the broad outlines of some of that hg mix must have been present 45,000 years ago when those hgs supposedly didn’t exist by this method.
The phylogeny is really only slightly different, finding that A1b is much more different from and older than A1a than previously realized, when both are rare to start with, but the age estimates are much younger except for A1a which is essentially squeezing everything else in the phylogenetic treee to younger dates.
In short, there are all sorts of reasons to be deeply skeptical of the absolute ages found in this study, a long standing problem with Y-DNA mutation dating that is at best miscalibrated and at worst fundamentally based on flawed assumptions about the homogeneity of mutation rates.
If there is a simple calibration problem, and one doubles all the dates, one starts to get dates that look more reasonable across the board, and the TMRCA of all men is ca. 280,000 years ago, which is still well within the range of the possible for the earliest anatomically modern human. Also, given the immense amount of migration that has taken place even within Africa in the time frame involved, accompanied by some known major population replacements and probably others buried in pre-history, the presence of old Y-DNA hgs in populations today is particularly uninformative when it comes to telling us where those hgs originated.