Here's another picture of HMP. I lifted this off of Maju's blog, Leherensuge, which is full of neat stuff like this. Thanks, Maju.
(Click to enlarge.)
What we see here is a phylogenetic tree somewhat similar in content (though different in design) to that displayed in my previous post, in that it represents mitochondrial DNA (mtDNA), i.e. female lineages only. But in this case, the M haplogroups, in red, and the N haplogroups, in blue, are greatly expanded, so we can get a good look not only at HMP, but many of its descendants throughout the world.
Despite the fact that N appears lower down on this map than M, it's important to recognize that both are "sister" clades, branching off from the same "mother": L3 -- now found only in Africa. If M and N are sisters, does this mean there were two different Out of Africa migrations? I don't know enough about population genetics to get into this question very deeply, but perhaps Maju, who knows a lot more than I, will be willing to chime in with some suggestions. I found an interesting reference to this issue in a paper by Luca Cavalli-Sforza and Marcus Feldman, The Application of Molecular Genetic Approaches to the Study of Human Evolution, dating from 2003.
After referring to the widely accepted theory that "the spread of humans into Asia occurred through two routes," a southern route along the Indian Ocean coast, and a "central route through the Middle East, Arabia or Persia to Central Asia, from where migration occurred in all directions . . .", they note that "it is still unresolved whether the divergence between these two expansion routes occurred in Africa or after entry into West Asia . . ." (p. 270). If the former is the case, then there would have been two Out of Africa migrations, and therefore two different HMPs, one carrying M and the other N. If the latter is the case, it looks as though there was only one such migration, of an L3 group which produced its M and N branches only after HMP crossed over into what is now Yemen. This is all, of course, highly speculative still, but much is nevertheless generally accepted as what the authors refer to as the "standard model of modern human evolution" (p. 270).
Having established, however provisionally, a baseline for the consideration of the physical and cultural nature of the population ancestral to M, N, and L -- what I've been calling HBP and HBC -- we are now in a position to speculate further regarding the physical and cultural characteristics of their descendants, the Out of Africa migrants. This will not be anywhere near so easy, however. If you look carefully at the phylogenetic tree reproduced above, you'll notice an important difference between the uppermost branches, labeled L, and those lower down, dominated by M and N. Though not indicated on the diagram, the deepest branches, especially L0, L1, and L2, are dominated by Bushmen groups and both of the major Pygmy groups, Eastern and Western. The simplicity of this picture, along with the equally clear musical evidence, has made it relatively easy to characterize HBC on the basis of what we know about these three groups, using the triangulation method already described. The many branches emanating directly from M and N present a very different picture, however, with far less structure. Unlike the situation with L, it's not at all clear where we can look if we wish to triangulate, or even quadrangulate or quintangulate (have I just invented two new words?).
We can, of course, look backward to HBC for signs that certain ancestral features might have survived in HMC. But we will also need to draw on what we know about a great many indigenous societies in many different parts of Asia, Oceania, Australia and Europe as well. And since there are so many differences among so many of these groups, our task will not be easy. It is nevertheless, in my opinion, possible to develop some very interesting and to some extent testable hypotheses regarding both HMP and HMC, and to the extent that we can do so, we will have established another important baseline for the study of human history.
Having established, however provisionally, a baseline for the consideration of the physical and cultural nature of the population ancestral to M, N, and L -- what I've been calling HBP and HBC -- we are now in a position to speculate further regarding the physical and cultural characteristics of their descendants, the Out of Africa migrants. This will not be anywhere near so easy, however. If you look carefully at the phylogenetic tree reproduced above, you'll notice an important difference between the uppermost branches, labeled L, and those lower down, dominated by M and N. Though not indicated on the diagram, the deepest branches, especially L0, L1, and L2, are dominated by Bushmen groups and both of the major Pygmy groups, Eastern and Western. The simplicity of this picture, along with the equally clear musical evidence, has made it relatively easy to characterize HBC on the basis of what we know about these three groups, using the triangulation method already described. The many branches emanating directly from M and N present a very different picture, however, with far less structure. Unlike the situation with L, it's not at all clear where we can look if we wish to triangulate, or even quadrangulate or quintangulate (have I just invented two new words?).
We can, of course, look backward to HBC for signs that certain ancestral features might have survived in HMC. But we will also need to draw on what we know about a great many indigenous societies in many different parts of Asia, Oceania, Australia and Europe as well. And since there are so many differences among so many of these groups, our task will not be easy. It is nevertheless, in my opinion, possible to develop some very interesting and to some extent testable hypotheses regarding both HMP and HMC, and to the extent that we can do so, we will have established another important baseline for the study of human history.
"The simplicity of this picture, along with the equally clear musical evidence, has made it relatively easy to characterize HBC on the basis of what we know about these three groups, using the triangulation method already described. The many branches emanating directly from M and N present a very different picture, however, with far less structure."
ReplyDeleteThis is, IMO, one of the attributes of a wrong historical scenario: if "older" states are better "known" than "younger" states, it means history is being written in reverse.
"Despite the fact that N appears lower down on this map than M, it's important to recognize that both are "sister" clades, branching off from the same "mother": L3 -- now found only in Africa. If M and N are sisters, does this mean there were two different Out of Africa migrations?"
First, even in the existing tree topologies M and N are sister clades of African-specific L3 lineages. This is what your image demonstrates. Second, both M and N lineages are found in India, Australia and the Americas, suggesting that, if one accepts (for a second) the out of Africa theory, there was only one migration out of Africa and (since M and N lineages are found in every corner of the Americas) there was only one migration into the Americas. Archaeologically, however, there's no evidence for a migration out of Africa or for a migration into the Americas. As Luis, Terry and I discussed over at Dienekes(http://dienekes.blogspot.com/2009/11/y-chromosome-diversity-human-expansion.html), the earliest technological complex associated with modern human behavior (call it Aurignacian) shows older dates in West Asia and South Siberia and younger dates in Europe and North Africa (Dabban). A migration into Africa evident in the distribution of the Y-DNA CFDE clade may be the genetic correlate of the spread of this technology into Africa. Then it's traces get lost but at least we have a good chronological and technological link between earlier non-African populations and later African populations.
So far we're lacking some essential evidence to support geneticists' claim that there was any migration out of Africa whatsoever.
German: "This is, IMO, one of the attributes of a wrong historical scenario: if "older" states are better "known" than "younger" states, it means history is being written in reverse."
ReplyDeleteOne would assume so, yes. Which is why scientific research should not be based solely on assumptions.
"both M and N lineages are found in India, Australia and the Americas, suggesting that, if one accepts (for a second) the out of Africa theory, there was only one migration out of Africa"
I appreciate your point, it does make sense, but I'd be curious to learn what Maju thinks of this possibility. As I see it, there is not presently enough evidence for us to fully understand what happened so far back in history. Which is why it's especially interesting, for me at least, to see if we can recreate a coherent picture of HMP and HMC. If not, then we might have to consider two OOA migrations. Or, as you would prefer: none.
"Archaeologically, however, there's no evidence for a migration out of Africa or for a migration into the Americas."
It seems to me that there is a great deal of archaeological evidence consistent with Out of Africa. For one thing, there is a long list of modern human remains in Africa going back to well over 100,000 years ago, far older than in Asia and far far older still than in the Americas. Even the multi-regionalists have been forced to acknowledge that.
"So far we're lacking some essential evidence to support geneticists' claim that there was any migration out of Africa whatsoever."
To my knowledgte you are the only one who holds that view. Even the multiregionalists concede that there is very strong genetic evidence for an out of Africa migration by modern humans.
"To my knowledgte you are the only one who holds that view. Even the multiregionalists concede that there is very strong genetic evidence for an out of Africa migration by modern humans."
ReplyDeleteVictor, you have a very stereotypical perspective on what's out there in specialized literature. Most archaeologists actually would agree with me. People can see technological continuity everywhere (South Siberia, Western Europe, you name it), but nothing suggestive of a replacement out of Africa. There's a vague sense, among evolutionists, that out of Africa is the most parsimonious explanation of fossils (anatomically modern humans) and genes, but if it comes down to actually pinpointing archaeologically how and when people actually left Africa at about 60-50K and how and when they entered Europe and Asia, I think everyone would concede there's very little evidence to this effect. A movement into Africa at 40K is easier do demonstrate both genetically and archaeologically.
"One would assume so, yes. Which is why scientific research should not be based solely on assumptions."
You are the one making assumptions.
Thanks, Maju.
ReplyDeleteYou're welcome. Glad that you found it useful.
The peculiarity of this tree is that branches reflect the length of the haplogroup-defining mutations (per PhyloTree), however I did not make any difference between coding and control region mutations and I have been chewing on redrawing it using only coding region ones, because the control region (also known as hyper-variable region or sequence) mutates much faster and is sometimes ambiguous. But from my preliminary notes the results should not be too different in the overall picture.
Also, from memory, it's possible that one or two minor M clades have the wrong geographic description (not sure if I corrected that). Finally it's worth noticing that L6 is found in at least as high frequencies in Yemen as in Africa (Ethiopia only), though the coalescence of the clade at either side of the Red Sea has not been determined (diversity is similar at both sides).
Despite the fact that N appears lower down on this map than M, it's important to recognize that both are "sister" clades...
This is because the root is at the left. If you think convenient, you can rotate the image and place it at bottom or top as you wish.
Notice also that R (in green) is part of N (blue) but marked in different color because of its unusual spread.
If M and N are sisters, does this mean there were two different Out of Africa migrations? -
ReplyDeleteIt is of course a possibility but, as German points out, the distribution of these two lineages is so extremely intertwined that it's much more likely that the two participated in a singe migration and/or they converged before the great Eurasian expansion, which was some time after the OoA as such.
In my opinion some L3 people were at the origin of all Eurasians via maternal lineages and these "amorphous" L3 evolved into M and N probably already in Asia (meanwhile L3 evolved into five other clades in Africa).
However I conjecture that, while M may have boomed (and what a boom!) in South Asia, N appears to have a more easterly center of gravity, so maybe it starred a secondary expansion in SE Asia while some of the early M lineages expanded there too. Whatever the case, South and SE Asia (the latter projecting into East Asia and Sahul) seem to have been interacting dynamically in the early stages of Eurasian expansion.
Maybe the most intriguing feature is macro-haplogroup R, which has highest diversity (and hence likely origin) in South Asia and that would seem to be the main clade involved in the colonization West Eurasia (inhabited by Neanderthals, and hence less immediately available initially). However R also sent sublineages to the East (pre-F, pre-B, P, etc.), so it shows expansion in all directions, probably at a moment when there was already some density in the Tropical Asian belt.
In general all lineages appear to show southernly coalescence areas (highest top tier diversity) within Asia. The main exceptions is (N-derived) haplogroup A but this one also shows a rather long stem, indicative of a long period of coalescence at low numbers before expansion, via founder effect, in NE Asia.
Anyhow, I must mention that some people have argued for N expanding before M (what I find unlikely but well) and also (related) for an average age of South Asian M clades that is younger than East Asian and Sahulian. But this last seems caused by the fact that the estimated ages of all clades were averaged and can't itself shed any light on the origins and timing of the M node, as different subclades may have expanded (become consolidated by expansion of their numbers) at different times.
Some people have also argued, rather weakly, for a different migration pattern for both lineages, with N taking the northern steppary route. Considering that N derivatives seem very old in Australia and that all nodes seem located rather to the south, I think this is a wrong interpretation of the data.
Another issue is Y-DNA and putting together the two halves of the story. However, for the three main Y-DNA lineages of Eurasia, two (C and D) seem centered in SE Asia, while the largest one (F) is clearly centered in South Asia, somehow confirming the "rapid coastal migration" model.
Though not indicated on the diagram, the deepest branches, especially L0, L1, and L2, are dominated by Bushmen groups and both of the major Pygmy groups, Eastern and Western.
ReplyDeleteThe smaller haplogroup L5, which is also quite deep, is also reported to be found among the Mbuti (15%) and at smaller frequencies in East Africa (Sandawe and others).
On the other hand I would not correlate L2 tightly with the most ancient groups, because it's a very important haplogroup among Niger-Congo peoples (or Mande, see below), particularly the large haplogroup L2a and its subclade L2a1. I suspect that L2 expanded at about the same time as L3 did (see how they branch out at the same level in this tree). It is argued to have also an East African origin but it also seems to represent a process of colonization of West Africa.
I was speculating (private notes) that maybe it signals an expansion leading to the Mande linguistic group, which is not confirmed to be Niger-Congo but nowadays treated as a separate language family. If so, the lineages leading to Niger-Congo would be rather from L3, specially L3b and L3d, though of course, these two populations have interacted rather horizontally since long ago, diluting these differences.