RSS

Kerguelen mice: a model for human expansions?

27 Oct
I will today stop at a new genetic paper that does not deal with humans but with a not-so-distant relative: house mice. Specifically with mice in the remote Kerguelen islands, which have been known to humans (and hence to mice) only since 1772.

Abstract (provisional)

Background

Starting from Western Europe, the house mouse (Mus musculus domesticus) has spread across the globe in historic times. However, most oceanic islands were colonized by mice only within the past 300 years. This makes them an excellent model for studying the evolutionary processes during early stages of new colonization. We have focused here on the Kerguelen Archipelago, located within the sub-Antarctic area and compare the patterns with samples from other Southern Ocean islands.

Results

We have typed 18 autosomal and six Y-chromosomal microsatellite loci and obtained mitochondrial D-loop sequences for a total of 534 samples, mainly from the Kerguelen Archipelago, but also from the Falkland Islands, Marion Island, Amsterdam Island, Antipodes Island, Macquarie Island, Auckland Islands and one sample from South Georgia. We find that most of the mice on the Kerguelen Archipelago have the same mitochondrial haplotype and all share the same major Y-chromosomal haplotype. Two small islands (Cochons Island and Cimetiere Island) within the archipelago show a different mitochondrial haplotype, are genetically distinct for autosomal loci, but share the major Y-chromosomal haplotype. In the mitochondrial D-loop sequences, we find several single step mutational derivatives of one of the major mitochondrial haplotypes, suggesting an unusually high mutation rate, or the occurrence of selective sweeps in mitochondria.

Conclusions

Although there was heavy ship traffic for over a hundred years to the Kerguelen Archipelago, it appears that the mice that have arrived first have colonized the main island (Grande Terre) and most of the associated small islands.The second invasion that we see in our data has occurred on islands which are detached from Grande Terre and were likely to have had no resident mice prior to their arrival. The genetic data suggest that the mice of both primary invasions originated from related source populations. Our data suggest that an area colonized by mice is refractory to further introgression, possibly due to fast adaptations of the resident mice to local conditions.

The abstract alone is quite explanatory.
I must say that the reason for the refraction may not be adaptations to local conditions as much as mere demographic pressure: any new mice would be automatic minority and have poor chances of perpetuation, even in the absence of adaptations. Numbers alone make the difference, proving statistically difficult for any new arrival to leave a mark after the first population is consolidated.
Another detail worth discussing is:
In the mitochondrial D-loop sequences, we find several single step mutational derivatives of one of the major mitochondrial haplotypes, suggesting an unusually high mutation rate, or the occurrence of selective sweeps in mitochondria.
The mutation rate does not look too striking to me: 200 years for mice is like 10,000 for us maybe, as they reach reproductive maturity in a matter of months. So it’s like one (or two in a few sub-lineages)  surviving mutations downstream of the founder haplotype, which is anyhow still massively dominant.

Fig.2A mtDNA HVS haplotype network
In other words: there were some distinct founder effects of some derived haplotypes, mostly in Kerguelen itself but also overseas: in Falkland and Auckland islands and even back to Europe (Britain).
The diversity of derived basal lineages clearly indicates that this Kerguelen haplotype (yellow) exploded in the islands, regardless that it may be original from mainland Africa (black) and regardless that it has also spread overseas. This in the equivalent of 10,000 human years (very roughly).
It reminds me somewhat of what we can see in Eurasian human mtDNA, with the yellow lineage resembling human mtDNA M somewhat and the magenta one like N. And that is one reason I wanted to comment this paper, really. Of course the stage of diversification and other aspects are clearly different but it does illustrates how founder effects proceed and how demic pressure alone tends to fend off new colonizations.
Advertisements
 

99 responses to “Kerguelen mice: a model for human expansions?

  1. terryt

    October 27, 2010 at 10:08 pm

    "It reminds me somewhat of what we can see in Eurasian human mtDNA, with the yellow lineage resembling human mtDNA M somewhat and the magenta one like N". Exactly. And note: separate regions. "So it's like one (or two in a few sub-lineages) surviving mutations downstream of the founder haplotype, which is anyhow still massively dominant". So doesn't that suggest that mtDNA M and N also coalesced in separate regions? The two may have left Africa together, but they must have soon parted company and not moved through South Asia together.

     
  2. Maju

    October 27, 2010 at 11:35 pm

    "So doesn't that suggest that mtDNA M and N also coalesced in separate regions?"Yes. I have sustained for long that M coalesced in South Asia and N in SE Asia. The shortest line between SE Asia and the Horn or the Nile is via South Asia (it is also the only one with a warm climate). Anyhow, M and N must have arrived to SEA roughly about the same time, because N did not impede the advance of M in SEA and surroundings (NEA, Oceania). It is also interesting that both mice populations have the same Y-DNA lineage anyhow.

     
  3. terryt

    October 28, 2010 at 2:05 am

    "Anyhow, M and N must have arrived to SEA roughly about the same time, because N did not impede the advance of M in SEA and surroundings" But M sure impeded N's survival in South Asia. I have to confess I have consistently carried one preconception. The conception that humans would not be too different from all other species. I'd like to draw your attention to something about your stars that you don't seem to have noticed for yourself. In all cases, but one, all the members of the star can be arranged in geographically connected regions. Haplogroup A through northern and eastern Asia. B from the Philippines and Tengarra through most of east and SE Asia. P through New Guinea. R from New Guinea through South Asia to Europe. M through Southern Eurasia all the way from SW Asia, through South Asia and on to Australia, New Guinea and East Asia. Similarly with all the others. Except there's a huge gap between eastern and western Ns. Unless …

     
  4. Maju

    October 28, 2010 at 2:11 am

    "But M sure impeded N's survival in South Asia".Not enough. It's not just on/off: in the early period there may have been a number of lineages, you know. Some survived and became so dominant that could not be moved, some went extinct and some managed to exile themselves into a more promising niche before suffering the same destiny. "Except there's a huge gap between eastern and western Ns. Unless" … Precisely.

     
  5. terryt

    November 1, 2010 at 3:54 am

    "Precisely". So how do you explain it?

     
  6. Maju

    November 1, 2010 at 8:35 am

    Unless there's no such huge gap because all Western N sublineages (except X) are also found in South Asia (India and such). When a lineage is removed geographically, it's probably because it traveled all that as a minor lineage. You have exactly the same problem with L3 and that's where Arabia and South Asia have to be integrated: first because they are in the middle of the only logical route and second because you see archaeological and genetic evidence in those regions as well. That's it. I think you're wasting your time and mine, sincerely.

     
  7. terryt

    November 2, 2010 at 7:51 am

    "Unless there's no such huge gap because all Western N sublineages (except X) are also found in South Asia (India and such)". Possibly so. But in downstream mutations only, so they've come in, not originated there. "When a lineage is removed geographically, it's probably because it traveled all that as a minor lineage". If it travels as a minor lineage surely it's unlikely to survive. "You have exactly the same problem with L3" Not really. You have N haplogroups N1'5 and N2/W, and M haplogroup M1, in SW Asia. So yes, that is right on the only logical route, and you see archaeological and genetic evidence there as well.

     
  8. Maju

    November 2, 2010 at 3:32 pm

    "But in downstream mutations only, so they've come in, not originated there".That's a pseudo-reasoning. The same reasoning applies for the SA and WEA branches, and there are no others, so it's undetermined. There's no root (underived) haplogroups anymore, we are always looking at derived branches. "If it travels as a minor lineage surely it's unlikely to survive".Unlikely is not the same as impossible. There were surely many minor lineages which did not survive but some did indeed. "Not really. You have N haplogroups N1'5 and N2/W, and M haplogroup M1, in SW Asia"…And X, don't forget X, which is the only so phylogenetically high WEA-only lineage comparable with M1. N1'5 and N2 are shared with SA. But it doesn't matter. Because what we do (what I do, what most geneticists do, what most other amateurs do, and they do correctly) is to consider ALL basal sublineages and not some arbitrarily chosen. You do differently but that's the core failure in your "method" (or lack of it thereof). You cannot understand a haplogroup without looking at all its parts, the same that the blind man could not figure out the elephant by just gouging its trunk/tail/leg. And while the blind men of the tale were justified by the limits that blindness (and the narrator) imposed on them, you do not have such limitations so it is an error on your side. A most grave and conscious error and why I accuse you of pseudoscience.

     
  9. terryt

    November 3, 2010 at 3:09 am

    "consider ALL basal sublineages and not some arbitrarily chosen". Maju. It is you who is not considering ALL the basal lineages. Remember A? And N9/Y? "Unlikely is not the same as impossible". True. But you're claiming the 'nearly impossible'. "The same reasoning applies for the SA and WEA branches" Exactly. And I have given them due consideration. The same reasoning aplies to them all. Branches with long tails are very likely to have remained in the one isolated region for some time, rather than having migrated huge distances as minor representatives within larger populations.

     
  10. Maju

    November 3, 2010 at 7:51 am

    "It is you who is not considering ALL the basal lineages. Remember A? And N9/Y?"Your insistance on those two lineages (out of 12) is what I call "arbitrarily chosen". I am considering them all the time but they are just two out of 12: 1/6 of the overall picture. "But you're claiming the 'nearly impossible'".I don't see it that way and I don't see how would this change in your hypothesis. There's almost always some haplogroups that are distant from the overall origin of their parent haplogroup. From M and N in relation to L3 to X2g in relation to overall X2 (and so on). So it does happen and rather often. "Branches with long tails are very likely to have remained in the one isolated region for some time"…That's a nonsense: those are clearly minor branches that have been bouncing around, they are not any useful reference. Except when you consider all subclades together, then they weight as any other.Much more logical is to pay special attention to the branches with very short stems, as those expanded almost at the same time as (immediately after) the parent haplogroup did, so their track is more directly related if any. "… rather than having migrated huge distances as minor representatives within larger populations".There are so many possible stories about each haplogroup of that kind that we cannot tell them all in all our lives. Maybe it was part of a somewhat larger population that went almost extinct and was absorbed, maybe they were pushed around by more powerful bands till they found a place where nobody else had arrived, maybe they were 1/30 and by chance became, after many generations, 30/30. Maybe… there are infinite possible stories but I am not a novelist so I have to let your imagination flow… Anyhow neither A nor N9 are so clearly detached from SE Asia: they are found all the way from SEA to NEA and the case is not very clear anymore for a NEA origin for them.

     
  11. terryt

    November 4, 2010 at 6:23 am

    "Maybe it was part of a somewhat larger population that went almost extinct and was absorbed, maybe they were pushed around by more powerful bands till they found a place where nobody else had arrived, maybe they were 1/30 and by chance became, after many generations, 30/30. Maybe…" Maybe, maybe, maybe. It's still most likely that they eventually expanded from the region they first became isolated in.

     
  12. Maju

    November 4, 2010 at 7:28 am

    No. That's just one of the zillion possible scenarios. When the stem is long, the track is cold and you can't say much. Full stop.

     
  13. terryt

    November 5, 2010 at 4:00 am

    "That's just one of the zillion possible scenarios". OK. Let's turn to your scenarios: "Maybe it was part of a somewhat larger population that went almost extinct and was absorbed" That doesn't shift them one metre from where they are found today. "maybe they were pushed around by more powerful bands till they found a place where nobody else had arrived" If they were being 'pushed around by more powerful bands' it's unlikely they'd be able to find a previously uninhabited region. Some haplogroups are today found in mountainous regions, for example, but they are usually survivors of a more widespread population thay has been reduced elsewhere. It's not that they entered that region on their own to escape persecution. "maybe they were 1/30 and by chance became, after many generations, 30/30". Again that is most likely to happen in an isolated population. And haplogroup that is 1/30 is far more likely to become extinct rather than take over. Unless it develops some huge advantage the remainder of the population fails to adopt. "Anyhow neither A nor N9 are so clearly detached from SE Asia: they are found all the way from SEA to NEA and the case is not very clear anymore for a NEA origin for them" But mainly derived haplogroups are found in SE Asia. So your comment is a bit like saying that, because Y-hap R is very common and diverse in New Zealand it must have originated there. Especially as many men who are obviously of indigenous phenotype have the haplogroup. "those are clearly minor branches that have been bouncing around" Surely if they had been 'bouncing around' to any extent they are much more likely to have become extinct.

     
  14. Maju

    November 5, 2010 at 4:29 am

    "Let's turn to your scenarios"…It's not worth the effort because there are, as I said, a zillion possible cases. We only need one individual, maybe one small family, to trigger a founder effect where and when the conditions are good. "… it's unlikely they'd be able to find a previously uninhabited region".Notice that all these founder effects we are discussing here are in the last "frontier" areas of the Eurasian expansion: North and West Asia. Those frontiers allowed for founder effects that were impossible (or most difficult) elsewhere. My whole understanding of Eurasian colonization is one of expansion followed by the "closure" of areas and routes by mere demographic pressure, so in this stage only the remote and cold NE Asia (north of Beijing) and the Neanderthal territory of West Eurasia remained open for new founder effects. These two processes seem to have happened at about the same time, maybe indicating a strong pressure from the already colonized areas of the Tropical and middle East Asian zones pushing some peoples to find even more remote and difficult new places to live – in some cases at least with success. "Some haplogroups are today found in mountainous regions, for example, but they are usually survivors of a more widespread population thay has been reduced elsewhere".I'd have to assess each case on its own merits. They may well be localized founder effects too. If they are refuge areas, the lineage should be found, even if at lesser frequencies, all around, if they are only found in a particular mountain are it's probably a case of founder effect in the colonization of that particular area. "And haplogroup that is 1/30 is far more likely to become extinct rather than take over".Absolutely correct. But once in 30 cases (or whatever the precise figure) it will happen as I say. Getting three sixes in a three dice throw is most unlikely but happens anyhow. There were many many dices cast all around Eurasia, so some had to be triple sixes. This is an important understanding when we are dealing with probability. And we are."But mainly derived haplogroups are found in SE Asia".This is what I do not have so clear, because I have yet to see a comprehensive study on A and/or N9 for their whole area, certainly not in a language I can read. If you have clear data, please provide it."So your comment is a bit like saying that, because Y-hap R is very common and diverse in New Zealand it must have originated there".Nonsense: Modern migrations and genocides have nothing to do with what could happen in Paleolithic or almost any other pre-modern conditions. And you guys lack the key R2 half part of the issue, not to mention related Q and P*, so at most you'd be talking of R1. "… they are much more likely to…"Again it's just a matter of casting enough dices. Maybe 200 or 15,000 "private" haplogroups had to go extinct for X to make a founder effect in the Levant. We will never know the exact figures.

     
  15. terryt

    November 7, 2010 at 1:58 am

    "There were many many dices cast all around Eurasia, so some had to be triple sixes". Any other examples of a haplogroup with a long tail that can be shown to have come into a region as a minority within a wider population, and then having expanded on its own?

     
  16. Maju

    November 7, 2010 at 7:17 am

    Ok, I'll try to list all mtDNA haplogroups with a long tail (5 or more SNPs) and then a meaningful expansion:L0, L0a2, L0k, L0d, L0d1, L0d2L1"6, L1b, L1c, L1c1a2, L1c2, L1c3L5a, L5cL2'3'4'6, L2, L2a, L2b'c, L2b, L2c,L6L4a, L4b2,L3b, L3f1b, L3d, L3e1, L3e3, L3x, L3h1N, N1b, W, N9b, A, R1, T, R7, F2, F3, R11, B4c1b1, B4f, B5b, R31a, P4a, U1, U2b, U2e, U3a, U4, U8a M1, M51, M2b, M7a, M7b'd, M7b3, M8a1, C, M10, M11, M12, M13a, M28, M31a1, M41, M53, M73, D4b1a1, D4h3aThat's it I think (notwithstanding whatever errors I may have committed). Lots of them.

     
  17. terryt

    November 8, 2010 at 3:37 am

    "I'll try to list all mtDNA haplogroups with a long tail (5 or more SNPs) and then a meaningful expansion" Thanks. Now show me which of these cannot be assumed to have been in the region it expanded from over the whole period of that long tail. All the L0s you mention expanded from somewhere in southeast or east Africa. Presumably they'd originated there. All the L1s look to have been long established in Central Africa, perhaps northeast of the Congo basin. The L5s look to be long established somewhere near, or in, Ethiopia. All the L2s in the Sahel region. L6 somewhere between Egypt and Ethioia. And so on. Each of the haplogroups has expanded from somewhere within their main range today. They haven't migrated for huge distances before they expanded. And almost certainly none 'have come into a region as a minority within a wider population'.

     
  18. Maju

    November 8, 2010 at 5:33 am

    I will provide an example with L3: if your nonsense theory would be correct, L3 would have to expand not from East Africa, where all basal lineages have short stems (1 CR mutation) but from SE Asia, where N seems to have coalesced. As I said that's a nonsense. Obviously N is the one that made the long journey (as L3n or pre-N). Another example can be found in U, which, following your conjecture would have coalesced in Egypt (U1), something nobody will agree with.

     
  19. terryt

    November 9, 2010 at 4:12 am

    "if your nonsense theory would be correct, L3 would have to expand not from East Africa, where all basal lineages have short stems (1 CR mutation) but from SE Asia, where N seems to have coalesced". I don't know how you reach that conclusion. I was trying to point out that each of those haplogroups with long tails are actually no different from the ones with short tails. Haplogroups of both type have individually coalesced somewhere within their parent haplogroup's geographic range. The parent haplogroup had previously expanded and broken into regional varieties, exactly the same as have genera, species and subspecies. No haplogroup has moved far from where they coalesced from their parent haplogroup, before they have in turn expanded, no matter how long the tail. So look at L3's parent haplogroup, L3'4. That's what tells us most about L3's origin. The length of tail is not relevant. And technically L3 in its wider sense did make it all the way to SE Asia. "Another example can be found in U, which, following your conjecture would have coalesced in Egypt (U1), something nobody will agree with". And for a very good reason. It's immediately ancestral haplogroup is R, which coalesced in SE Asia. But R expanded as far as Egypt, or somewhere near it. Perhaps just R's descendant U itself actually reached Egypt, amoung other places. But back to the main point. It is very difficult, if not impossible, to place the origin of either N9/Y or A in SE Asia. And it makes sense that the development of an ice age would eliminate many haplogroups across Central Asia, dividing mtDNA N into eastern and western haplogroups. However it is again very difficult to explain the disappearance of N haplogroups through India.

     
  20. Maju

    November 9, 2010 at 6:14 am

    "I don't know how you reach that conclusion".Hopefully I misunderstood you. But I understand that you are all the time saying that certain subclades of N, your capricious choice, must have been located all the time, prior to their consolidation where they are found at expansion. These clades you choose are almost always the ones with longer stems (= longer obscurity "private clade" periods). So if you say that A means that N expanded from North Asia, then N means that L3 expanded from SE Asia (or maybe North Asia if we follow the thread). Instead the opposite is true: the location of the expansion of N and, to lesser extent, M are less informative regarding L3 as a whole, than those clades with short stems, which are all the African ones."I was trying to point out that each of those haplogroups with long tails are actually no different from the ones with short tails".They are different: the ones with short tails expanded soon after their parent haplogroups, while the ones with long stems did a lot later, time in which they may have wandered (or not). "Haplogroups of both type have individually coalesced somewhere within their parent haplogroup's geographic range".No. This is not a valid reasoning. Technically you are maybe correct because one mutation before N it is not yet N but something else (L3*, L3n, pre-N) but this does not apply if we are talking of the root haplogroup in its pure form (L3-root). Between L3 and N there are a series of private (or otherwise extinct now) clades which are distinct evolutionary steps:L3-rootpre-N (1)pre-N (2)pre-N (3)pre-N (4)N (N-root)In this sequence we know (I understand) the locations of L3-root and N-root but there are four intermediate steps which are totally unknown. Let's see:pre-N (1) is parallel to all the African L3 basal clades (five in number)pre-N (2) has no parallel other than two steps-derived L3 subclades, of which I can only find L3i. So maybe at that stage pre-N was in Arabia (or not, but in this case it makes good sense). pre-N (3) is parallel to M, so maybe then pre-N was in South Asia. pre-N (4) is the most difficult to compare, so I won't. N-root finally was in SE Asia, we have deduced. N did not coalesce in the L3 range… unless you consider the obscure pre-N (4) or the better know one-step derived M subclades to be that. N in fact coalesced half a world away from where L3 existed when it took shape as such distinct haplogroup.

     
  21. Maju

    November 9, 2010 at 6:29 am

    "No haplogroup has moved far from where they coalesced from their parent haplogroup, before they have in turn expanded, no matter how long the tail"The stem matters because the parent lineage of N is not in this sense you say L3-root but pre-N(4). However we do not call such kind of long lost lineages 'haplogroups' but this is a mere convention, as the word 'haplogroup' means group of extant (present day) haploid lineages. If pre-N(4) existed in, say Bangla Desh or maybe even already SE Asia, then what you say is correct… in relation to pre-N(4). But certainly N coalesced a long distance away from where L3 existed and there is absolutely no problem with that precisely because its stem is five CR mutations long: enough. Somewhat more problematic is when stems are short and yet distances are long, as happens with, specially, mtDNA M (see again this thread's map). You have sub-haplogroups, just one mutation away, in such distant places as South Asia, Japan and New Guinea. So what do we have to conclude? That M-root spanned all that range? Sure, technically yes or something quite similar. But that does not apply for the moment in which M began expanding for sure, which must have been from some more specific location and a more specific population. What we have to accept is that there are many sub-mutation steps, which cannot be analyzed, but are the real people who carried mtDNA M-root around, eventually (soon) coalescing in all those subclades. For this reason one-step mutations are most informative of the range reached by a haplogroup before deriving into recognizable sub-haplogroups. And for this reason they are more important.

     
  22. Maju

    November 9, 2010 at 7:01 am

    "So look at L3's parent haplogroup, L3'4. That's what tells us most about L3's origin".Nope. It's L3 and L4 which inform about L3'4. We don't know shit about L3'4 without looking at its daughter clades. "And technically L3 in its wider sense did make it all the way to SE Asia".But only very technically and very much in its wider sense. IMO you are word-playing here. If we consider single-step lineages (which must have existed in the past), L3-root never made all the way to SE Asia, only its descendants N (maybe pre-N(4) already) and M (in the form of would-be or maybe even already formed daughter lineages) did. But never anyone with the pure L3 sequence arrived to SE Asia: if they left Africa at all, they only reached as far as Arabia most probably."It is very difficult, if not impossible, to place the origin of either N9/Y or A in SE Asia".I'm not sure right now about where they coalesced but in any case what I am saying is that the wandered from N's origin to wherever they did. Obviously the first N person was one single person who surely lived in one single place. Then, she or one or several of her descendants (not yet mutated into something else) was quite successful in expanding. What I'm saying is that, in the case of N, these people lived essentially in SE Asia and there's where they expansion began. For me A is not different than HV, as both are five-steps removed from N. Just that HV's ancestors left a more clear trail (R, R0), while A's ancestors slided silently to their similarly remote destination. "And it makes sense that the development of an ice age would eliminate many haplogroups across Central Asia"…Excuses, nothing but pathetic pretexts! Why would that happen when some haplogroups clearly expanded to that same direction? "However it is again very difficult to explain the disappearance of N haplogroups through India".They just never left much of a mark, except for the most dynamic lineage, probably the largest of that N subpopulation. But all other "Western N" clades did leave some noticeable mark, with only one exception: X.The more we discuss the more convinced I am all these Western N represent a single back-migrant population, which could not expand much in South Asia (as it was already densely colonized by the M clan).

     
  23. terryt

    November 10, 2010 at 3:52 am

    "We don't know shit about L3'4 without looking at its daughter clades". Surely it's the other way round. Its daughter clades surely formed and then expanded in turn from where their parent haplogroup L3'4 got to. "They just never left much of a mark, except for the most dynamic lineage, probably the largest of that N subpopulation". That 'apparent' explanation is totally inadequate. R left plenty of descendants as it moved through India, yet its parent haplogropup N did not. Strange. "The more we discuss the more convinced I am all these Western N represent a single back-migrant population, which could not expand much in South Asia (as it was already densely colonized by the M clan)" But R managed to colonise India very effectively. "Just that HV's ancestors left a more clear trail (R, R0), while A's ancestors slided silently to their similarly remote destination" Leaving no trace along the way. Strange. "L3-root never made all the way to SE Asia, only its descendants N (maybe pre-N(4) already)" Presumably N*. Where it formed haplogroups N21, N22, N13, N14, S and O. And presumably R, possibly somewhere between N21/N22 and N13/N14/S/O. "I'm not sure right now about where they coalesced but in any case what I am saying is that the wandered from N's origin to wherever they did". Surely it was N* that wandered at first. Its daughter haplogroups wandered once they had formed in the separate regions N had expanded through beforehand. "What I'm saying is that, in the case of N, these people lived essentially in SE Asia and there's where they expansion began". I suppose it's possible to manipulate the evidence sufficiently to claim that haplogroups A and N9/Y may have originated in SE Asia. But you have an even harder job placing X, N1'5 N2/W's origin in SE Asia. I realise you dismiss Wikipedia unless it agrees with you, but this is what it says about various N haplogroups: Haplogroup A: "Its subgroup A1 is found in northern and central Asia, while its subgroup A2 is found in Siberia and is also one of five mtDNA haplogroups found in the indigenous peoples of the Americas … 7.5% of the Japanese belong to haplogroup A (mostly A4 and A5).[3] 2% of Turkish people belong to haplogroup A". Nothing about SE Asia (although derived haplogroups are found there) and certainly nothing about India. The three basal haplogroups of N9: "Haplogroup N9 – found in Far East.[18] Haplogroup N9a – East Asia, Southeast Asia and Central Asia. Haplogroup N9b – found in Japan. Haplogroup Y[23] – found especially among Nivkhs and Ainus, with a moderate frequency among Koreans, Mongols, Tungusic peoples, Koryaks, Itelmens, Chinese, Japanese, Tajiks, Island Southeast Asians (including Taiwanese aborigines), and some Turkic peoples[17]" Just one branch of one subhaplogroup in SE Asia (N9a). So It's difficult to make the case for an SE Asian origin for these two haplogroups.

     
  24. terryt

    November 10, 2010 at 3:56 am

    "But certainly N coalesced a long distance away from where L3 existed" Aren't you forgetting Haplogroup X? Not to mention N1'5 and N2/W. Why do you feel it's necessary to postulate huge distances?

     
  25. terryt

    November 10, 2010 at 4:02 am

    Sorry. I missed another one. "L3-rootpre-N (1)pre-N (2)pre-N (3)pre-N (4)N (N-root)In this sequence we know (I understand) the locations of L3-root and N-root but there are four intermediate steps which are totally unknown". We know L3 is African, but we don't know the location of N's root. Why are you absolutely certain it did not coalesce just outside Africa? "So maybe at that stage pre-N was in Arabia (or not, but in this case it makes good sense)". And there's no evidence that it did not remain there for the whole series of mutations. Then it expanded. "N-root finally was in SE Asia, we have deduced". No we haven't. You have decided. With no evidence. In fact evidence to the contrary.

     
  26. Maju

    November 10, 2010 at 7:22 am

    Terry: we know nothing of L3'4 without looking at its daughter clades. Nobody is L3'4(xL3,L4) anymore and we do not have a time machine to go back in time and make genetic tests to the Idaltu clan or whatever. All we know of parent haplogroups is because of descendant ones living today (or in some cases of aDNA research as well). Reconstructing where L3'4 existed depends on first reconstructing where L3 and L4 did exist and is independent of where horizontal relatives such as L6, L5, L2 and L1 exist. The process is always from today's actual haplotypes to yesterday's inferred haplogroups, what ca be a pain to calculate in large scale but can be done as I demonstrated in the African reconstruction maps (notwithstanding possible errors in the process).It is obvious that you have never looked at haplogroups this way so, I'd suggest you to start doing some reconstructions yourself. Reconstructing N's original location from modern data should not be that hard, so go ahead and do it. Yah, take a map of Asia-Australasia and begin the task of joining the dots. "R left plenty of descendants as it moved through India, yet its parent haplogropup N did not. Strange".Exactly what happened in West Eurasia, where R left plenty of descendants (what's that? 80-90%) and N only a bunch of lesser haplogroups, such as N1 (the most important of all, and also the first to branch out, after R, in the Western group), X and W. You are aggrandizing the importance of residual N. Most N is R through the World, excepted some random peripheral areas like Aboriginal Australia. So again thanks for revealing to me (albeit if confusingly) the SE Asian origin of R, it saves ancient humankind a whole journey back there and helps to explain N much better.

     
  27. Maju

    November 10, 2010 at 7:42 am

    "Presumably N*"N* means N-other in present day. There are always some unclassified lineages (or just not tested for in a particular research) that are placed under that umpbrella. N alone or, more specifically, N-root or something like that should be used instead when talking of past pre-evolved "pure" N. "Surely it was N* that wandered at first. Its daughter haplogroups"…N* has no daughter haplogroups. It's residual N-other of today (a paragroup). I insist. Itself is descendant from N-root just like all other haplogroups in that clade, including all within R. "I suppose it's possible to manipulate the evidence sufficiently to claim that haplogroups A and N9/Y may have originated in SE Asia".I am not even saying that. I am talking of N-root before all these clades arose. It is the same in regard to the four Australian N subclades and the two known Negrito ones, etc."I realise you dismiss Wikipedia unless it agrees with you"…I do not. As former active wikipedian, I know well that some stuff is accurate and other is not so much, so I appreciate when it's well documented and can be double-checked. Double-checking Wikipedia is generally a good idea, specially if you're working with specialist data, as we are doing here. "Just one branch of one subhaplogroup in SE Asia (N9a)"…And Y2, what makes it 1.5 (out of 3). Also I am quite sure that when I decided in the past to place N9's origin in SE Asia instead of further to the North it was because N9b is also found there. Anyhow, it won't make a big difference re. the origin of N as a whole, because Negrito and Australian haplogroups weight a lot (6/12), so you'd need to place all other 6 in Lappland or something like that to argue for an Altai origin. The real problem for your hypothesis is that there are not enough basal clades of N in Siberia or near it. So unless you want to argue that all SE Asian and Oceanian basal N subclades are ill researched and in fact (conjecturally) are a single haplogroup, you cannot argument your model at all. That's the only chance your model has: that somehow it's demonstrated in the future that SE Asian/Oceanian clades are tied by a single umbilical cord, now missing. Only that would displace the origin of N sufficiently to the North and West as to support your conjecture.

     
  28. terryt

    November 10, 2010 at 11:33 pm

    "that somehow it's demonstrated in the future that SE Asian/Oceanian clades are tied by a single umbilical cord, now missing". They are. They're all N haplogroups. "Only that would displace the origin of N sufficiently to the North and West as to support your conjecture". I've never claimed N 'originated' in the north. Only that it passed that way on its journey to East Asia. It's very difficult to avoid the conclusion that it originated just outside Africa when you look at the evidence. "Negrito and Australian haplogroups weight a lot (6/12)" You cannot use that sort of 'logic' to determine the region of origin of any haplogroup. There are any number of reasons why diversity might survive in one region and become reduced in another. "N alone or, more specifically, N-root or something like that should be used instead when talking of past pre-evolved 'pure' N". OK. I'm talking of the last of those N-root haplogroups when I say, 'And there's no evidence that it did not remain there [in Arabia] for the whole series of mutations. Then it expanded'. "we know nothing of L3'4 without looking at its daughter clades. Nobody is L3'4(xL3,L4) anymore" OK. The region of origin for L3 is likely to be somewhere near where the haplogroups most closely related to it are found. In this particular case L4. Same with haplogroups M and N. "It is obvious that you have never looked at haplogroups this way" Maju. I have always looked at haplogroups this way. "N only a bunch of lesser haplogroups, such as N1 (the most important of all, and also the first to branch out, after R, in the Western group), X and W". Haplogroups that had been in the west in some form from the moment the pre-N (L3N?) haplogroup left Africa. There is no evidence at all that they made the journey from to SE Asia. "So again thanks for revealing to me (albeit if confusingly) the SE Asian origin of R, it saves ancient humankind a whole journey back there and helps to explain N much better". We obviously still have a little way to go before you will be able to see the SW Asian origin of M and N though.

     
  29. Maju

    November 11, 2010 at 3:48 am

    "You cannot use that sort of 'logic' to determine the region of origin of any haplogroup".That's how it is. "There are any number of reasons why diversity might survive in one region and become reduced in another". I do not think so, only extinction does that. And in any case the now extinct hypothetical Arctic sublineages of N are not any better sort of evidence than the now extinct such lineages in Madagascar, Ceylon or the Moon. Dead people do testify, hypothetical vanished lineages do not weight. First go and find them.Because they have even tested for aDNA… but it is not N, M, L3 nor anything "ours": it is H. erectus mtDNA what is there (Denisova).So there is even a negative evidence against your rant. Now please shut up and think. "OK. I'm talking of the last of those N-root haplogroups when I say, 'And there's no evidence that it did not remain there [in Arabia] for the whole series of mutations. Then it expanded'".Not sure what you're talking about here because your terminology is sloppy. If for you anything between the L3 and the A nodes is "N", but R is not, for instance, then we have a problem of understanding. "The region of origin for L3 is likely to be somewhere near where the haplogroups most closely related to it are found".Not necessarily. And again N is a great example: as its origin is half a planet away from where L3, L4, and all Ls are found, excepting itself and M. "Same with haplogroups M and N".M and N are not more related to each other than to L3x for instance. "Maju. I have always looked at haplogroups this way".Then why you are saying the opposite? Why are you saying that the origin of haplogroups must be located based not on their descendants but their ancestors and sibling clades?"Haplogroups that had been in the west in some form from the moment the pre-N (L3N?) haplogroup left Africa. There is no evidence at all that they made the journey from to SE Asia".There is one BIG evidence tatooed on the DNA: ALL them have exactly ALL mutations defining N at the root. If pre-N (L3n) branched out as you suggest, Western clades would be pre-N not fully evolved N.So they were together (in a single personal line) for 5 full "eras" (single-mutation time periods) and only then in a single "era", they scattered around fast (but thinly, R excepted). But there were still another five "eras" before A showed up, four for X, etc."We obviously still have a little way to go before you will be able to see the SW Asian origin of M and N though".Little is not the word here: forever is the word. It is you who are absolutely wrong in this.

     
  30. terryt

    November 12, 2010 at 3:06 am

    "Little is not the word here: forever is the word". You argued exactly that whenever I tried to persuade you of mtDNA R's origin in SE Asia too. So we'll await further evidence. "And again N is a great example: as its origin is half a planet away from where L3, L4, and all Ls are found" There is no evidence for that statement at all. Just your belief. "M and N are not more related to each other than to L3x for instance". Exactly. So why would we expect them to have coalesced half a planet away from somewhere near where L3a, L3b'f, L3c'd'j, L3e'i'k'x and L3h coalesced? "Why are you saying that the origin of haplogroups must be located based not on their descendants but their ancestors and sibling clades?" Because their descendant could have moved far from where they originated. Any haplogroup's descendants provide very little evidence regarding its origin, although the haplogroup presumably usually coalesced in a region where its descendants are still found. "If pre-N (L3n) branched out as you suggest, Western clades would be pre-N not fully evolved N". Why so? Not if its expansion, once fully evolved, was rapid. Surely you should argue the same for mtDNA R. Its eastern branches should be less fully evolved if it originated in SE Asia. "So they were together (in a single personal line) for 5 full 'eras' (single-mutation time periods) and only then in a single 'era', they scattered around fast (but thinly, R excepted)". Is that really so unbelieveable? I find it hardly remarkable at all. And there's no evidence at all that mtDNA N accompanied R on its expansion. In fact the evidence seems against the connection. "But there were still another five 'eras' before A showed up, four for X, etc". Again. What's so remarkable about that? The individual haplogroups remained isolated in various regions that N had passed through during its rapid route to SE Asia. They then remained isolated in those regions for some time, giving each haplogroup a long tail, until they were eventually able to indulge in expansions of their own. "If for you anything between the L3 and the A nodes is 'N'" And I was talking about N, not A. Conveniently for your belief N has a tail of five mutations. This allows you to shift its origin anywhere around the world that you find convenient. So you have it moving across India to SE Asia, leaving no trace of its route. And then moving back through India, again leaving no trace of its route. Amazing. You use the same 'logic' for haplogroup A. You simply shift it around to fit your current belief. And you reckon I make things up!!!

     
  31. Maju

    November 12, 2010 at 6:45 am

    "You argued exactly that whenever I tried to persuade you of mtDNA R's origin in SE Asia too. So we'll await further evidence".I hate how you put this matter as some sort of crusade or jihad you are in. We need no preachers, thanks. What we need is curious flexible-minded people who can see all the viewpoints of any matter, so the best solution(s) are found. I may have some fun with preachers for a while but eventually I do get bored. So open your mind to what others say or get lost. "So why would we expect them to have coalesced half a planet away from somewhere near where L3a, L3b'f, L3c'd'j, L3e'i'k'x and L3h coalesced?"No "expectations" involved, sir, just factual data: it has nothing to do with ancestors but descendants and descendants say: South and SE Asia with impressive clarity. "Any haplogroup's descendants provide very little evidence regarding its origin"…Ok, it's clear we do not agree in method. We cannot therefore agree in anything else. I have been all the time working that way: from descendant to ascendant, from bottom to top. And there is no other way, because we have no data for the past ancestor nodes other than the descendants of today. And in the past you have followed and cheered my method, so what's the problem now? That it does not fit with your doctrine. Sorry but happens a lot in science. [re. N] "Why so? Not if its expansion, once fully evolved, was rapid".Oh, but, according to you, they had no boats, so they could not paddle their way. So now we have the "rapid migration through Tarim Basin" or what? Did they already have domestic camels? Or maybe they are like Santa with flying reindeer? "Is that really so unbelieveable?"It has very low likelihood. Occam rules against. …

     
  32. Maju

    November 12, 2010 at 6:45 am

    …"And there's no evidence at all that mtDNA N accompanied R on its expansion"They are found together in almost all places. So it becomes a strong possibility. "Again. What's so remarkable about that? The individual haplogroups remained isolated in various regions that N had passed through during its rapid route to SE Asia".But, in spite of being such a fantastic route by such marvelous lands for human habitation (whoa, Siberia, better than Bahamas!), they did not expand until much later. Please… it's a total rant! Do you even read what you write? Do you even bother criticizing your own conjectures? "Conveniently for your belief N has a tail of five mutations".Yes, so extremely conveniently that it cannot be just luck. "This allows you to shift its origin anywhere around the world that you find convenient".No. I follow the geography of descendants: the descendants of L3 for the previous node (L3) and the descendants of N for N itself. It cannot be anywhere in the world: it must be the two specific areas I am saying, I cannot arbitrarily move these: the descendants won't allow me. "So you have it moving across India to SE Asia, leaving no trace of its route. And then moving back through India, again leaving no trace of its route. Amazing".How is it more amazing than a conjectural Siberian route with even less traces, which does not even fit the geometry of the descendants? And anyhow they did leave traces in South Asia in the migration back to the West. Not many but enough. I'm now even pondering if W (most fequently found in North Pakistan) could be of South Asian origin, what would place all N2 in South Asia until the expansion of W itself. You always accuse me of "belief" and nonsense. This is absolutely tiresome and boring. But you never look at the facts yourself: the descendants define the ancestor node in all aspects, including geography. At least that's how it works for us, looking at the matter from present-day data. That's how it is: the descendants dictate where the ancestor was. I do nothing but listening to them. If you could listen (instead of actively searching for unlikely twists that might in your imagination support your preconceptions) you would see it crystal clear. I do not manipulate shit: the haplogroups just speak themselves.

     
  33. terryt

    November 13, 2010 at 9:04 am

    "I hate how you put this matter as some sort of crusade or jihad you are in". I'm trying to help people understand what actually happened, not what they prefer to believe happened. "And in the past you have followed and cheered my method, so what's the problem now?" Because you're not using the same method as you've used to trace European haplogroups, and which you are still doing a very good job of in your latest blogs. "No 'expectations' involved, sir, just factual data" What 'factual data'? "They are found together in almost all places. So it becomes a strong possibility". But not in India. "And anyhow they did leave traces in South Asia in the migration back to the West. Not many but enough". What traces? "I'm now even pondering if W (most fequently found in North Pakistan) could be of South Asian origin, what would place all N2 in South Asia until the expansion of W itself". No it wouldn't. N2 breaks into two haplogroups after four mutations: N2a and W. According to Wiki, 'Haplogroup N2a – small clade found in West Europe.[20] Haplogroup W[21] – found in Western Eurasia and South Asia[22]'. Nothing exclusively South Asian all. Just that you wish it to be so. "You always accuse me of 'belief' and nonsense". Well, well, well. And there's an example of your own 'belief and nonsense'. "But you never look at the facts yourself: the descendants define the ancestor node in all aspects, including geography". But just because one line of descendants is found in a particular region, say SE or South Asia, doesn't mean the haplogroup must have originated it that region. The region of early branches can give a very good idea but when you use downstream mutations to justify a particular belief your argument moves onto shakey ground. "How is it more amazing than a conjectural Siberian route with even less traces, which does not even fit the geometry of the descendants?" It fits exactly the 'geometry of the descendants' but you prefer to place the descendants elsewhere. "But, in spite of being such a fantastic route by such marvelous lands for human habitation (whoa, Siberia, better than Bahamas!), they did not expand until much later". Ever heard of 'climate change'? "So now we have the 'rapid migration through Tarim Basin' or what?" Presumably something like that.

     
  34. Maju

    November 13, 2010 at 9:53 am

    "I'm trying to help people understand what actually happened"…What you think, or believe, happened. Instead what I demand is to look at matters for what they are: to look at the data without any preconceptions. Can you do that?I think you cannot. "Because you're not using the same method as you've used to trace European haplogroups"…I think I am using the same method all the time. Otherwise explain me the differences, please. "What 'factual data'?"Where the haplogroups and the subhaplogroups, and the sub-sub-sub..-haplogroups (f)actually exist. "But not in India".In India too. "What traces?"R, N2 (incl. W) and N5 (half of N1'5). "According to Wiki, 'Haplogroup N2a – small clade found in West Europe.[20] Haplogroup W[21] – found in Western Eurasia and South Asia[22]".Metspalu 2006 (supplement) seems to indicate N2a being South Asian but maybe it's found in both regions, as happens with W and other haplogroups. Whatever the case, let's assume for a moment all three Western N(xR) haplogroups are not found in South Asia (which is not true but whatever), where is the evidence pointing to Central Asia/Siberia? Nowhere! So both hypothetical routes would be on the same situation. You still cannot prove that N migrated back to West Asia via the North or that it stayed all the time in the Far West, while the rest of N spread ultra-rapidly eastwards (without boats, to make things even more difficult). "But just because one line of descendants is found in a particular region, say SE or South Asia, doesn't mean the haplogroup must have originated it that region".Exactly my point: it is all the lineages together which describe where the ancestor was (by quite simple geometry). "It fits exactly the 'geometry of the descendants' but you prefer to place the descendants elsewhere".It does not. Have you tried to plot the dots on a map yourself and then find their centroid? It falls in SE Asia for N, there's no other solution and moving a few clades a bit here or there in Asia does not change that substantially because most of the weight is in SE Asia and Australia. The only way that could happen would be if several of the SEA/Australasia clades would be joined into a larger N subhaplogroup, what would reduced their overall weight. But right now we have 7/12 clades pulling strongly towards SE Asia. And the rest do not pull towards West or Central Siberia anyhow but towards NEA and WEA. There's no subarctic diversity so early: it formed only at later times, at more downstream levels. "Ever heard of 'climate change'?"No matter: Bahamas will always be better than Siberia… unless they become a desert. Today we have some of the warmest conditions ever in the history of Humankind and Siberia is still a barren desert essentially, where human life is possible but extremely hard. And it's fucking cold!

     
  35. terryt

    November 15, 2010 at 2:45 am

    "Instead what I demand is to look at matters for what they are: to look at the data without any preconceptions. Can you do that?" Yes I can. But can you? You have several times claimed that I fail to consider all the evidence when trying to explain the long tails of mtDNA haplogroups N, A and N9/Y. So let's look at some other haplogroups with long tails. Within haplogroup N we find three other basal haplogroups with long tails: N13 has 13 mutations, N14 has 10 and N21 has 8. To apply your theory consistently would necessitate us assuming that the two Australian haplogroups really took their time moving from SE Asia to Australia while their tail grew, similar to the situation you claim for the original N migration from Africa to SE Asia. But the other two Australian basal N haplogroups, S and O, have respectively just 1 and 3 mutations in their tails. So, according to your theory, N13 and N14 must have been much later arrivals in Oz than were O and S. And the Malay haplogroup N21 has an 8 mutation tail. So, applying your theory consistently, we have to assume it must have moved very slowly from SE Asia to SE Asia! When we turn to mtDNA R we find very long tails in several basal haplogroups. In the east we have R22 (in Indonesia, especally in the Lesser Sunda Islands) with a 10 mutation tail, and R23 (in Bali and Sumba) with an 18 mutation tail. These two haplogroups must also have taken a very long time to move from SE Asia to SE Asia! And what about R14 from New Guinea and the Nicobar Islands? A massive 23 mutation tail. That's enough to move to Africa and back again, possibly several times. And in the west we have two more R haplogroups with long tails: R1 with 17 mutations and R3 with 19. According to your theory these haplogroups must have left SE Asia long after the other R haplogroups departed, and they moved extremely slowly west leaving no descendants as their tail developed. In all the above cases isn't it much more likely that their long tail is a result of drift acting for some time on each separate haplogroup once it had become isolated somewhere within their ancestral haplogroup's geographic distribution? They each later embarked on their own individual expansion. So isn't it extremely likely that we have the same general explanation for the tails of mtDNA haplogroups N, A and N9/Y, and even for the relatively short M tail? M and N each coalesced as a result of drift acting for some time on each separate haplogroup once it had become isolated somewhere within their ancestral haplogroup's geographic distribution, in this case L3. And A and N9/Y each coalesced as a result of drift acting for some time on each separate haplogroup once it had become isolated somewhere within their ancestral haplogroup's geographic distribution. No need to postulate long migrations that left no descendants along the way while the tail developed.

     
  36. Maju

    November 15, 2010 at 7:02 am

    Read my lips: my "theory" does not say lineages provided of a long stem traveled necessarily… it just enables them to do that more calmly than if they have a single-transition stem. I'm not saying that stem means travel… I'm just saying it does not mean staying put in a shell as you claim. Is this clear? Please respond YES/NO. If no explain your doubts.

     
  37. Maju

    November 15, 2010 at 7:14 am

    "So, according to your theory, N13 and N14 must have been much later arrivals in Oz"…They might but they do not need to be so. The only thing long stems says is that the lineage survived in near-extinction conditions for long (for example as minority clade in a larger population or in an isolated clan in an island or even maybe that the lineage was violently pruned by a catastrophe of natural or human origin). "In all the above cases isn't it much more likely that their long tail is a result of drift acting for some time on each separate haplogroup once it had become isolated somewhere within their ancestral haplogroup's geographic distribution?"You want to keep each lineage in a box and that surely not possible. Drift of course was in action but did not need to act on a single isolated monophyletic "castes" (from Port. "casta": lineage, breed, race) but more probably in larger and more variegated real populations and population networks. There are just not enough isolated islands and deep gorge valleys to keep each lineage separated, specially when we know people moved around a lot in th early period of the Eurasian colonization.

     
  38. terryt

    November 16, 2010 at 2:55 am

    "Is this clear? Please respond YES/NO. If no explain your doubts". No. It's not clear. You are inconsistent as to when you claim that 'stem means travel'. But you are consistently prepared to claim 'stem means travel' when it fits your belief, and, on the other hand, claim it doesn't mean travel when that perspective fits your belief. So how do you decide when to claim 'stem means travel'? "Read my lips: my 'theory' does not say lineages provided of a long stem traveled necessarily…" You're arguing with a fair degree of certainty that it does so in the case of mtDNA N, and pretty much so in the case of A. Why the inconsistency? "for example as minority clade in a larger population" You've already claimed elsewhere that a 'minority clade in a larger population' is unlikely to survive very long. And I agree. "in an isolated clan in an island" Not necessarily just on an island. All sorts of things can lead to population isolation. "Drift of course was in action but did not need to act on a single isolated monophyletic 'castes' (from Port. 'casta': lineage, breed, race) but more probably in larger and more variegated real populations and population networks". Quite. But if drift acts for any length of time on any 'larger and more variegated real populations and population network' it will very quickly eliminate all haplogroups but one. "There are just not enough isolated islands and deep gorge valleys to keep each lineage separated, specially when we know people moved around a lot in th early period of the Eurasian colonization" And isolation is so much more likely when the overall population is relatively small. Presumably that's why basal N (and Y-hap C) clades are so widely scattered.

     
  39. terryt

    November 16, 2010 at 3:04 am

    Oh. By the way. Have you seem Wiki's entry on Y-hap C lately? http://en.wikipedia.org/wiki/Haplogroup_C_(Y-DNA)Have a look at the map. It shows the haplogroup splitting into two. C5, C3 and C1 moving through Central Asia and C4 and C2 moving through India to SE Asia and Australia. I strongly suspect that there were no 'two routes' east though. We need another explanation. But going back to a comment you made the other day: "Bahamas will always be better than Siberia… unless they become a desert. Today we have some of the warmest conditions ever in the history of Humankind and Siberia is still a barren desert essentially, where human life is possible but extremely hard. And it's fucking cold!" Bahamas is batter, but if it's already full, or you can't get there anyway, you have to make do with Siberia.

     
  40. Maju

    November 16, 2010 at 11:53 am

    "You are inconsistent as to when you claim that 'stem means travel'".I do not claim that. Long stems only mean long coalescence times (as small "invisible" lineages), just that. They could have coalesced locally or they could have traveled or anything in between. What means travel is when a haplogroup is found to have expanded far away from what is the logical reconstructed core of its parent haplogroup. Is this clear?"You're arguing with a fair degree of certainty that it does so in the case of mtDNA N, and pretty much so in the case of A. Why the inconsistency?"No. I am arguing that N expanded far away from the logical origin of its parent lineage L3. I believe you agree with me in this. While a long stem helps to explain this (long time of migration) it is not a condition, as the migration can always be faster. You don't really need 5000 or 50,000 years to walk or sail from Arabia to Burma. Not counting modern obstacles such as borders, you can well do that in few years or even months. But people was surely not in such a hurry in the Paleolithic and only migrated when they felt the need to do it (the details of why we can only speculate about). "You've already claimed elsewhere that a 'minority clade in a larger population' is unlikely to survive very long".Depends. If the population is large and growing the odds increase a lot. Drift acts only slowly (faster in smaller and decreasing or stagnated populations). It also depends on how much minoritary it is the haplogroup: it's not the same to be 1/1000 than to be 1/20, right?Anyhow unlikely is not the same as impossible. Maybe hundreds or even thousands of other small lineages went extinct… but a few survived. Probability works that way: it's not mechanically straightforward, you know."All sorts of things can lead to population isolation".Not sure if this is correct but certainly is difficult to explain why a clade would not expand for a long time and then suddenly do it massively, except if it happened to have arrived to a new "empty" land where to thrive. This applies specially to above average star-like expansions, which are all concentrated in "Greater Eurasia" (i.e. not in Africa) and must therefore, almost necessarily, reflect demographic expansions into virgin or quasi-virgin lands….

     
  41. Maju

    November 16, 2010 at 11:53 am

    …"But if drift acts for any length of time on any 'larger and more variegated real populations and population network' it will very quickly eliminate all haplogroups but one".But that assumes no population interactions, no expansions, no substructure… and even no stochastic randomness (or even real individual lives)… it is a good but highly theoretical lab model rather than a realistic one. There's no 1:1 map, reality always beats fiction (theory), etc.Also I really disagree with your use of the phrase "very quickly". At the limit, when time tends to infinite, it unavoidably happens but time is always finite, and that's why the equation does not work so well in reality, even if it does reflect a real tendency. If time=2500 generations (roughly 50,000 years), then it is still not infinite. And that is the largest figure I could came up with. "Very quickly" may mean what? Four generations? 50 generations? Sure, the likelihood of "drift out" increases each generation that passes without expansion but it is never absolute. You can probably estimate how many other minor lineages had to go extinct, given X and Y circumstances, for one to survive, statistically speaking, but that's about it. You cannot ever confirm that any one lineage had necessarily to go extinct, much less when we know so little about the exact details. The odds may be against survival, the same they are against me winning lottery (assuming I'd buy a ticket, what I do not). But I can still win the big prize: there is a slim chance I do (and that's why people play lottery). "And isolation is so much more likely when the overall population is relatively small. Presumably that's why basal N (and Y-hap C) clades are so widely scattered".Population should have grown very fast in the conditions of a "virgin" Asia, so I think there must have been other factors at play, like multi-lineage expansions or second waves. "Have a look at the map. It shows the haplogroup splitting into two. C5, C3 and C1 moving through Central Asia and C4 and C2 moving through India to SE Asia and Australia".They must have changed the map in the last hours because now it shows the classical "because of C, then coastal migration" model. Central Asia is empty. Checked the history and nope: the page has not been edited since September. You are suffering from some delusion. "Bahamas is batter, but if it's already full, or you can't get there anyway, you have to make do with Siberia".Fair enough (in the metaphoric realm). In fact Siberia was, in my understanding only colonized by AMHs after Tropical Asia was full. c. 50 Ka ago or later.

     
  42. terryt

    November 18, 2010 at 3:17 am

    "Not sure if this is correct but certainly is difficult to explain why a clade would not expand for a long time and then suddenly do it massively" The development, or introduction, of new technology springs to my mind immediately. "Long stems only mean long coalescence times (as small 'invisible' lineages), just that. They could have coalesced locally or they could have traveled or anything in between". But they're far more likely to have coalesced locally. "What means travel is when a haplogroup is found to have expanded far away from what is the logical reconstructed core of its parent haplogroup. Is this clear?" Yes. Very clear. And, unlike you, I have nowhere claimed any haplogroup has 'expanded far away from what is the logical reconstructed core of its parent haplogroup'. Quote: "I am arguing that N expanded far away from the logical origin of its parent lineage L3". Why? It's not necessary to do postulate that N moved a huge distance from where it first emerged from Africa before it,in turn, expanded further. "I believe you agree with me in this". (continued)

     
  43. terryt

    November 18, 2010 at 3:48 am

    "I believe you agree with me in this". Try, for just a moment, to consider the possibility that mtDNA haplogroups M and N did actually first coalesce in separate regions somewhere near Africa. And try to look at the evidence with an open mind to see if it might be possible to untangle the data sufficiently to support such a possibility. The tails show that the two haplogroups, pre-N and pre-M, were both subject to a considerable period of drift through either restricted population size or bottleneck before they were able to diversify, although M's population diversified rapidly once it had managed to burst into India after 3 mutations. It's quite possible that M1 remained for some time in the region where M had originally coalesed though. Members of M1 are found in Africa, but almost certainly a result of back migration. So that's M out of the way. After 5 mutations N too broke free from the region where it had coalesced (Middle East or, very unlikely, SE Asia) and began to diversify, rapidly. However it's possible that the members of N who remained behind were not able to escape the established pressure from drift that had beset the original pre-N population. No SE Asia haplogroups demonstrates such a situation but X had 7 more mutations before it too began to diversify. N2/W experienced 5 mutations before it began its own population expansion, especially in Pakistan according to the information you supplied a few days ago. But by that time N5 had almost certainly already reached India because N1'5 split after just 1 mutation, and N5 is Indian. But neither N5 nor W moved very far through South Asia towards the east. Haplogroup A underwent an 8 mutation long period of restricted population size after it had separated from the other N haplogroups. We'll leave open for now the question of where that population survived over that period. But as soon as members of haplogroup N had reached the relatively benign habitat around the Yellow Sea the population expanded rapidly. After just 1 mutation haplogroup N9/Y split into N9a, N9b and Y. Interestingly, in his diagram of haplogroup N, Ian Logan still has haplogroups P and R basically descending from N9/Y. Further south R began its own very rapid expansion after just 1 mutation. As did S in Australia. Haplogroup O expanded in Australia after 3 mutations. The other two Australian N haplogroups, N13 and N14, survived a long period of restricted population size (13 and 10 mutations respectively) and have not undergone any real population expansion. Nor have SE Asian haplogroups N21 and N22 with 8 and 7 mutations. So that's it. All the mtDNA N haplogroups fitted snugly into a coherent sequence of expansion. No complicated two-way migrations leaving no descendants along the way. Nothing complicated at all in fact. "In fact Siberia was, in my understanding only colonized by AMHs after Tropical Asia was full. c. 50 Ka ago or later". But we don't know that for certain. And it was certainly colonized by some sort of humans long before 50 Ka.

     
  44. Maju

    November 18, 2010 at 7:53 am

    "Try, for just a moment, to consider the possibility that mtDNA haplogroups M and N did actually first coalesce in separate regions somewhere near Africa".It does not work that way because then it would not have been M and N but some derived downstream clades the ones expanding. So pre-M and pre-N possibly did what you say but they did not quite expand yet. "although M's population diversified rapidly once it had managed to burst into India after 3 mutations. It's quite possible that M1 remained for some time in the region where M had originally coalesed though".Pre-M1 you mean? Possible but rather unlikely, it'd be easier for a pre-M clade to do that. Something like L3i, which is there to attest such kind of possibility, it seems, but is not the same line (a sister one). "Ian Logan still has haplogroups P and R basically descending from N9/Y"Logan's graphs are admittedly obsolete. Old stuff. Look at PhyloTree. I'm done with this discussion, I think. You are not the least convincing to me. I do not think haplogroups the way you do: if two or more separate expansions happened in the line leading to N as you say we'd have Western pre-N and NEA pre-N as well, not N. That would be the evidence in favor of your hypothesis, which lacking that key evidence is nothing but wild speculation.

     
  45. terryt

    November 19, 2010 at 3:53 am

    "I'm done with this discussion" So am I. Except for one thing: "if two or more separate expansions happened in the line leading to N as you say we'd have Western pre-N and NEA pre-N as well, not N". I have never claimed any more than one expansion of N. Just one expansion. It's your theory that has N moving backwards and forwards across half the world leaving no descendants anywhere along the way.

     
  46. Maju

    November 19, 2010 at 10:05 am

    "It's your theory that has N moving backwards and forwards across half the world leaving no descendants anywhere along the way".I think it's you who says that: you have N spawning subclades in West Asia, in Siberia and then again in SE Asia and possibly Australia too with your idea of large range of proto-haplogroups.This is not possible. Even if the pre-N clade would have been hidden in some corner and then suddenly, with no preliminary track, you still claim that N-root spread all around Asia and Australia, yet without evolving to anything new (such as R or whatever). It is not just unbelievable but totally breaks the concept that we can track lineages at all, and that they were carried by normal people like you and me. Because if, even before expansion, a haplogroup could be everywhere at the same time… we are before something more proper of quantum mechanics or relativistic physics, not about normal people living normal lives.

     
  47. terryt

    November 19, 2010 at 8:24 pm

    "I think it's you who says that" No it's not. Let me clarify what you believe happened. You claim that pre-N and pre-M emerged from Africa into the virgin habitat of SW Asia. They then traveled to the virgin habitat of India without leaving any descendants along the way. What's more pre-N then moved through India into SE Asia, again without leaving any descendants along the way. At which stage N diversified instantly into 12 different haplogroups, which expanded in various directions. Some of these haplogroups moved back through already occupied India, along with the N daughter haplogroup R. In that region R managed to leave many descendant haplogroups along the route but, yet again, N failed to leave any descendants. This whole scenario is surely very unlikely. "you have N spawning subclades in West Asia, in Siberia and then again in SE Asia and possibly Australia too with your idea of large range of proto-haplogroups". In contrast to your scenario I have N spawning various subclades as it moved (presumably rapidly) from Africa to Australia. Surely a much more likely scenario than yours. "claim that N-root spread all around Asia and Australia, yet without evolving to anything new" Again it is you who claim that situation. You have N moving huge distances around the world, often through virgin habitat, yet failing to leave any descendants behind. I totally accept the N-root left various haplogroups along its route, that eventually (or in some cases rapidly) gave rise to the surviving N-derived haplogroups. So this last comment of yours is completely stupid: "if, even before expansion, a haplogroup could be everywhere at the same time… we are before something more proper of quantum mechanics or relativistic physics, not about normal people living normal lives".

     
  48. Maju

    November 19, 2010 at 10:15 pm

    I understand your reasoning but for me the scenario is most likely. In Arabia it was several lineages (L3i, L6, the various L0s…) because several sub-regions were at play (Yemen, Persian Gulf…) and Africa was just across the Red Sea (which for me is no real obstacle) bringing new people now and then, keeping the diversity somewhat high that way. In South Asia insteadthe initial population was soon M-dominated, maybe with some pre-N and some other minor lineages but these eventually faded with drift. It'd be interesting that both of us looked in detail at the M subgroups because the expansion of N corresponds, IMO, not to the M stage but to two steps downstream (later in time). So the question is not why N left no trace in SA but why it did expand so dynamically in SEA, when there were surely earlier arrivals within the dominant M clan which do not show such dynamism. I have no answer to this question but it's clear that N and R right after it, showed unusual dynamism, not just in SEA but at continental scale. So we have a peculiar, highly mobile and expansive "tribe" dominated by the N haplogroup. Why was this? Without enough archaeological knowledge available I cannot tell. "In contrast to your scenario I have N spawning various subclades as it moved (presumably rapidly) from Africa to Australia. Surely a much more likely scenario than yours".Doesn't make sense because then the core expansion of N would be in West Asia, where the diversity is low (3/12) and then in NEA (as you defend the northern route), where diversity is even lower (2/12). Logically the center must be where diversity is high: in the SE (7/12 basal lineages, once we include R) and where the geometry of the whole haplogroup is centered (all agrees). You put too much emphasis in L3 but ignore the fact that N does not show any sign of expansion for four full mutations/periods (at least 12,000 years, probably 16-20 millennia). And then suddenly it expands everywhere at the same time (according to you). "You have N moving huge distances around the world, often through virgin habitat, yet failing to leave any descendants behind".The people left descendants, no doubt but the lineage did not, as it was not dominant and was drifted out. Other co-migrant lineages did instead (M specially but several L clades too in Arabia). What I think instead is of populations which had several lineages not always just one. Even in Papua, it's difficult to find such extreme homogeneity when we are talking of something bigger than a village: populations hold some diversity. They do tend to homogenization but this tendency is seldom fulfilled to its extreme. Even Basques have two Y-DNA lineages (excluding erratics), even Native Americans have a bunch of them. Total homogenization in a population is rare.

     
  49. terryt

    November 21, 2010 at 1:10 am

    "In South Asia instead the initial population was soon M-dominated, maybe with some pre-N and some other minor lineages but these eventually faded with drift". Seems difficult to explain why one basal haplogroup would expand hugely while the other dissappeared entirely. As you pointed out: "Even in Papua, it's difficult to find such extreme homogeneity when we are talking of something bigger than a village: populations hold some diversity". And that diversity in Papua includes haplogroups of different origin, not just one. "The people left descendants, no doubt but the lineage did not, as it was not dominant and was drifted out". It which case it's absolutely stunning that N made it to SE Asia in the first place. "Total homogenization in a population is rare". Especially today. But I strongly suspect it was not too uncommon in the Paleolithic, although there has probably always been some wandering to and fro. But in the Paleolithic as any particular population moved into a new region, or developed a new way to exploit existing resources, their numbers would have expanded to the maximum sustainable. During the following period of relatively stable population numbers we get a bottleneck, in that haplogroups start to dissappear. "Logically the center must be where diversity is high: in the SE (7/12 basal lineages, once we include R)" But that diversity may be the product of major population expansion in the region as opposed to restricted population expansion, or bottlenecks, further north. "So the question is not why N left no trace in SA but why it did expand so dynamically in SEA, when there were surely earlier arrivals within the dominant M clan which do not show such dynamism. I have no answer to this question but it's clear that N and R right after it, showed unusual dynamism, not just in SEA but at continental scale". I have a theory but first we should do this: "It'd be interesting that both of us looked in detail at the M subgroups because the expansion of N corresponds, IMO, not to the M stage but to two steps downstream (later in time)". I agree. let's do it. M is very huge though, si it will be a big task.

     
  50. terryt

    November 21, 2010 at 1:27 am

    "ignore the fact that N does not show any sign of expansion for four full mutations/periods (at least 12,000 years, probably 16-20 millennia). And then suddenly it expands everywhere at the same time (according to you)". It certainly expanded rapidly. But the interesting thing is that we can see the route it took, as long as we're prepared to concede it may not have moved far from Africa before it suddenly 'expanded everywhere'. A haplogroup's long tail doesn't necessarily indicate any previous reduction in population numbers, just a population of restricted numbers over a period. Either situation could be described as 'a bottleneck', a restriction in population expansion. "it's clear that N and R right after it, showed unusual dynamism" Surely it is not usually the individual downstream haplogroups that migrate. For example let's look at mtDNA R, a haplogroup we both agree coalesced in SE Asia. Surely it is extremely unlikely the U, or even R0, were fully formed in SE Asia. It is much more likely that the group that departed from SE Asia contained a majority of what we might call, for convenience, mtDNA R*. Once this R* population had expanded U coalesced in SW Asia and replaced its parent potential haplogroup R*, through drift in the local population. Same with R0. Same with the other R haplogroups. In SE Asia itself R22 and R23 replaced their parent haplogroup R* along with many other pre-R22 and pre-R23 potential haplogroups. They have a very long tail but are found very near to where R itself coalesced. So their tail did not develope during a long migration. The same process presumably holds for mtDNA N, and M. I'll get back to you on M when I have time.

     

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: