March 10, 2014

Dark pigmentation of Eneolithic and Bronze Age kurgan groups from eastern Europe

This is a very exciting new study that seems to parallel some results from early west Europeans. The authors invoke selection as a possible cause for the massive change in frequency between the Bronze Age and present-day Ukrainians.

An invocation of selection as an explanation requires evidence population continuity, otherwise changes in allele frequency may involve migration of a new frequency-differentiated new population; for example, the massive change in pigmentation in North America over the last 500 years is not due to selection but to migration of Europeans. The authors cannot reject population continuity on the basis of mtDNA haplogroup frequencies, although autosomal data may be more informative for that purpose.

In any case, the fact that the limited sample from western Europe and the much more extensive sample from eastern Europe both show a darker pigmentation than modern Europeans does suggest that interesting changes happened in Europe over the last few thousand years and samples from more recent time periods may better determine the pace of this change.

From the paper:
In sum, a combination of selective pressures associated with living in northern latitudes, the adoption of an agriculturalist diet, and assortative mating may sufficiently explain the observed change from a darker phenotype during the Eneolithic/Early Bronze age to a generally lighter one in modern Eastern Europeans, although other selective factors cannot be discounted. The selection coefficients inferred directly from serially sampled data at these pigmentation loci range from 2 to 10% and are among the strongest signals of recent selection in humans.
UPDATE: 

The classical Greeks did of course notice that the inhabitants of the north Pontic hinterland, collectively known as Scythians, were extraordinarily light-pigmented. This would imply that major pigmentation change occurred in the steppe over a time span of Bronze Age-Classical Antiquity rather than Bronze Age-present; this would imply even higher selection coefficients (if selection over a population exhibiting continuity is at play).

The Scythians were also thought to be recent arrivals from the east so it is not clear if they were descended from the Bronze Age population of eastern Europe; the crazy selection coefficients that would need to be assumed if there was indeed population continuity might imply that Herodotus got it right again, and the Scythians did in fact arrive from elsewhere. That would of course also imply that people from Central Asia and Siberia (where the Scythians may have come from) were originally lighter than Europeans which does find support from an older study on southern Siberian remains. Ironically, if that is the case, it would mean that the famous light-pigmented mummies of different parts of Inner Asia may not be long-lost European descendants -- as it has sometimes been presumed on the basis of modern-day clines of pigmentation. As usual, ancient DNA continues to surprise.

PNAS doi: 10.1073/pnas.1316513111

Direct evidence for positive selection of skin, hair, and eye pigmentation in Europeans during the last 5,000 y

Sandra Wilde et al.

Eye, hair, and skin pigmentation are highly variable in humans, particularly in western Eurasian populations. This diversity may be explained by population history, the relaxation of selection pressures, or positive selection. To investigate whether positive natural selection is responsible for depigmentation within Europe, we estimated the strength of selection acting on three genes known to have significant effects on human pigmentation. In a direct approach, these estimates were made using ancient DNA from prehistoric Europeans and computer simulations. This allowed us to determine selection coefficients for a precisely bounded period in the deep past. Our results indicate that strong selection has been operating on pigmentation-related genes within western Eurasia for the past 5,000 y.

Link

51 comments:

andrew said...

By my calculations, if the time frame were 83 generations (until the time of classical Greece) instead of 172 (until the present), the selection coefficient would have to be 0.04-0.20, i.e. about twice as large. By comparison, the selection coefficient of lactase persistence alleles in Europe has a 95% confidence interval of 0.008-0.018.

It is entirely plausible that skin pigmentation would have selective impact. But, there is no way that it would have an impact 5 to 10 times as great as lactase persistence in the relevant time period of the early Bronze age to early Iron Age in that particular part of Europe. It should be less selectively important than LP, not more, unless it is a proxy for something else, like superior technology or other extreme cultural advantages.

Admixture or population replacement, quite plausibly male dominated so that mtDNA profiles didn't change nearly so much, make much more sense to explain the change in this time period, which includes the periods before and after major Indo-European expansions in the region.

Bill Lipton said...

LOL I'll invoke the 2011 book "Grandpa Was A Deity" -- which has the Scythian connected to the Brahman and north to the Aryas from Central Asia.
There are multiple levels of connection which extend east and west -- the light skin would be connected to females in whatever population the overcame in their journey.

barakobama said...

Dienekes, the chart of these ancient east Europeans have frequencies like modern near easterns. They have a higher amount of the "light skin" mutation in gene TYR than modern Europeans and near easterns but less than modern Europeans in SNP SLC24A5 and around the same amount as modern near easterns. It is also somewhat similar to modern Sardinians(nearly full blooded descendants of early European farmers) who have the lowest amount of the one in SNP SLC24A5 in Europe at a little over 50%. Stuttgart did not have the one in SLC24A% either but did have the one in SNP TYR like 43% of these ancient east Europeans.

Also only 16% of them had the "blue eye mutation" in gene OCA2. Which probably means they had a high amount of EEF or just plain near eastern ancestry and a small amount of Mesolithic European ancestry. The reference European has 71% of this mutation which means they should be labeled as north European. These ancient east Europeans may have been typical EEF(like Stuttgart and Otzi) and are not the main ancestors of modern east Europeans.

I really doubt these ancient Ukrainians were the ancestors of the mainly light haired and eyed bronze age Indo Iranians from Asia. We know that light colored eyes had already developed and were popular in the Mesolithic, plus these ancient Indo Iranians had a very high amount of typical Mesolithic east European mtDNA haplogroups U5a, U4, and U2e. In my opinion they had a very similar genetic makeup as modern north Europeans especially north-east Europeans.

Dienekes I think you and many others make the assumption that everything that causes light skin in Europe has been discovered. The skin color of Loschbour and La Brana-1 is unknown, according to current knowledge they had dark skin that's what you should say. All three of the main 'European" light skin mutations are JUST AS POPULAR IN THE NEAR EAST AS IN EUROPE except the one in SNP SLC24A5 which is close to 100% in most of Europe and only around 50% in the near east. Current knowledge can not explain skin color difference between Europeans and near easterns and north Europeans(majority WHG+ANE) and south Europeans(majority near eastern).

I still don't think the correlation between light hair-eyes-skin-Mesolithic European ancestry is random. I think that once a genomes from Mesolithic Russia are sampled there will probably be a high amount of light eyes and light hair. I don't think they will have the mutations associated with European light skin(they are near eastern not European) but those genomes will be evidence there are unknown factors to creating light skin in Europe which have Mesolithic/Upper Palaeolithic origin.

pconroy said...

quote:
"it would mean that the famous light-pigmented mummies of different parts of Inner Asia may not be long-lost European descendants"

Exactly!!!

That's why I suggested a few weeks ago on another blog, that the ANE=ANI and the Kalash/Nuristani may be a remnant of the original light-pigmented people, not a far-flung European remnant...

Grognard said...

Would imply some serious 'elite dominance' if these numbers prove correct.

I am guessing ancient peoples would be more heterogenous compared to near neighbors, though. Finding some dark tribe next to a lighter tribe would not surprise me much as travel was slow and language and cultural barriers more serious.

curious student said...

Or just higher reproduktion due to farming technology...

Va_Highlander said...

Dienekes:

"That would of course also imply that people from Central Asia and Siberia (where the Scythians may have come from) were originally lighter than Europeans which does find support from an older study on southern Siberian remains. Ironically, if that is the case, it would mean that the famous light-pigmented mummies of different parts of Inner Asia may not be long-lost European descendants -- as it has sometimes been presumed on the basis of modern-day clines of pigmentation."

Which, if true, should surprise no one. Population movements across Eurasia, from deep prehistory to historic times, have generally been east to west. The only proven exceptions seem to be the expansion of the oasis culture in the Early Bronze Age and the Russian expansion eastward in late historical times.

Dienekes said...

Stuttgart did not have the one in SLC24A% either but did have the one in SNP TYR like 43% of these ancient east Europeans.

That's 4.3%. They have lower frequency in all 3 SNPs.

About Time said...

Classical Scythians (the ones noted by Greeks as light pigmented) came from Central Asia and maybe West Asia / Mideast before that (based on new mtDNA hg that show up in Iron Age West Siberia).

They were active at fringes of West Asia (Assyria, Persia) prior to that, and some (like Partatua) married in with royal families of the major Middle Eastern empires in their heyday.

They were not the same as 3000-2000 BC Kurgan populations, which were more Punjabi like morphologically.

Derek P. Moore said...

Eratosthenes was the first to locate Scythia on a map, and he placed it near modern day Germany. Later Greek and Egyptian map makers relocated Scythia to the Steppes regions.

Many nations' historical traditions document descent from the Scythians, but these traditions are mostly disregarded even as genetics begins to corroborate the migration stories they contain.

Grey said...

General point:

If the selection pressure for light skin was the shift to agriculture from foraging because of the consequent reduction of meat in the diet why would that selection pressure apply more strongly to pastoralists?

.

@Va_Highlander

"Population movements across Eurasia, from deep prehistory to historic times, have generally been east to west."

How did MA1 get there then?

The very earliest moves were plainly west to east with the mammoth steppe although the main direction of travel reversed later.

barakobama said...

"That's 4.3%. They have lower frequency in all 3 SNPs.
"

Thanks for the correction Dienekes. Wow, that's very surprising for any modern European or near eastern population. I want to look over the study before making any big conclusions. I am looking at the mtDNA right now and there is a surprisingly high amount of mtDNA U 18/62(I didn't count ones with unknown HG) and by just glancing at their HVR1 differences to rCRS I can see most of the U's have U5a1, like bronze and iron age mainly light eyed and haired Indo Iranians and Mesolithic Russian hunter gatherers.

Is this that German study, that said copper age Pontiac steppe people were very dark eyed.

Raimo Kangasniemi said...

Later steppe groups tended to be heterogenous to differing degrees, so I don't think that we should assume that Scythians were necessarily much different. Culture and political organization might well have been more important than shared genealogical background - which could always be invented, as is common with nomads and semi-nomads when the need arises to form larger groups and connections between groups. When it comes to Greek claims, we have to remember that they were basically using stereotypes to separate different people (in the manner of Berbers have red hair and blue eyes) and would have taken the more extreme examples as the basis of that stereotype. So, we can assume that there would have been pretty light-skinned - to the eyes of the Greeks - Scythians, but we should not hastily assume that light-skin was the norm. It would have just been more common than Greeks would have been used to.

Nirjhar007 said...

@Pconroy
''That's why I suggested a few weeks ago on another blog, that the ANE=ANI and the Kalash/Nuristani may be a remnant of the original light-pigmented people, not a far-flung European remnant...''
Well i think it is quite logical to suggest!
Good Day.

Grey said...

MC1R or IRF4?

eurologist said...

"Current knowledge can not explain skin color difference between Europeans and near easterns and north Europeans(majority WHG+ANE) and south Europeans(majority near eastern)."

Barak,

As you might know, that's what I have been saying for quite a while. Current oversimplified methodologies to determine skin (and hair, and eye) color from just a couple of sites is ludicrous, given that we know better, and have known better for a while.

We know that the often-used markers don't explain northern light skin, and we know that light skin is imperative to survive at high latitude (or, more precisely, in areas of low medium-term and long-term UV exposure, also due to the requirement of protective clothing).

There is no question that northern-central and northern Mesolithic Europeans had such features way before any Neolithic genetic impact, same way NE Asians have had for a long time, but based on entirely different mutations.

Rokus said...

'Population movements across Eurasia, from deep prehistory to historic times, have generally been east to west. The only proven exceptions seem to be the expansion of the oasis culture in the Early Bronze Age and the Russian expansion eastward in late historical times.'
Seems you have been missing most of the relevant results lately. West to east migrations are attested genetically since the earliest times: the Altai Neanderthal and Mal'ta boy both have a western connotation. East to west migrations that involve Amerindian admixtures may be interpreted as backmigrations of a differentiated population without East Asian component. I think you are also missing the Indo-Iranian presence in northern Eurasia, that clearly has a western origin? The Russian expansion to Siberia was not the exception, it was the last chapter. The prehistoric west-eurasian presence in Eurasia was such that historic 'Asiatic' expansions rarely reached the west without being already heavy west-eurasian admixtured.
BTW, we may conclude from the result of the present study that the relationship between the 'Skeletal material from 150 Eneolithic and Bronze Age individuals from the
west and north Pontic region', and Bouakaze's 'from 25 archeological human remains from southern central Siberia dating from the Bronze and Iron
Ages' that 'most probably had typical European pigment features, i.e., blue or green eye color, light hair color and skin type, and were likely of European
individual ancestry', is not at all that obvious. Especially since light pigmentation was already present long before in LBK Europe.

Nathan said...

Pconroy
"that's why I suggested a few weeks ago on another blog, that the ANE=ANI and the Kalash/Nuristani may be a remnant of the original light-pigmented people, not a far-flung European remnant... "

Am I right in concluding from your above comment that ANI (Ancestral North Indian) is made up 1 linguistic group, that being Indo-Iranian?

It is far more likely ANI is made of multiple linguistic groups that arrived in the SubContinent at different times.

If we take ANI to be Indo-Iranian (Kalash/Nuristani as you mention), then how does one account for the not insignificant ANI in Southern India a region that has mostly escaped Aryanization.

If ANI is Indo-Iranian only, that would invariably mean Dravidian,Munda,Veddah and Onge are ASI yet the Onge who are stated to be unadmixed ASI are distinct from the other groups.

I believe based on the history of the region and the genetic profile of the various peoples, that both ANI and ASI are composed of multiple distinct groups .

Va_Highlander said...

Grey:

"How did MA1 get there then?"

I think it was Terry who suggested that the deep ancestry of MA-1 may lie in southern Central Asia or northern South Asia. That seems reasonable and a quite different trajectory than those usually proposed west-to-east across the Eurasian continent.

Simon_W said...

Interesting, this makes me involuntarily think of Stanislav Grigoriev's theory that most of the European Indo-Europeans (save the Old Balkan languages) reached Europe via Siberia, in the middle Bronze Age. David Anthony thinks the Celts and Italics descended from Yamnaya groups in the Carpathian basin. But this wouldn't explain how the Celts got lighter pigmented.

Tobus said...

@andrew: By my calculations, if the time frame were 83 generations (until the time of classical Greece) instead of 172 (until the present), the selection coefficient would have to be 0.04-0.20, i.e. about twice as large

The authors state that the process was already ongoing at the time of their samples, and the sweep is "estimated to have begun between 11,000 and 19,000 y ago". You should be using a start date in this range, not the date of the earliest sample in the study.

Dienekes said...

The authors state that the process was already ongoing at the time of their samples, and the sweep is "estimated to have begun between 11,000 and 19,000 y ago". You should be using a start date in this range, not the date of the earliest sample in the study.

The forward simulations in the paper take place from the date of the samples to the present.

The quote you refer to is a simple literature reference.

Tobus said...

@Dienekes/@andrew:The forward simulations in the paper take place from the date of the samples to the present.

I stand corrected - I mistakenly thought it was referring the entire sweep, not just the last part.

@eurologist:
Current oversimplified methodologies to determine skin (and hair, and eye) color from just a couple of sites is ludicrous, given that we know better, and have known better for a while.

Multiple studies by multiple labs on multiple populations have confirmed that certain alleles cause skin colour to be darker or lighter to various degrees. The combined affect of these alleles has been quantified into an absolute value (about 30% of the global range), and we can use this to estimate a baseline for the *lightest* a population can possibly be if it has the "dark" known variants, by assuming all of the unknown factors are the "light" versions and appling the known darkening effect to the lightest samples we have. In the cases of MA-1, La Brana and Loschbour, the skin tone they would have if all the unknown elements were "light" variants is roughly in the middle of the range of modern South Asians (using measurements from Jablonski 2000). Just because we don't know the whole picture, that doesn't mean we can't put together the pieces we do know to point us in the right direction.

we know that light skin is imperative to survive at high latitude

Except Siberia, Alaska, Aleutian Islands, North America...

There is no question that northern-central and northern Mesolithic Europeans had such features way before any Neolithic genetic impact

It's extremely far-fetched to think that the most strongly selected phenotype is the history of modern humans was redundant - which it would be if the phenotype was already widespread. The fact the selection for it was so strong indicates it provided a significant difference to its carriers - we wouldn't see this strength if it was turning white people white.

Rajarshi Banerjee said...

Thinking about naturally occurring vitamin D in milk and cultured milk, fish. is a negative correlation between pastoralism or other forms of subsistence and light skin expected or not?

anomalous positive selection may also be prompted by sudden scarcity.

Grey said...

@Va_Highlander
@Simon_W

I think it has to be divided into two stages: the HG stage and the pastoralist stage.

For MA1's ancestral connection to both Amerindians and Europeans to work you need the move onto the mammoth steppe to have started at the western edge (which just happens to be where the mammoth steppe extended furthest south) or the entirety of Europe to have been replaced at some point by people from the East.

It seems more likely to me that the mammoth steppe was originally dominated by people coming from the western end of the range simply because that's where it extended furthest south.

The center is second most plausible and Siberia the least.

Although if the mammoth steppe was its own little eco niche for a long time with a lot of movement back and forth - including into (and possibly back out of) America at certain times - then at some levels it doesn't matter much where the start point was as there would simply be an ANE population extending from France to Siberia.

This would make Europeans partly a remnant of that "Hyperborean" population that originally lived in north-central and north-east Eurasia as well.

.

The second stage is the pastoralist bit.

The problem with that not being west (or west-central maybe) to east is it involves the dudes north of the Black Sea domesticating horses, trading them and then waiting to be conquered by the people they sold their horses to.

I think that is a very silly theory.

Although the direction of travel reversed later to become east to west: Iranians, Huns, Turks, Mongols etc, I think the earliest waves must have come from the people who first domesticated horses which means west (or west-central) to east.

terryt said...

"West to east migrations are attested genetically since the earliest times: the Altai Neanderthal and Mal'ta boy both have a western connotation".

Very ancient. I would guess such a movement carried Y-DNA C and mt-DNA N eastward originally. It is difficult to make a convincing case that either moved eastward through South Asia and yet both apparently were the first haplogroups to reach Australia.

"I believe based on the history of the region and the genetic profile of the various peoples, that both ANI and ASI are composed of multiple distinct groups".

I'd be very surprised if it turns out to be anything other than that.

Va_Highlander said...

Grey:

"The second stage is the pastoralist bit.

"The problem with that not being west (or west-central maybe) to east is it involves the dudes north of the Black Sea domesticating horses, trading them and then waiting to be conquered by the people they sold their horses to."


That would indeed be silly. Luckily, the "dudes north of the Black Sea" do not appear to have been the first to domesticate the horse, an event that seems to have occurred several hundred miles to the east.

A second difficulty arises when we note the archaeological time line. The horse was first domesticated as a food source, long before the development of nomadic pastoralism on the Eurasian steppe.

About Time said...

@Grey, interesting point about mammoth steppe

Maybe Kurgans were a specialized "third group" that kept one foot in the N world (hg tending, leaning to Europe) and the other foot in the S world (Asia, leaning to state societies). They would be specialist go betweens, maybe mixed in origins but also tending to develop their own niche.

That's exactly what we see with Scythians, Goths, Cinmerians, etc. They sometimes served the empires as muscle/military specialists, sometimes traded (Medes were famous horse breeders), and sometimes found themselves at the top of the heap marrying princesses etc (Madyes was king of the Medes).

So yes in a sense they were "conquered", but really it can also be said they were "drawn in" to the southern world by their own actions (and benefited from it--becoming much more than backwoods horse breeders).

Simon_W said...

@ Grey

Judging from the recent Lazaridis et al. study, Western European Hunter-Gatherers and ANE were related, true, yet it's clear from that study that modern Europeans received an additional ANE-related input that was neither from Western European nor from Scandinavian Hunter-Gatherers, nor from the Early Farmers. There was no complete replacement for sure, but some additional ANE admixture. When exactly, is a question that needs to be explored further.

The discussion has often turned around the Corded Ware culture, it's putative key role in the Indo-Europeanization of central to eastern Europe, and it's possibly eastern origins (the latter assumption having come under heavy critique by archeologists, but to some extent vindicated by recent mt-DNA and y-DNA evidence). In comparison, Grigoriev's suggestion to look for later, bronze age connections to the east, and a much later arrival of Italo-Celtic, and Germano-Balto-Slavic groups looks refreshingly new and unconventional.

As for the horse domestication, I'm not sure that this first happened north of the Black Sea. I think there was some recent evidence pointing more towards the Botai culture in Kasachstan. And anyway, the movements Grigoriev thinks about were much later than the suggested time of the horse domestication in the Botai culture, I think the latter cannot rule out these much later events. And for what it's worth, the middle Bronze Age is the time when there is the first unambiguous evidence for horse riding and chariotry in Europe.

Grey said...

@Rajarshi

"Thinking about naturally occurring vitamin D in milk and cultured milk, fish. is a negative correlation between pastoralism or other forms of subsistence and light skin expected or not?"

The things I've read say Vitamin D in milk is only at trace levels but I haven't seen a calculation of how much is enough?

Or even if it wasn't enough what kind of percentage milk could provide if a population drank a lot of it: 1% or 50%?

So I don't think there's a clear answer except that pastoralists should have *less* selection pressure for light skin even if it is only slightly less.

(Or maybe not if farmers ate a lot more eggs - which are the other good source of Vitamin D6 other than fish - but that's the point really; what the details of the diet were.)

terryt said...

"For MA1's ancestral connection to both Amerindians and Europeans to work you need the move onto the mammoth steppe to have started at the western edge (which just happens to be where the mammoth steppe extended furthest south)"

I think that is the only possible conclusion we can draw from the data. And, in spite of what German claims, the first Americans also have their deeper origin from the same mammoth steppe population.

Grey said...

Oops, apparently no eggs at that time in those regions.

http://en.wikipedia.org/wiki/Chicken

"From India, the domesticated chicken was imported to Lydia in western Asia Minor, and to Greece by the fifth century BC.[3] Fowl had been known in Egypt since the mid-15th century BC, with the "bird that gives birth every day" having come to Egypt from the land between Syria and Shinar, Babylonia, according to the annals of Thutmose III."

Hector said...

The new ISOGG tree reveals that C4(Australian) is also within C1-C2-C5-C6 clade. So C is divided into C3 and the rest. It will take a fairly elaborate theory to claim that C6 parted with the rest in West Asia(thus West Eurasian all the way).

You guys really make me laugh big time. Especially Rokus who fantasizes about the supremacy of the Caucasoid race in Siberia from time immemorial... If you can claim R as West Eurasian, you should admit it is Papuan originally.(both view points are wrong but to say in a manner of speaking that is your way)

Rokus said...

'The new ISOGG tree reveals that C4(Australian) is also within C1-C2-C5-C6 clade. So C is divided into C3 and the rest. It will take a fairly elaborate theory to claim that C6 parted with the rest in West Asia(thus West Eurasian all the way).'

Indeed, last year november I already said so:
'Australian C4 is still suspected to make a Holocene cluster with Indian C* (Redd 2002, Pugach 2013), what is a feat since this would imply considerable Y-DNA replacement from a single Indian source not so very long ago, against 11% contemporaneous Indian admixture. I wonder if Indian C* and Australian C4 are truly different from the disparate C-K29 branch or just in need of more investigation.'

Moreover, I posed the question: 'Even without excluding C4 as a reliable ancient marker in Oceania, C3 against CXC3 might be indicative of an originally northern ancestral branching that may be equally reminiscent in MP?'

In Restoring the Gap, Part 1 – The Delayed Non-African Expansions of Modernity I elaborated this suspicion and framed it up with the overall genetic relatedness of non-African modern humans to conclude this pattern could indeed br confirmed and actually part of a 'continuous process in the period between the earliest Eurasian AMH and Gravettian expansions, that spawned new CF-derived YDNA lineages (C1, G, H, IJ, L, T, MP, NO) all the time from a central or northern position all over Eurasia.'

Hector: 'You guys really make me laugh big time. Especially Rokus who fantasizes about the supremacy of the Caucasoid race in Siberia from time immemorial... If you can claim R as West Eurasian, you should admit it is Papuan originally.(both view points are wrong but to say in a manner of speaking that is your way)'
If you'd read my blog you could have known that in my view Cromagnoid modernity (I suppose this transforms into despicable "Caucasoid" personal "race" attacks in your paranoid attitude) already carried an important Southeast Asian (including southern Chinese) element. However, ever since non-African populations find an important Upper Paleolithic source in basal populations of West Eurasian signature.

Chad Rohlfsen said...

Indo European trash pits go from almost all fish, to almost all bovine in the Bronze Age. This sudden drop of a great Vitamin D source could be the culprit.

Chad Rohlfsen said...

Excuse me, that was late Neolithic/Chalcolithic. Perhaps enough time to begin the skin lightening. Did anyone test for SLC24A5, to see how common it was?

terryt said...

"The new ISOGG tree reveals that C4(Australian) is also within C1-C2-C5-C6 clade".

Thanks for that reminder. I was told the new phylogeny would be up by now but had forgotten to check it out.

"So C is divided into C3 and the rest".

That makes it impossible that Y-DNA C moved east through South Asia. The first split is between East Asia and the rest, with the rest being divided into one each of Wallacean, Australian, South Asian and Central Eurasian.

Rokus said...

"So C is divided into C3 and the rest".

'That makes it impossible that Y-DNA C moved east through South Asia. The first split is between East Asia and the rest, with the rest being divided into one each of Wallacean, Australian, South Asian and Central Eurasian.'

I don't think this conclusion is justified. The breakup of YDNA C(xC3) (nowadays C1) into the four main subclades must have been subsequent to an initial expansion that links Oceania and India with Central Asia. This pattern is not unlike the breakup of YDNA MP into an Oceanian subclade M and the Indian/north Eurasian subclades of P. This means ancestral C1 and ancestral MP must have moved into one direction or the other. At least in the case of Australian C1 it has been fairly substantiated there was a Holocene link with India. Moreover, we have to be conscient that modern subclades are the result of continued drift and population history that strongly tend to homogenize regional haplogroups into the modern units of reduced MRCA's. In this view North Eurasian subclades just experienced a longer period of shared drift. In other words, the resolution of YDNA subclades has all appearance to be too shallow to be sure of any migrational direction at all. We have to dig deeper into the patterns that link Oceania, India and Northern Eurasia to find other circumstancial evidence that could lead to the identification of direction.

terryt said...

"The breakup of YDNA C(xC3) (nowadays C1) into the four main subclades must have been subsequent to an initial expansion that links Oceania and India with Central Asia".

Possibly, although I am informed that C(xC3)'s diversification may not actually be star-like. Australian C's position has not been fully determined yet. Apart from that we have to consider Indian C1b's distribution. It forms two branches neither of which are evenly, or even randomly, distributed through the subcontinent. That is not what we'd expect if its presence there was ancient. C1b2 especially is confined to Bangladesh and even C1b1 is concentrated in Gujarat. And Eurasian C1a has all the hallmarks of an Upper Paleolithic expansion, whereas we can be sure Australian C1d arrived on the continent before the UP.

"This means ancestral C1 and ancestral MP must have moved into one direction or the other".

I see no reason at all to believe they moved together, anywhere. However I agree C1a probably moved onto the steppe with P. But by that time humans were already widespread.

"At least in the case of Australian C1 it has been fairly substantiated there was a Holocene link with India".

The Holocene link between India and Australia has nothing to do with Y-DNA C. Y-DNA C's diversification is much more ancient than the Holocene.
A connection between India and Australia appears through the presence in both regions of mt-DNA M42, with separate branches found in the two countries. But we don't know the direction of the Holocene link at all, especially considering M42's closest relation is M74, a southern Chinese haplogroup.

"We have to dig deeper into the patterns that link Oceania, India and Northern Eurasia to find other circumstancial evidence that could lead to the identification of direction".

That deeper digging has basically already been done. Polynesian origins are well understood with complete consistency between genes, archeology and linguistics. What we see from well before they originated is a strong connection between Wallacea and Oceania with a lesser connection between these two and Australia. Not surprising at all. On the other hand we find very little indeed to link any of those three regions with India, even though many people have been searching for such a link (and assuming such a link exists) since haplogroups were first classified. The connection between the three regions is much closer to SE and East Asia and almost non-existent with India.

Rokus said...

BTW, I wonder if Australian former C4 and possibly Oceanian former C2 attest a single link with the earliest AMH of northern Eurasia (C1), or indeed indicate a rather prolonged contact if C4 could be confirmed as a Holocene innovation from South Asia. I hope one day Tasmanian aDNA could be tested to find out, since apparently Tasmania was excempted from the Holocene innovations of the Australian mainland. I suppose Australians would have been much more like Tasmanians and Papuans before this - and the advent of YDNA C4! - happened.

Narcissus said...
This comment has been removed by the author.
terryt said...

"I hope one day Tasmanian aDNA could be tested to find out, since apparently Tasmania was excempted from the Holocene innovations of the Australian mainland".

It is extremely doubtful that Australian C1d-M347 is anywhere near as recent as 'Holocene'. It is the most common and most widely distributed Australian Aboriginal Y-DNA, more than 60% in most groups according to Ebizur's data from some years ago. Around the same time Ebizur claimed an age of 42,000-61,000 years for C1c-M38 (as it is now called) in Southern Wallacea. It cannot be Holocene there either.

"I wonder if Australian former C4 and possibly Oceanian former C2 attest a single link with the earliest AMH of northern Eurasia (C1)"

Qyuite possibly so. But the link must have been much earlier than the Holocene.

"I suppose Australians would have been much more like Tasmanians and Papuans before this - and the advent of YDNA C4! - happened".

Australian C has all the characteristics of being the
oldest surviving Y-DNA in Australia, probably the first Y-DNA there. Any change in Australia is not through C's arrival. The 'later' arrival in Australia is almost certainly K, mostly concentrated in the northwest of the continent. And I think you are forgetting that C of any sort is hardly present in New Guinea and any C1c-M38 that is there is confined to the Bird's Head and the northern coastline. It is almost certainly associated with the Austronesian expansion.

"if C4 could be confirmed as a Holocene innovation from South Asia".

Certainly not 'Holocene contact'. In fact doubtfully 'from South Asia'. From the other day:

"The breakup of YDNA C(xC3) (nowadays C1) into the four main subclades must have been subsequent to an initial expansion that links Oceania and India with Central Asia".

But C(xC3) (nowadays C1) most likely formed somewhere near where C2-M217 formed. I fully accept it is possible for a single male to move a huge distance and leave a Y-DNA legacy in a new region. However to me it is far more likely that a genetic mutation builds up in a restricted population before being able to expand when conditions allow. Such expansion is likely to be a slow geographic expansion rather than a small group moving huge distances. To me is is almost certain C1 and C2 emerged originally from the same region. C2 almost certainly first formed somewhere to the northeast of the Tibetan Plateau. C1 may have formed at the western end of the plateau and been separated as climate cooled. However its prevalence in Southern Wallacea and Australia, and relative scarcity in Southern and Central Asia, argues against a western origin for C1-F3393 and in favour of an eastern origin.

Annie Mouse said...

I would have expected the first Y haplogroup in Australasia to be present in both PNG and Australia. But there is no clear candidate.

Only K is common to both Australia and PNG. And K is less common in the highlands than the coast of PNG. I would have expected the original group to be retained more in the highlands than the coast. In PNG it is M and S that are higher in the inland than the coast. Neither of which have been found in Australians.

In Australia K seems higher in the South (60% in South Australia), which might indicate that it arrived before C1d (aka C4). As new immigrants arrive from the north.

Because of the general instability of Y haplogroups, C1d (aka C4) has to be a potential candidate given the lack of a very strong alternative. But in my opinion K is just pipping it at the post because it is still present in both areas.

Rokus said...

'C1 may have formed at the western end of the plateau and been separated as climate cooled. However its prevalence in Southern Wallacea and Australia, and relative scarcity in Southern and Central Asia, argues against a western origin for C1-F3393 and in favour of an eastern origin.'

The Indian subcontinent has often been pictured as having suffered major post-Neolithic immigration from Central Asia, especially from the Dravidian and Indo-Aryan kind. The flipside of this must be that pre-Neolithic populations became largely overruled by new elements, and that nowadays few traces remain of the Paleolithic situation. Sri Lanka still has a degenerated Veddah element that is thought to have been much more important in the past, including India. The Andaman Islands are another example of ancient South Asian populations that hardly represent the current Indian populations nowadays.
It is from one of these s ancient (Veddoid) populations that Redd et al. (2002) draws his hypothesis in Gene Flow from the Indian Subcontinent to Australia. Evidence from the Y Chromosome.
Unfortunately Aboriginals even worse than Native Americans to cooperate in genetic investigations, so we still don't know much of their structure. However, I am very inclined to take Redd et al serious to infer more recent Aboriginal contact with South Asia, including massive YDNA replacements.
This new point of view is also debated by John Hawks, who announced a lecture where one of the subjects will be "the surprising influence of South Asians upon Aboriginal Australians."

Hector said...

A Filipino C* has been recently tested and it is reported that he also harbors K29. So he will be reclassified as C1* as of now.

It may be that many, if not most, C*s are actually C1*. It will be interesting if we can find C* closer to C2 than to C1.

There seems to be confusion over MRCA and bifurcation point. Though theoretically tri or higher-furcation is possible most Y-lineages diverge through bifurcation. And it actually represents a single farther and two sons. Thus naturally MRCA of C1 and C2(just about every C except rare cases) was a man, and he had two sons that represent C1 and C2 lineage each. However when you consider the origin of C1 and C2 it is usually about MRCA of C1 and likewise for C2. And they can be temporally and geographically quite distant.

terryt said...

"I would have expected the first Y haplogroup in Australasia to be present in both PNG and Australia. But there is no clear candidate".

The same goes for mt-DNA. That suggests the two regions were settled separately to a large extent.

"In Australia K seems higher in the South (60% in South Australia), which might indicate that it arrived before C1d (aka C4)".

Where did you get that information from? The data I have suggests the opposite.

"Thus naturally MRCA of C1 and C2(just about every C except rare cases) was a man, and he had two sons that represent C1 and C2 lineage each".

I doubt it is anywhere near as simple as that. If we regard the ancestor of both C1 and C2 as C* then there is no need at all that the separate ancestors of C1 and C2 were literally brothers. The mutations that gave rise separately to C1 and C2 could have been separated by a considerable amount of time. However I agree it is convenient to regard the two ancestors as 'brothers'.

"The Andaman Islands are another example of ancient South Asian populations that hardly represent the current Indian populations nowadays".

We actually have no evidence that they were ever closely connected to Indian populations, just assumptions. Y-DNA D is extremely uncommon in India and where found is most closely associated with Tibeto-Burman speakers who are almost certainly relatively recent arrivals in India.

"Unfortunately Aboriginals even worse than Native Americans to cooperate in genetic investigations, so we still don't know much of their structure".

We do have a reasonable idea of their mt-DNA distribution although samples from the far northeast and southwest are lacking. I printed off a map some time ago but unfortunately can't find a link to the article. It provides pie charts for four regions: Kalumbaru, NT/Arnhem Land, Desert/Yuendumu and Riverine. Easch region has a distinct proportion of mt-DNA haplogroups.

"This new point of view is also debated by John Hawks, who announced a lecture where one of the subjects will be 'the surprising influence of South Asians upon Aboriginal Australians'."

That will be interesting no doubt.

terryt said...

Found the map. It's in a pay per view paper, that's why it was so difficult to find. The map is available to the left:

http://www.sciencedirect.com/science/article/pii/S1040618211002278

Annie Mouse said...

@Terry
("In Australia K seems higher in the South (60% in South Australia), which might indicate that it arrived before C1d (aka C4)".

Where did you get that information from? The data I have suggests the opposite.)

http://www.ncbi.nlm.nih.gov/pubmed/22297081

Hector said...

When I said C1 and C2 were brothers I meant it in the literal sense. Here C1 and C2 are not necessarily as defined currently but represnet the maximum expansion of each lineage absorbing some if not most of C*.

To explain this, suppose some C* is M130+ but shares a hitherto undiscovered SNP with C1 but not C2. C1 will be expanded(updated) with that SNP as upstream to all of SNPs defining the previous C1.(that the C* in question harbors none of the SNPs currently defining C1 is naturally assumed by the original assignment of it as C*)
(Of course there will be C*s, especially those that are M216+ but M130-, that will remain outside of C1 and C2 even when they are maximally expanded. They will in turn update the C haplotree in the future.
For example
C1 may be redefined as M130+
C2 will be redefined by yet to be discovered SNPs
The current C1 will become C1a so on so forth.)

Anyway C1 and C2 will have a point where they share a MRCA and at that point it literally represents a father and two sons.

However this MRCA is in the genealogical sense. He exists genealogically but the man may not be identifiable precisely by SNPs(ie if both of the two sons are exact matches of their father in Y chromosome). But the error of margin cannot exceed a few generations when the Y chromosomes are fully sequenced.(ie 30 Mbp or so)

terryt said...

@ Annie Mouse:

Thanks. I hadn't seen that paper.

"In S.A., K* (∼60%) has a much higher frequency than C4 (∼40%), and the subgroup of C4, C4(DYS390.1del), comprised only 17%".

Interesting. That doesn't fit the data from Arnhem Land and the central desert, the only regions I had data on where K is present at 30% and 17% respectively. South Australia would most likely be an extension of the central desert so the results are confusing. It would be great to know the relative distributions of the two haplogroups through Australia. Another thing to keep in mind is that 'K' has three branches and its relationship to MNOPS is uncertain. I think most Australian K is K* while K1 is present in both Australia and New Guinea, K2 is found in the Melanesian islands which were presumably settled after New Guinea, while K3 is notably present in Bali.

@ Hector:

"Anyway C1 and C2 will have a point where they share a MRCA and at that point it literally represents a father and two sons".

As I tried to explain before, I doubt it is ever as simple as that.

terryt said...

"Only K is common to both Australia and PNG. And K is less common in the highlands than the coast of PNG. I would have expected the original group to be retained more in the highlands than the coast. In PNG it is M and S that are higher in the inland than the coast. Neither of which have been found in Australians".

Yes. I must point out that it is not only the relative distributions of K and C in Australia/New Guinea that leads me to believe C was the first haplogroup into the region. I think we can assume that the first crossing was made by the shortest route: Timor (Southern Wallacea) to the Kimberly. Later groups moved further along the coast. C1d is exclusively Australian and, although C1c is found in New Guinea it is primarily in the Bird's Head, then along the north coast, and out into the Pacific. The latter in the form C1c1 with C1c1a in Polynesia, about 4000 years old. Some years ago Ebizur calculated that C1c was some 45,000 years old in Southern Wallacea but just some 8-10,000 years old in New Guinea/Melanesia. Presumably C1 arrived in Australia very soon after it arrived in Southern Wallacea. On the other hand Ebizur remarked that the most widespread clades of K, M and S in New Guinea/Melanesia were surprisingly young, nothing older than 10,000 years. Of course that doesn't mean the ancestral form was absent from the region, just that it was present only in isolated patches, unable to expand until technology allowed a more efficient use of New Guinea/Melanesian resources. The improved technology perhaps associated with the expansion of the Hoabinhian.

As a result of all the above I long ago concluded C was the earlier haplogroup into Australia while New Guinea was populated some period later, and Melanesia later still.