PDA

View Full Version : Genetic Drift Sharing: Bell Beaker & WHG



R.Rocca
04-03-2015, 01:23 PM
Haak's figure S7.7b shows populations that share the most genetic drift with Bell_Beaker_LN (or more specifically, German Bell Beakers). In order of most sharing, they are:

Group 1:


1. HungaryGamba_HG (a.k.a KO1, Neolithic of hunter-gatherer ancestry, halpogroup I2a)
2. Loschbour (based on an 8,000 year-old hunter-gatherers from Loschbour, Luxembourg, haplogroup I2a)
3. Motala_HG (based on seven 8,000 year old hunter-gatherers from Motala in Sweden., 5 of 5 males haplogroup I2)

Note: Interestingly these first three are Western or Scandinavian Hunter Gatherer groups and not EHG. All are I2.

Group 2:


4. Alberstedt_LN (possibly of mixed Corded Ware/Bell Beaker ancestry as per Haak, 2459-2345 cal BCE)
5. Esperstedt_MN (dated to 3360-3086 cal BCE, haplogroup I2a1b1a)
6. Karsdorf_LN (possibly Corded Ware, radiocarbon dating pending)
7. BenzigerodeHeimburg_LN (possibly all Bell Beaker)
8. Corded_Ware_LN (haplogroup R1a)
9. SwedenSkoglund_NHG (haplogroup I2)
10. Karelia_HG (haplogroup R1a)
11. Halberstadt_LBA (haplogroup R1a)
12. Unetice_EBA (3 of 3 males haplogroup I2)
13. Samara_HG (haplogroup R1b1)

Note: The bulk of this group is made up of possible Bell Beaker mixes or later groups with Bell Beaker ancestry. However, the Esperstedt_MN sample is the lone Neolithic sample and of interest because it groups with Spain_MN (2 of 2 males haplogroup I2a). It is not until the 10th (Karelia_HG) and 13th (Samara_HG) spot that we get to EHG samples.

Group 3:


14. Spain_MN (3900-3600 BCE, 2 of 2 samples haplogroup I2a)
15. Spain_EN (5311-5066 cal BCE, 1 sample R1b1, the other F*

Note: This group is interesting because Bell_Beaker_LN shares more genetic drift with these two Spanish Neolithic populations than all other Central European Neolithic samples except for the previously mentioned Esperstedt_MN. Of further interest, both male Spain_MN samples also belong to haplogroup I2a. One Spain_EN sample is I2a, the other is R1b1* and while the third sample could only be assigned to F*, he is a relative of the R1b1* sample and enough of the usual hunter/farmer haplogroups have been ruled out (e.g. G, I1, I2, T), that he could also be R1b1.

After that you get a mix of Neolithic and modern populations. Way down the list we have the HungaryGamba_CA (Baden Culture) sample, which is is a typical EEF, finding it's closest likeness to another Balkan Neolithic group in Starcevo_EN. Another Copper Age Central European, Otzi the Iceman, also shares his most genetic drift with Starcevo_EN. Theoretically, a Yamnaya movement westward should produce a Yamnaya mix with a HungaryGamba_CA/Iceman/Starcevo type population. Instead, we have the following:

Figure S9.8b the modeling of Bell_Beaker_LN as a mixture of earlier populations...

24.8% Karelia_HG + 75.2% Spain_MN
49.5% Spain_MN + 50.5% Yamnaya

So, what does this all mean for R1b? While the appearance of Yamnaya ancestry across all of Europe still makes a movement of R1b-L51 & R1b-L11 from east to west during the Late Neolithic very likely, these German Bell Beaker samples seem heavily mixed with Western European Hunter-Gatherers. They shed some of their EHG, more so that their distant R1a Corded Ware cousins. Unlike the Corded Ware samples, the German Bell Beakers do have Western Mediterranean ancestry. Was this a mix with a Western European Megalithic population, perhaps mainly females belonging to mtDNA haplogroup H...perhaps they themselves the daughters of I2a males? Is genetic drift hinting at a P312+ Rhenish Bell Beaker or Iberian Bell Beaker re-expansion...or perhaps both? IMO, a clean and uniform east to west movement and growth is not backed by Haak's genetic drift nor admixture tests. Certainly both leading Bell Beaker expansion models (Iberian/Dutch) show expansions from Western Europe, not from the Danube. Of course we will only know for sure once we get ancient DNA from places like the Netherlands (Protruding Foot Beaker Culture), France (Seine-Oise-Marne culture), Italy (Remedello Culture) and across different periods of time, but we need to keep all options open, because one thing that the predominantly Z2103+ Haak results showed us, is that ancient DNA is full of surprises.

razyn
04-03-2015, 04:42 PM
Certainly both leading Bell Beaker expansion models (Iberian/Dutch) show expansions from Western Europe, not from the Danube.
But there are other ways to get from the steppe to the Rhine, or Iberia, than via the Mediterranean and/or Danube.

R.Rocca
04-03-2015, 04:47 PM
But there are other ways to get from the steppe to the Rhine, or Iberia, than via the Mediterranean and/or Danube.

True, but none of them are overwhelmingly backed by archaeology.

rms2
04-03-2015, 09:17 PM
Bell Beaker is an enigma. It's a culture that is supposed to have begun in Portugal and expanded east, but not in full-blown Beaker form. In full flower it looks so kurgan-like that Gimbutas described the Bell Beaker people as eastern European "Kurgans" and referred to them as the "cousins" of the Corded Ware and Catacomb peoples. She derived Beaker from Vucedol and its immediate successors. Coon and Hubert noted that the Spanish Beaker Folk were not the same people physically as the eastern Beaker Folk.

To top it off, thus far the only Beaker y-dna results have all been R1bxU106, and one of those is a confirmed P312+. R1b in Europe has an east-to-west phylogeography.

So, what are we to make of all that and the presence of ANE and a substantial Yamnaya component in both Beaker and in most modern Europeans and descendants of Europeans?

I think Mike Walsh has suggested that perhaps Beaker began with a y-dna profile in which there was no R1b. It moved east and encountered an R1b element in central or eastern Europe which ultimately became the predominant y haplogroup in Beaker and completed the full-blown Beaker package.

Jean M has the Stelae People come from the Pontic-Caspian steppe to Iberia, bringing their anthropomorphic funeral stelae with them and perhaps some P312, as well. They blazed an avenue of communication and trade across Europe from the east to the west that would see steppe reinforcements travel to Iberia and Beaker Folk move eastward as far as Hungary.

I'm not sure who is right.

I suspect the following mix from Figure S9.8b quoted by Rich above is pretty close to correct:

49.5% Spain_MN + 50.5% Yamnaya

Somehow some elements of Beaker came out of Iberia and mixed with people who were derived from Yamnaya steppe pastoralists, perhaps Gimbutas' Vucedol and Vucedol successor people. There in central Europe a hybrid, kurganized culture was born: full-blown, classic Beaker. And that is where Beaker acquired its P312, too, if not from Jean M's Stelae People, who were themselves of steppe origin.

rms2
04-04-2015, 02:46 PM
Interesting to note that the greatest amount of genetic drift shared with Beaker comes from KO1 (HungaryGamba_HG) from the Körös River region in eastern Hungary. That's a hotspot for steppe pastoralist tumulus burials and spin-off cultures and the great mixing bowl where it is likely that Beaker acquired its steppe pastoralist y-dna. Maybe it's not surprising that the WHG element in Beaker resembles KO1 more closely than any other, since that was the region in which the eastern Beaker mix took place.

jeanL
04-04-2015, 04:21 PM
Interesting to note that the greatest amount of genetic drift shared with Beaker comes from KO1 (HungaryGamba_HG) from the Körös River region in eastern Hungary. That's a hotspot for steppe pastoralist tumulus burials and spin-off cultures and the great mixing bowl where it is likely that Beaker acquired its steppe pastoralist y-dna. Maybe it's not surprising that the WHG element in Beaker resembles KO1 more closely than any other, since that was the region in which the eastern Beaker mix took place.

Wouldn't that imply that there was a population in that was a relic WHG and extremely similar to KO1, but we know that the population with which the pastoralist likely mix was possibly similar to Hungary EN and CO, neither of which is in the top list of drift sharing with Beaker.

Augustus
04-04-2015, 11:44 PM
So, what are we to make of all that and the presence of ANE and a substantial Yamnaya component in both Beaker and in most modern Europeans and descendants of Europeans?
.

Quite a few of the Haak et. Al samples in Central Europe had high Yamna without R1. We have examples of Yamna autosomal component penetrating into a population without a mark on Y-Dna. Even Otzi had some very slight Samara component.

Coldmountains
04-05-2015, 12:21 AM
Quite a few of the Haak et. Al samples in Central Europe had high Yamna without R1. We have examples of Yamna autosomal component penetrating into a population without a mark on Y-Dna. Even Otzi had some very slight Samara component.

Something is very odd about his Yamna admixture results. Norwegians having more Yamna ancestry than eastern Europeans sounds unrealistic and his ANE estimates look simply wrong (not totally but not accurate for sure). I am sure that Norwegians indeed have much Yamna "related" ancestry (over 40%) but not more than Ukrainians, Poles or Belarusians. The ANE admixture among them is higher for sure and Norwegians scored only higher ANE admixture because of using MA-1 as reference, which creates noise. Some Ma-1 samples like from Human origin look contaminated and Eurogenes results for ANE were a bit different when he used illumina datasets and other methods. I think that Eurogenes results for ANE admixture are the most accurate till yet.
https://docs.google.com/spreadsheets/d/1x8pm8sVcHqceiNFJMO082kxaBF5ePr4__bAK05VQRFw/edit?pli=1
https://docs.google.com/spreadsheets/d/1v4zYizoWtsoW1MNBN7SUrLf8R62NHPbMRySUJ2J48_Q/edit?pli=1#gid=1410860471
Otzi having some Samara component is really strange and maybe there is something which makes the admixture results inaccurate for them and modern Europeans in Haak et al.

rms2
04-05-2015, 11:37 AM
Quite a few of the Haak et. Al samples in Central Europe had high Yamna without R1. We have examples of Yamna autosomal component penetrating into a population without a mark on Y-Dna. Even Otzi had some very slight Samara component.

Since all of the male Yamnaya samples upon which the Yamnaya component is based were R1, and the three samples upon which the eastward genomic shift of Europe's population is based are R*, R1a, and R1b, it would be pretty difficult to justify an attempt to separate the Yamnaya component from the expansion of R1 westward.

Is that what you were attempting? Do you think the Yamnaya component was connected to the spread of some other y haplogroup?

rms2
04-05-2015, 11:45 AM
Wouldn't that imply that there was a population in that was a relic WHG and extremely similar to KO1, but we know that the population with which the pastoralist likely mix was possibly similar to Hungary EN and CO, neither of which is in the top list of drift sharing with Beaker.

From what I can see (Figure 3 in Haak et al), the two samples to which you refer are largely Early Neolithic (LBK_EN) and, in the case of HungaryGamba_EN, have little Yamnaya and WHG, or, in the case of HungaryGamba_CA, no Yamnaya and only slightly more WHG than Beaker.

Net Down G5L
04-05-2015, 09:07 PM
True, but none of them are overwhelmingly backed by archaeology.


My view for what it is worth.....

I believe that there is plenty of evidence - if not 'overwhelming'.

Including the:
(1) anthropology evidence of Coon (e.g. longheads and roundhead mix within later Megalithic);
(2) mixed mother godess (Old Europe) and sun worship (Indo-European) culture of the late megalithic culture;
(3) literature describing the distinct waves of megalithic monument building;
(4) patterns of change in middle and late Neolithic cultures for cental Europe and Atlantic Europe; and
(5) emergence of cremation in Tisza-Polgar, Baden, some Megalithic monuments/locations, Isles etc.

The evidence strongly suggests to me that Indo-Europe (Gimbutas' wave 1 or 2? we could debate that extensively) merged with the megalithic I2 people in pre-Beaker times.

German Beaker could have emerged from this mixed IE -WHG "culture" in to the mixing bowl adjacent to corded ware.
Beaker was heading from Iberia / Netherlands to Germany / Central Europe with partner H - hence the mt trail of Brotherton 2013. Also Brandt's 2013,2014 conclusions.

The complication is how/when Beaker picked up its metalworking skills. I do not think that they were carried directly from the Steppe so there must have been interaction - probably in the Balkans / Alps - with metalworking specialists - possibly out of Anatolia. Whether the metalworker skills came from people with J or another type of DNA is an open question.

I believe we have to disentangle two/three waves out of the Steppe, waves of I2 megalithic that mixed with Indo-Europeans in its later stage, and finally the metalworking specialists - wherever they joined the party. I think the evidence is probably already available to pin this down in some detail - but there is a massive amount of literature to distil. Also to look at the existing DNA information again to test different hypotheses. As I have previously stated, for me the Haak data never supported a straight forward east-West movement of P312 with Yamnaya. More aDNA will make it much easier to get to the answer - and will eventually give us a pretty definitive solution.

I know most of you reject Coon's work. However, I suspect that studying his Cypriote Dinarics (for example) could be quite enlightening for all of us.

I also strongly suspect that a lot of material and discussion on M269, L11 etc that you, Richard, shared with this forum in 2012/13 - is still of particular relevance. It certainly started me thinking in a new way and looking for alternative solutions to my old 'Northern route out of the Steppe model'.

alan
04-06-2015, 01:44 AM
Its pretty clear beaker is a very complex phenomenon and not a simply movement from A to B. It clearly was migratory but it lacks the relative simplicity of corded ware. It could yet transpire that in pre-beaker times eastern genes went west and are indicated by the arrival of single burial. They then mixed with locals in Iberia, developing a hybrid we call early beaker culture before finally spreading back east again.

If that was what happened then c. 2800BC some group with steppe y lineages and perhaps autosomal mixes from other groups must have passed from central or Alpine Europe rapidly through France leaving little trace and on to Iberia where they should have added some Iberian late Neolithic autosomal and mtDNA. The beaker dates then track along the south of France.

One thing crucial about beaker that has never been really understood is beaker. after a century or two confined to Iberia and then southern France, appearing suddenly over large swathes of Europe c. 2500BC and exactly where this spread from. That still appears unclear to me. Around that date we suddenly have the central European beaker east group, the Rhenish group, the isles groups, Csepel etc. It seems totally obscure to me how that happened the nature of this, the directions of spread etc. Whatever the cause, something c. 2500BC made beaker very expansive.

Net Down G5L
04-06-2015, 05:40 AM
Its pretty clear beaker is a very complex phenomenon and not a simply movement from A to B. It clearly was migratory but it lacks the relative simplicity of corded ware. It could yet transpire that in pre-beaker times eastern genes went west and are indicated by the arrival of single burial. They then mixed with locals in Iberia, developing a hybrid we call early beaker culture before finally spreading back east again.

If that was what happened then c. 2800BC some group with steppe y lineages and perhaps autosomal mixes from other groups must have passed from central or Alpine Europe rapidly through France leaving little trace and on to Iberia where they should have added some Iberian late Neolithic autosomal and mtDNA. The beaker dates then track along the south of France.

One thing crucial about beaker that has never been really understood is beaker. after a century or two confined to Iberia and then southern France, appearing suddenly over large swathes of Europe c. 2500BC and exactly where this spread from. That still appears unclear to me. Around that date we suddenly have the central European beaker east group, the Rhenish group, the isles groups, Csepel etc. It seems totally obscure to me how that happened the nature of this, the directions of spread etc. Whatever the cause, something c. 2500BC made beaker very expansive.

I have hypothesised that a proto-Beaker network could have already been in place. For example, if R1b did move out of the Steppe with wave 1 - and moved in two routes - Danube to Baden and Cyprus/Mediterranean - West (also some areas between), then a family network would be there already for Beaker to exploit.
I do not think there are many aDNA samples in the right place at the right time to test this possibility. And then there is cremation - although the older writers seem to want to link that to Anatolia.

alan
04-06-2015, 09:43 AM
I have hypothesised that a proto-Beaker network could have already been in place. For example, if R1b did move out of the Steppe with wave 1 - and moved in two routes - Danube to Baden and Cyprus/Mediterranean - West (also some areas between), then a family network would be there already for Beaker to exploit.
I do not think there are many aDNA samples in the right place at the right time to test this possibility. And then there is cremation - although the older writers seem to want to link that to Anatolia.

I am still wary of use skull change. Across much of Europe, skulls apparently noticeable broadened in the high Medieval period for a while in the full light of history which provides no migratory explanation. They then reverted to the more usual longer shape again. The timing does not correspond with a cold period - it actually was a warm period which means climate doesnt explain it either. I am not saying it cannot ever be an indicator of movement but when there is a wide switch towards a rounder skull from NW Europe to SW Asia at the same time it sounds more to me like the mysterious Medieval pan-European shift. Note too the vital flaw in the idea of a link with Bronze Age round skulls and Cyprus etc - early SW European beaker skulls c.2800/2700-2500BC are not of the round type -which is apparently a central, Alpine and NW European beaker phenomenon of 2500BC onward and seems to commence at the period of the sudden mega-expansion of beaker across Europe c. 2500BC. So, I dont think the Coon inspired idea of a east Med origin of spread of the skull type into beaker stands up to scrutiny.

A much more interesting thing in the Alpine Italian Remedello II beaker type skulls of c. 2900-2600BC which are much closer and overlap in time with earlier beaker although also pre-date beaker altogether. In fact in the period c. 2800/2700-2500BC beaker lacked these skulls while Remedello II had them.

I have my doubts about skulls and migration but if you wanted to look for an explanation for their appearance in Italy, SW Asia and then in beaker etc I would tend to look towards the Balkans area.

alan
04-06-2015, 10:10 AM
I have hypothesised that a proto-Beaker network could have already been in place. For example, if R1b did move out of the Steppe with wave 1 - and moved in two routes - Danube to Baden and Cyprus/Mediterranean - West (also some areas between), then a family network would be there already for Beaker to exploit.
I do not think there are many aDNA samples in the right place at the right time to test this possibility. And then there is cremation - although the older writers seem to want to link that to Anatolia.

An important thing to remember is the plano-occipital head of beaker type also known in other cultures around this time is not really a round head as in a thickset round bullet head with a broad round face anyway. Its more like a long head and with long narrow dolichomorphic facial features but with the rear part flattened. The other trait is tallness.

There is a repeated formulaic phrase for the ideal face in early Irish literature 'broad above, narrow below' which suggests a very ancient standard of beauty or handsomeness held by the Celts which sounds a little beakerish to me. It is possible that such a standard goes back to some time in the copper or Bronze Age in another part of Europe.

alan
04-06-2015, 10:24 AM
Its pretty clear beaker is a very complex phenomenon and not a simply movement from A to B. It clearly was migratory but it lacks the relative simplicity of corded ware. It could yet transpire that in pre-beaker times eastern genes went west and are indicated by the arrival of single burial. They then mixed with locals in Iberia, developing a hybrid we call early beaker culture before finally spreading back east again.

If that was what happened then c. 2800BC some group with steppe y lineages and perhaps autosomal mixes from other groups must have passed from central or Alpine Europe rapidly through France leaving little trace and on to Iberia where they should have added some Iberian late Neolithic autosomal and mtDNA. The beaker dates then track along the south of France.

One thing crucial about beaker that has never been really understood is beaker. after a century or two confined to Iberia and then southern France, appearing suddenly over large swathes of Europe c. 2500BC and exactly where this spread from. That still appears unclear to me. Around that date we suddenly have the central European beaker east group, the Rhenish group, the isles groups, Csepel etc. It seems totally obscure to me how that happened the nature of this, the directions of spread etc. Whatever the cause, something c. 2500BC made beaker very expansive.

what puzzles me about the sudden expansion around 2500BC is how this fits in with the idea of the fall of the early beaker people at the hands of the central European beaker group around this time. It seems odd that they could be toppled by an other group of beaker people who had only barely come into existence...and what was the relationship between these groups when one had different type of skulls. Also we could ask where did these brachycephalic skulls appear from if the early beaker people didnt have them and neither did the corded ware people. I tend to think that the appearance of this skull type in beaker at the moment beaker suddenly moves from a couple of centuries as a culture limited to south-west Europe to a pan-European one cannot be a coincidence. Perhaps something caused both phenomenons. IMO its crucial to get yDNA and DNA in general from the early beaker phase pre-2500. There seem to have been beaker users in Iberia from at least the 2700s and in southern France by 2600BC. Were they the same peope or were some of the exotic and desirable parts of their culture taken up by other peoples in central Europe c. 2500BC through intermarriage. Only a study of beaker ancient DNA across the period from the earliest to after 2500BC will show this. Clearly it is now possible to do this if someone gets the funding.

Net Down G5L
04-06-2015, 03:07 PM
what puzzles me about the sudden expansion around 2500BC is how this fits in with the idea of the fall of the early beaker people at the hands of the central European beaker group around this time. It seems odd that they could be toppled by an other group of beaker people who had only barely come into existence...and what was the relationship between these groups when one had different type of skulls. Also we could ask where did these brachycephalic skulls appear from if the early beaker people didnt have them and neither did the corded ware people. I tend to think that the appearance of this skull type in beaker at the moment beaker suddenly moves from a couple of centuries as a culture limited to south-west Europe to a pan-European one cannot be a coincidence. Perhaps something caused both phenomenons. IMO its crucial to get yDNA and DNA in general from the early beaker phase pre-2500. There seem to have been beaker users in Iberia from at least the 2700s and in southern France by 2600BC. Were they the same peope or were some of the exotic and desirable parts of their culture taken up by other peoples in central Europe c. 2500BC through intermarriage. Only a study of beaker ancient DNA across the period from the earliest to after 2500BC will show this. Clearly it is now possible to do this if someone gets the funding.

Yes, agree re. need for maritime Beaker samples for aDNA. I have already trawled the Dorset archives for likely samples but as Dorset was in the area of eastern/mixed finding a clearly maritime sample is difficult. I think S W Ireland may be the best bet (or Brittany / Iberia...but sourcing samples??). Then institutions will not release material without a well worked up project plan that is fully resourced. That really means an academic institution project - unless the citizen scientist community formed a consortium bid and applied for funding and bought in lab time. Very difficult to achieve.

Re early Beaker being "longhead". Two possibilities that occur to me are:

1. Beaker was a mixed group of, for example, J2 and Rib L11/P312. A J2/I2 dominated group headed to Iberia. Then later when heading east they became dominated by R1b. Seems a little complex to me though.
2. L11/P312 were longheads. When Beaker moved inland towards central Europe it moved via the balkans and picked up "Dinaric" genetic mutation to become roundheaded.
However, I just do not think that fits with the archaeology evidence.

So I do not have an elegant solution to suggest.

I do think Coon's work is worth considering. His description of skull types and migrations fits so well with what we now know from DNA - when applied to the Neolithic farming migrations and early megalithic migrations. I see know reason why his later Neolithic migration observations should not be equally valid.
I think Cypriote Dinaric is potentially compatible with martitime Beaker being 'longheaded' because I see Cypriote Dinaric as an earlier occurance. If it hypothetically linked to a Gimbutas wave it would be wave 1. Beaker would be wave 2 at the earliest. Even if both were R1b (and we don't know if either are R1b) they could be very different lines and we know skull shapes change over time due to admixture.
My own view is that we need to consider all lines of evidence available to us. By all means weight their importance and reliability and value them accordingly.

R.Rocca
04-06-2015, 03:48 PM
Its pretty clear beaker is a very complex phenomenon and not a simply movement from A to B. It clearly was migratory but it lacks the relative simplicity of corded ware. It could yet transpire that in pre-beaker times eastern genes went west and are indicated by the arrival of single burial. They then mixed with locals in Iberia, developing a hybrid we call early beaker culture before finally spreading back east again.

In Iberia, I don't think there is any likelier scenario (than the one you've outlined) that would explain high P312+ there combined with the heavy amount of Bell Beaker material plus their collectively older age. A non-R1b early Iberian Bell Beaker presence followed by a full DF27 advance from central Europe into Iberia fails on my levels...


1. Almost non-existent traces of a Bell Beaker movement from Central Europe into Iberia to explain DF27 there

2. Unlike say L21, DF27 seems as old as P312 itself

3. While much is made of the differences between Iberian Bell Beaker and Central European Bell Beaker, they are much closer related to each other than say either of them are to Corded Ware


If that was what happened then c. 2800BC some group with steppe y lineages and perhaps autosomal mixes from other groups must have passed from central or Alpine Europe rapidly through France leaving little trace and on to Iberia where they should have added some Iberian late Neolithic autosomal and mtDNA. The beaker dates then track along the south of France.

I think the rarity of stelae in the Balkans and Central Europe shows just how rapid that movement was. One minute they are in places like the Ukraine, the next they are in the Alps, France and Iberia.


One thing crucial about beaker that has never been really understood is beaker. after a century or two confined to Iberia and then southern France, appearing suddenly over large swathes of Europe c. 2500BC and exactly where this spread from. That still appears unclear to me. Around that date we suddenly have the central European beaker east group, the Rhenish group, the isles groups, Csepel etc. It seems totally obscure to me how that happened the nature of this, the directions of spread etc. Whatever the cause, something c. 2500BC made beaker very expansive.

The migration path was likely so complex, it is nearly impossible to guess all of it correctly. As I've said previously, it could be that the diversity of subclades we see in Central Europe is due to a lot of smaller clans settling in fringe Corded Ware areas and never really making the treck to Iberia. It certainly looks like U106 never made the trek. Perhaps some types like L21 and DF19 and some specific sublades of U152 or DF27 didn't either?

Hando
04-06-2015, 04:30 PM
In Iberia, I don't think there is any likelier scenario that would explain high P312+ there combined with the heavy amount of Bell Beaker material plus their collectively older age. A non-R1b early Iberian Bell Beaker presence followed by a full DF27 advance from central Europe into Iberia fails on my levels...


1. Almost non-existent traces of a Bell Beaker movement from Central Europe into Iberia to explain DF27 there

2. Unlike say L21, DF27 seems as old as P312 itself

3. While much is made of the differences between Iberian Bell Beaker and Central European Bell Beaker, they are much closer related to each other than say either of them are to Corded Ware



I think the rarity of stelae in the Balkans and Central Europe shows just how rapid that movement was. One minute they are in places like the Ukraine, the next they are in the Alps, France and Iberia.



The migration path was likely so complex, it is nearly impossible to guess all of it correctly. As I've said previously, it could be that the diversity of subclades we see in Central Europe is due to a lot of smaller clans settling in fringe Corded Ware areas and never really making the treck to Iberia. It certainly looks like U106 never made the trek. Perhaps some types like L21 and DF19 and some specific sublades of U152 or DF27 didn't either?

Sorry, but are you agreeing or disagreeing with Alan? Do you mean that what made Beaker, first started in Iberia via a non R1b peoples, and then an R1b people from Central Europe (with ultimate origins in the Steppes) entered Iberia and mixed with these pre-Beaker Neolithic Iberian people and then migrated east as Beaker?

Jean M
04-06-2015, 04:40 PM
Whatever the cause, something c. 2500BC made beaker very expansive.

Catriona Gibson read an interesting paper (not yet published sadly) about the signs of trouble and strife in Portuguese BB at the time and the abandonment of BB walled settlements there, just as BB appears elsewhere in Europe. It's pretty suggestive, though it is not clear who or what was causing trouble for our bell-beaker makers.

The Eastern BB Group clearly started with Csepel. The Atlantic route goes to Brittany and the Isles. No real difficulty there.

Jean M
04-06-2015, 04:55 PM
I think the rarity of stelae in the Balkans and Central Europe shows just how rapid that movement was. One minute they are in places like the Ukraine, the next they are in the Alps, France and Iberia.

Actually the more I dig into the literature the more stelae I find. One trail clearly goes from the Balkans to Greece (which I relate to Proto-Greek). Another goes up the Danube to the Carpathian Basin.

4275

In my view, initially only a few prospectors peeled away from the mother group in the Carpathian Basin to arrive in the Alps, and move on to Italy and Iberia. They blazed the trail that others could travel for centuries, to and fro. The Carpathian Basin was not abandoned.


As I've said previously, it could be that the diversity of subclades we see in Central Europe is due to a lot of smaller clans settling in fringe Corded Ware areas and never really making the trek to Iberia. It certainly looks like U106 never made the trek.

While I agree entirely that the mother group of P312 did not move en masse to Iberia, there seems no particular reason for U106 to be in the Carpathian Basin in the first place. It is more likely to have travelled east and north of the Carpathians.

Net Down G5L
04-06-2015, 05:03 PM
[QUOTE=Richard A. Rocca;77913]In Iberia, I don't think there is any likelier scenario that would explain high P312+ there combined with the heavy amount of Bell Beaker material plus their collectively older age. A non-R1b early Iberian Bell Beaker presence followed by a full DF27 advance from central Europe into Iberia fails on my levels..[/QUOTE

What is known about the location of the earliest branches of DF27? Also diversity? (I know some do not think diversity important but I am still interested to know).

I am afraid I have no particular knowledge of these facts.

Is there a suggestion of earliest branches occuring in Central Europe, Iberia, Isles...??

rms2
04-06-2015, 06:09 PM
Here's something that I wonder about that was posted by Piquerobi in another thread:



Decorated horse phalanges have also been reported from Bell Beaker sites in Spain (Maier 1961; Piggott 1983). They are perhaps the strongest cultural marker for the Botai, and show a connection with the Tersek, a contemporaneous Copper Age culture in the Turgay region to the west (Kalieva et al. 1989).
Source: Olsen, S. (2003). "The Exploitation of Horses at Botai, Kazakhstan". In Levine, Marsha; Renfrew, Colin; Boyle, Katie. Prehistoric Steppe Adaptation and the Horse. Cambridge: McDonald Institute.

Piquerobi was quoting something that was posted at some other forum by some unknown person. I wonder which works the "Maier 1961" and "Piggott 1983" references are pointing to; it would be nice to know. Maybe they are cited in the source article by Olsen, S.?

http://www.anthrogenica.com/showthread.php?3474-Bell-Beakers-Gimbutas-and-R1b&p=59628&viewfull=1#post59628

Jean M
04-06-2015, 06:27 PM
Piquerobi was quoting something that was posted at some other forum by some unknown person. I wonder which works the "Maier 1961" and "Piggott 1983" references are pointing to; it would be nice to know. Maybe they are cited in the source article by Olsen, S.?

You have access to Levine, Renfrew and Boyle 2003 in Mini-Library > Books > Steppe. That quote is from p. 94. But I'll do the donkey work.

Maier, R.A., 1961. Neolithische Tierknochen-Idole un Tierknochen-Anhänger Europas. Bericht der Römisch-Germanischen Kommission 42, 171–305.
Piggott, S., 1983. The Earliest Wheeled Transport. Ithaca (NY): Cornell University Press.

rms2
04-06-2015, 07:10 PM
Thanks. I tracked down something relevant to the decorated horse phalanges issue.

Chapter Four of the book, Are All Warriors Male?: Gender Roles on the Ancient Eurasian Steppe, edited by Katheryn M. Linduff and Karen Sydney Rubinson, is entitled "Women's Attire and Possible Sacred Role in 4th Millennium Northern Kazakhstan", and was written by Sandra Olsen and Deborah G. Harding. This is from page 73:



Decorated horse phalanges have also been identified at Neolithic and Bell Beaker sites in Spain (Almagro Basch 1966; Ayala Juan 1985; Cardoso 1995). The Spanish ones are more simply and abstractly decorated, primarily by series of dimples drilled into the front (volar) surface.

I'm not sure it's all that meaningful, however, since that chapter describes decorated animal phalanges from a number of different peoples and cultures. From what I could see, not all peoples used horse phalanges; Beaker, Botai, and Tersek peoples did, and there was one example cited from the Epipaleolithic site of Cuina Turcului in Romania in the Iron Gate region of the Danube.

vettor
04-06-2015, 07:30 PM
This supplementary paper released a fortnight ago by Anna Szécsényi-Nagy might help on the genetic drift ......some interesting maps as well

http://rspb.royalsocietypublishing.org/content/royprsb/suppl/2015/03/20/rspb.2015.0339.DC1/rspb20150339supp1.pdf

newtoboard
04-06-2015, 11:04 PM
Actually the more I dig into the literature the more stelae I find. One trail clearly goes from the Balkans to Greece (which I relate to Proto-Greek). Another goes up the Danube to the Carpathian Basin.

4275

In my view, initially only a few prospectors peeled away from the mother group in the Carpathian Basin to arrive in the Alps, and move on to Italy and Iberia. They blazed the trail that others could travel for centuries, to and fro. The Carpathian Basin was not abandoned.



While I agree entirely that the mother group of P312 did not move en masse to Iberia, there seems no particular reason for U106 to be in the Carpathian Basin in the first place. It is more likely to have travelled east and north of the Carpathians.

Where would the migration of Proto Greek start and when?

And would U106 in your scenario be found in Central (Polish?) Corded Ware?

Jean M
04-06-2015, 11:33 PM
Where would the migration of Proto Greek start and when?

And would U106 in your scenario be found in Central (Polish?) Corded Ware?

Let's not drift off the subject.

Anglecynn
04-06-2015, 11:50 PM
Let's not drift off the subject.

Lol, har har.

alan
04-07-2015, 03:16 AM
It damned unlucky for P312 folks that their yDNA has to be linked to beaker - one of the most complex and baffling archaeological phenomenons. Its a a situation where archaeology is not going to help the geneticists but more a case the geneticists will help the archaeologists to crack exactly what beaker was. Its not even as simple as establishing what DNA was in what areas and what beaker phases. We also need to establish in what cultures P312 and before it L11 and L51 were located in pre-beaker times. Only then will we properly understand.

My personal view now is that beaker was a very fast local response to something coming in from central Europe c. 2800BC or a little later. I have completely abandoned the idea that the spread of copper across the Med and southern Alps corresponds with the spread of R1b. IMO R1b is a subsequent spread. We have ancient yDNA, admittedly a tiny sample across a big aea. from the period c.3500-3000BC from Italy, southern France and Iberia and none is R1b.I think its best to run with the hard evidence that the spread of copper working across Med. and south Alpine Europe is not linked to R1b.

I suppose in Italian terms that means I am saying I dont believe R1b was connected to Remedello I and the early copper mines but I am prepared to believe it might well have arrived in Remedello II. In Iberian terms I dont believe it is linked to the Zambujal type groups c. 3100BC and now prefer the idea that it arrived right at the inception of Beaker in the 2700s and indeed beaker culture was a very rapid local outcome of an influx of single burial users.So I am guessing now that P312 is only fractionally pre-beaker in Iberia. So, I dont think P312 or L11 will be found pre-2900BC anywhere east of eastern Europe/the Balkans. If bell beaker is completely unconnected to Corded Ware then the best bet seem steppe to Balkans, then balkans to north Italy/SE France c.2900BC then Iberia in the 2700s.

Anyone spreading at that time and place was not offering the miracle of metal knowledge as that had preceded them some centuries earlier. IMO its far more likely that groups moving at that time were primarily offering mobility and the ability to establish a network that would provide an outlet for Iberian metals back into the central Med. and then central and northern Europe. It is interesting that around this time RC dates suggest the tailing off of the older mines in northern Italy so that could have been a push factor. I found interesting a paper that said pre-beaker use of copper in Iberia was mundane and non-prestigious because it was too common and easy to get hold of in Iberia. Its value only really came out if it could be shifted out of Iberia to areas less blessed in copper. It seems to me there was mutual benefit and push and pull factors for a mobile central European group to provide an outlet network for older Zambujal copper workers.

If that sort of concept is near the mark then it is interesting that the earliest part of the network ie the earliest beaker dates outside Iberia seem to be SE France and adjacent Alpine areas in Switzerland and Italy by at least c. 2600BC, the western part of the area where Remedello symbolism is found and also near the area where the failing older Italian mines were located. Now as I said, the whole Med. and Alpine area already had copper working knowledge and skills so it was the creating of a new wider network and implied mobility that was new.

vettor
04-07-2015, 06:10 AM
This supplementary paper released a fortnight ago by Anna Szécsényi-Nagy might help on the genetic drift ......some interesting maps as well

http://rspb.royalsocietypublishing.org/content/royprsb/suppl/2015/03/20/rspb.2015.0339.DC1/rspb20150339supp1.pdf

With pages 26 and 27 , it seems the Neolithic farmers did not come from the fertile crescent, but from north of the Zargos mountains, mostly Caucasus and north-Caucasus,........they must have followed the north-Anatolian coast and entered the Balkans from there.
They are Caucasian farmers and not levantine farmers.

Since women are usually the pot makers ( as I read elsewhere , I do not know why ), we should look at the mtdna of these women in the bell Beaker area.
Maybe as BritainDNA states............mtdna N1a1a..the pot makers is a clue

Net Down G5L
04-07-2015, 06:33 AM
[QUOTE=Richard A. Rocca;77913]In Iberia, I don't think there is any likelier scenario that would explain high P312+ there combined with the heavy amount of Bell Beaker material plus their collectively older age. A non-R1b early Iberian Bell Beaker presence followed by a full DF27 advance from central Europe into Iberia fails on my levels..[/QUOTE

What is known about the location of the earliest branches of DF27? Also diversity? (I know some do not think diversity important but I am still interested to know).

I am afraid I have no particular knowledge of these facts.

Is there a suggestion of earliest branches occuring in Central Europe, Iberia, Isles...??

OK - no takers.

So, my limited knowledge suggests it is possible DF27 was 'born' on the Atlantic, for example Iberia, and it could have arrived in the Isles with Maritime Bell Beaker and refluxed back in the Bronze Age re. the north-South split of str data. Now, I am sure there is much more upto date snp analysis that can support or disprove such ideas.
In that Maritime Bell Beaker could be P312, DF27 or Z2103 migrating with Stelae, J2 from metalworking origins......and possibly many more. Can we rule out any of these options. How likely is the DF27 option?

rms2
04-07-2015, 11:15 AM
In Busby, P312xL21,U152 in the Isles, which is probably mostly DF27, reaches its highest frequency in SE England (~21%). It seems to fade pretty drastically as one moves west and north, with the exception of the North Wales sample location, where P312xL21,U152 is at 17.5%. I suspect that stems from later English settlement, however, since the frequencies of U106 and U152 are also oddly elevated there (9.2% and 7.5% respectively - relatively high for Wales). In Ireland P312xL21,U152 is pretty consistent at around 7.5%, which is not too bad.

Personally, I would hesitate to connect DF27 to Maritime Beakers just because there is a lot of it in Iberia now. Its distribution in Britain looks like the product of a relatively late arrival to me, perhaps, as may be the case with U152, with the Belgae and, later, with the Romans. In Ireland it could represent refugees from the Roman conquest of Gaul.

moesan
04-07-2015, 11:30 AM
I am still wary of use skull change. Across much of Europe, skulls apparently noticeable broadened in the high Medieval period for a while in the full light of history which provides no migratory explanation. They then reverted to the more usual longer shape again. The timing does not correspond with a cold period - it actually was a warm period which means climate doesnt explain it either. I am not saying it cannot ever be an indicator of movement but when there is a wide switch towards a rounder skull from NW Europe to SW Asia at the same time it sounds more to me like the mysterious Medieval pan-European shift. Note too the vital flaw in the idea of a link with Bronze Age round skulls and Cyprus etc - early SW European beaker skulls c.2800/2700-2500BC are not of the round type -which is apparently a central, Alpine and NW European beaker phenomenon of 2500BC onward and seems to commence at the period of the sudden mega-expansion of beaker across Europe c. 2500BC. So, I dont think the Coon inspired idea of a east Med origin of spread of the skull type into beaker stands up to scrutiny.

A much more interesting thing in the Alpine Italian Remedello II beaker type skulls of c. 2900-2600BC which are much closer and overlap in time with earlier beaker although also pre-date beaker altogether. In fact in the period c. 2800/2700-2500BC beaker lacked these skulls while Remedello II had them.

I have my doubts about skulls and migration but if you wanted to look for an explanation for their appearance in Italy, SW Asia and then in beaker etc I would tend to look towards the Balkans area.


do'nt confuse a global phenomenon of brachycephalization (and today debrachycephalization) modifying the average indexes of 3/4 points (say 80 to 83/84 or 72 to 75.76, and the genetic individual variations from 67 to 100 (yes!) or collective variations from 72 to 89, in the same continent at the same time within not too different cultures!
the BB of the Isles was a mix were AT LEAST 1 dolicho type and 2 well differentiated brachy types were found, differing not only by cephalic index but by facial index, thickness of cranial bones, orbits, nasal shape and index, and other typical shapes : too often I read this confusion between some external mesological / cultural relatively light incidences on skeletons and true genetic facts to deny phenotypes accuracy and historic demic moves as do someones ... (I don't accuse you here)
interesting thread by the way - the Y-I2a connexion is something to be worked on, and not only I2a1...

R.Rocca
04-07-2015, 12:06 PM
In Busby, P312xL21,U152 in the Isles, which is probably mostly DF27, reaches its highest frequency in SE England (~21%). It seems to fade pretty drastically as one moves west and north, with the exception of the North Wales sample location, where P312xL21,U152 is at 17.5%. I suspect that stems from later English settlement, however, since the frequencies of U106 and U152 are also oddly elevated there (9.2% and 7.5% respectively - relatively high for Wales). In Ireland P312xL21,U152 is pretty consistent at around 7.5%, which is not too bad.

Personally, I would hesitate to connect DF27 to Maritime Beakers just because there is a lot of it in Iberia now. Its distribution in Britain looks like the product of a relatively late arrival to me, perhaps, as may be the case with U152, with the Belgae and, later, with the Romans. In Ireland it could represent refugees from the Roman conquest of Gaul.

A pretty significant portion of P312xL21,U152 in SE England may be DF19+, with DF27 likely making up much more of the percentages of P312xL21,U152 in SW England. So, I think the link between DF27 and Maritime Beaker is not something we should rule out quite yet.

rms2
04-07-2015, 12:12 PM
A pretty significant portion of P312xL21,U152 in SE England may be DF19+, with DF27 likely making up much more of the percentages of P312xL21,U152 in SW England. So, I think the link between DF27 and Maritime Beaker is not something we should rule out quite yet.

That's a good point. Do we know of many DF19+ results from SE England or from people with ancestry in SE England?

There does seem to be a fairly solid block of DF27+ folks in SW Ireland connected to a certain set of surnames (I forget which ones - someone here will probably pipe up and supply them). Is there much evidence of Maritime Beaker in Ireland? I thought the Beaker in Ireland mostly came over from Britain and was of Rhenish derivation, but I could be wrong on that.

R.Rocca
04-07-2015, 12:30 PM
That's a good point. Do we know of many DF19+ results from SE England or from people with ancestry in SE England?


Yes, there are plenty...and I at some point estimated that a large portion of the Barbant study's P312xZ195 could be DF19, but I forget the specifics and can't find the details now. DF19 has quite a nice representation in the Low Countries as per FTDNA maps.



There does seem to be a fairly solid block of DF27+ folks in SW Ireland connected to a certain set of surnames (I forget which ones - someone here will probably pipe up and supply them). Is there much evidence of Maritime Beaker in Ireland? I thought the Beaker in Ireland mostly came over from Britain and was of Rhenish derivation, but I could be wrong on that.

From what I've read, Irish Bell Beakers are both Iberian/NW French and British derived, with the latter of course being primarily Rhenish derived.

R.Rocca
04-07-2015, 12:38 PM
It damned unlucky for P312 folks that their yDNA has to be linked to beaker


I think it is quite the contrary and it is good to having a rough starting age for an SNP.



My personal view now is that beaker was a very fast local response to something coming in from central Europe c. 2800BC or a little later. I have completely abandoned the idea that the spread of copper across the Med and southern Alps corresponds with the spread of R1b. IMO R1b is a subsequent spread. We have ancient yDNA, admittedly a tiny sample across a big aea. from the period c.3500-3000BC from Italy, southern France and Iberia and none is R1b.I think its best to run with the hard evidence that the spread of copper working across Med. and south Alpine Europe is not linked to R1b.


I think the same thing, it looks like one very fast movement west within a matter of a few generations, perhaps even driven by one clan.



I suppose in Italian terms that means I am saying I dont believe R1b was connected to Remedello I and the early copper mines but I am prepared to believe it might well have arrived in Remedello II. In Iberian terms I dont believe it is linked to the Zambujal type groups c. 3100BC and now prefer the idea that it arrived right at the inception of Beaker in the 2700s and indeed beaker culture was a very rapid local outcome of an influx of single burial users.So I am guessing now that P312 is only fractionally pre-beaker in Iberia. So, I dont think P312 or L11 will be found pre-2900BC anywhere east of eastern Europe/the Balkans. If bell beaker is completely unconnected to Corded Ware then the best bet seem steppe to Balkans, then balkans to north Italy/SE France c.2900BC then Iberia in the 2700s.


I'm still up in the air about it, but if Remedello I is not R1b, then neither were the stelae people (or at least does not look compatible with), since the earlier stelae in Italy have what are clearly Remedello I daggers carved into them. These are very different from the later Italian stelae that have Bell Beaker daggers, which in in Iberia and elsewhere, resemble Yamnaya ones.



Anyone spreading at that time and place was not offering the miracle of metal knowledge as that had preceded them some centuries earlier. IMO its far more likely that groups moving at that time were primarily offering mobility and the ability to establish a network that would provide an outlet for Iberian metals back into the central Med. and then central and northern Europe. It is interesting that around this time RC dates suggest the tailing off of the older mines in northern Italy so that could have been a push factor. I found interesting a paper that said pre-beaker use of copper in Iberia was mundane and non-prestigious because it was too common and easy to get hold of in Iberia. Its value only really came out if it could be shifted out of Iberia to areas less blessed in copper. It seems to me there was mutual benefit and push and pull factors for a mobile central European group to provide an outlet network for older Zambujal copper workers.

If that sort of concept is near the mark then it is interesting that the earliest part of the network ie the earliest beaker dates outside Iberia seem to be SE France and adjacent Alpine areas in Switzerland and Italy by at least c. 2600BC, the western part of the area where Remedello symbolism is found and also near the area where the failing older Italian mines were located. Now as I said, the whole Med. and Alpine area already had copper working knowledge and skills so it was the creating of a new wider network and implied mobility that was new.

By mobility, I don't think it was just the wheel, but the wheel plus the horse.

Agamemnon
04-07-2015, 01:20 PM
Richard, where do you think U152 took root in the BB horizon? I tend to think that Central-Eastern Beaker and Csepel might've been mainly U152, but that's just me.

Webb
04-07-2015, 03:13 PM
That's a good point. Do we know of many DF19+ results from SE England or from people with ancestry in SE England?

There does seem to be a fairly solid block of DF27+ folks in SW Ireland connected to a certain set of surnames (I forget which ones - someone here will probably pipe up and supply them). Is there much evidence of Maritime Beaker in Ireland? I thought the Beaker in Ireland mostly came over from Britain and was of Rhenish derivation, but I could be wrong on that.

There is a DF17 grouping with native Irish surnames. They all share a cluster of downstream markers. Durkin, Mulvihill, and Meehan. McDonald is a Scot who is in this cluster. Upstream from them is Menge, which is German, and upstream from them is an Italian and another Scot. They are all positive for Y3470, which according to YFull has a TMRCA of 2300 ybp. So DF17 is formed 4300ybp, TMRCA 3500ybp and this goes down to Y3470 which was formed 3500ybp and TMRCA of 2300ybp. I do not know the age of this Irish cluster, so I do not know how far out the four families are related, however, their STR signatures are are not closely related.

Webb
04-07-2015, 03:34 PM
Another example is myself, CTS4065. I am just upstream of a Swede and a Dutchman. There is a Dane, waiting on results who is probably in this cluster as well. He and the other two are more closely related than I am, as they are just downstream from me. So CTS4065 was formed 3800ybp, and a TMRCA of 3200ybp. This drops to our cluster, Y13211. This was formed 3200ybp, and a TMRCA of 1100ybp. Van Den Vliet, the Duchman, came up with a TMRCA of 900ybp for the seperation of my line from theirs. So, my line seperated and ended up in Britain at around 900AD to 1100AD. We all can guess by these dates how my line most likely ended up in England. I will also supply the Welsh cluster, which quite possibly puts a DF27, Z195- cluster in Wales at around 0BC. I think at this point DF27 should be a cautionary tale. I say this because there is no general pattern to it. Some of the various subclades could be quite old in Britain and some could be quite young. It could have arrived as early as Maritime Beaker, and as late as Viking incursions, and even more recent than that.

moesan
04-07-2015, 09:01 PM
DF 27 seems a relatively recent subgroup of Y-R1b -P312 - the dates cited here don't contradict this and are all ot them later than the 2000 so, even if we attach BBs to R1b, what I do not, I don't see why specially to link this SNP to BBs -
It's true, even if Y-R1b is not linked to early BBs, it could have been associated to them in further developments - that said DF27 seems very more linked to Aquitanians or Basques as a whole than to southern Portugal - it doesn't reflect the BBs poles of spreading in Iberia nor elsewhere- As I see the Basque history closely tied to the western I-Eans in West Europe (I-Eans or indo-europeanized?), the Celts and close tribes for the most, I think the presence more northernly than in Iberia is not so amazing and is not by force linked to late BBs colonization in the Isles - Basque country shows us a not neglictible overlap of Y-R1b-L21 and DF27, and I think the most of the recent SNPs of R1b got birth in West-Central and Northwestern Europe, independant from older subclades of R1b (the most of the L23 bearers) found in Mediterranea - that said as always the most recent downstreams clades of DF27 could have reached far contries at more individual levels, more recently... my bets for the moment -

razyn
04-07-2015, 10:38 PM
that said as always the most recent downstreams clades of DF27 could have reached far contries at more individual levels, more recently... my bets for the moment -

My bet is that you have that almost 100% backward, geographically and chronologically. We shall see. Or -- admitting that I am kind of old -- one of these days, the truth about it will out. Speculation from insufficient evidence won't resolve it.

lgmayka
04-07-2015, 11:13 PM
DF 27 seems a relatively recent subgroup of Y-R1b -P312
Y-DNA evidence suggests otherwise. YFull dates DF27, U152, and L21 (http://yfull.com/tree/R-P312/) at roughly the same age, 4700 ybp. L21 is almost certainly a little younger than the other two, because L21 has 6 tree-equivalent defining SNPs whereas the other two major clades have only one defining SNP each.

Isidro
04-08-2015, 10:44 AM
Y-DNA evidence suggests otherwise. YFull dates DF27, U152, and L21 (http://yfull.com/tree/R-P312/) at roughly the same age, 4700 ybp. L21 is almost certainly a little younger than the other two, because L21 has 6 tree-equivalent defining SNPs whereas the other two major clades have only one defining SNP each.

And that is why I have a problem equating L21 and U152 with DF27.
What are the odds of 2 mega founders having the same age L21 and U152?.

How about 3 mega founders having the same age?.

I guess anyone can argue just about anything, but seriously I think it is time to consider how far fetched that sounds.

rms2
04-08-2015, 11:32 AM
Age estimates based on SNP counting aren't exact; they're approximate, with margins of error, so one should not think there could not be some gaps between the birth of one founder and another or between all three. I doubt they were large gaps, certainly not gaps of thousands of years, but gaps of one hundred to a few hundred years are possible.

The fact that the founders of DF27, L21, and U152 are about the same age indicates that P312 expanded rapidly and that the time in which those three major clade-defining SNPs were born must have been propitious for the survival of male offspring in sufficient numbers to insure their lines continued into the future. So, there must have been something relatively good going on for P312 around 2700 BC or so.

Perhaps it was the expansion of the Indo-European tribes into new territory in peninsular Europe (Europe outside Russia) that we know was taking place early in the 3rd millennium BC.

R.Rocca
04-08-2015, 11:59 AM
And that is why I have a problem equating L21 and U152 with DF27.
What are the odds of 2 mega founders having the same age L21 and U152?.

How about 3 mega founders having the same age?.

I guess anyone can argue just about anything, but seriously I think it is time to consider how far fetched that sounds.

Far fetched? They all have the same STR modal and Bell Beaker populated almost all corners of Western Europe. I think it's far fetched to not think they weren't closely related within 300-400 years of each other.

Jean M
04-08-2015, 12:28 PM
I think at this point DF27 should be a cautionary tale. I say this because there is no general pattern to it. Some of the various subclades could be quite old in Britain and some could be quite young. It could have arrived as early as Maritime Beaker, and as late as Viking incursions, and even more recent than that.

That's pretty much what I say in the forthcoming Blood of the Celts:


R1b1a2a1a2a (DF27) is common in Iberia. So the rare cases of R1b-DF27* (the basal form of DF27) in Ireland today may be a remnant of Bell Beaker movements up the Atlantic. If R1b-DF27 also travelled with Bell Beaker from Iberia into the Carpathian Basin, that would explain its present wide range. It has some subclades whose bearers cluster in south-western Europe, but others which are almost exclusive to northern Europeans and their descendants. Scandinavian branches of R1b-DF27 could have arrived in the Isles as late as Viking or Norman times.

Webb
04-08-2015, 01:02 PM
That's pretty much what I say in the forthcoming Blood of the Celts:

I am touched. Weepy actually. JeanM, you have made my day. :)

Jean M
04-08-2015, 01:11 PM
I am touched. Weepy actually. JeanM, you have made my day. :)

Not news, surely? I gave everyone the chance to comment on the DF27 section. http://www.anthrogenica.com/showthread.php?3218-Draft-DF27-section-in-Celts-book . The arguments went on for pages. ;)

[Added] I did take out the specifics 'Viking or Norman' at one point: http://www.anthrogenica.com/showthread.php?3218-Draft-DF27-section-in-Celts-book&p=57309&viewfull=1#post57309

But then I put them back in again at the editing stage. I think we have enough data to make that a reasonable guess.

ArmandoR1b
04-08-2015, 02:50 PM
Y-DNA evidence suggests otherwise. YFull dates DF27, U152, and L21 (http://yfull.com/tree/R-P312/) at roughly the same age, 4700 ybp. L21 is almost certainly a little younger than the other two, because L21 has 6 tree-equivalent defining SNPs whereas the other two major clades have only one defining SNP each.
One of the phylogenetic equivalent SNPs to L21 should have an age close to DF27 and U152 and the rest of the phylogenetic equivalents should be younger.

ArmandoR1b
04-08-2015, 03:03 PM
And that is why I have a problem equating L21 and U152 with DF27.
What are the odds of 2 mega founders having the same age L21 and U152?.

How about 3 mega founders having the same age?.

I guess anyone can argue just about anything, but seriously I think it is time to consider how far fetched that sounds.

SNPs appear every 3-4 generations on average. It is not unreasonable or far-fetched that a person during a population explosion had a lot of great-great-grandchildren and they went on to have a lot of descendants whose lineages lived on to this day.

razyn
04-08-2015, 03:30 PM
One of the phylogenetic equivalent SNPs to L21 should have an age close to DF27 and U152 and the rest of the phylogenetic equivalents should be younger.

That's true if equivalent SNPs only occur singly and sequentially. That's the hypothesis behind the currently fashionable estimation of age of a clade by counting SNPs. If that hypothesis is incorrect, and multiple mutations can actually occur in an individual organism at once (yet be stable in descendants of that organism), then the numerous equivalent SNPs one finds in all of L21 -- and at many other phylogenetic branching points -- need not be younger than the one chosen by modern researchers to represent the parent.

TigerMW
04-08-2015, 04:09 PM
Far fetched? They all have the same STR modal and Bell Beaker populated almost all corners of Western Europe. I think it's far fetched to not think they weren't closely related within 300-400 years of each other.
I agree, the P312 early branches are closely related. I would never say this was literally a father-son type relationship but a time span of 300-400 years is quite possible. I think we have to keep in mind that U106's likely ancestral haplotype is very much right on target too with the P312 family. Descendants of some early branches of U106, such as Z18, are not so easy to distinguish from P312.

It's really the P311 family and they appear to stretch beyond Bell Beaker types. This is why I'm persistently interested in understanding U106. It's the cousin that branched away early so it might be a very instructive indicator.

SNP branch length-wise, in what order did the P311 family peel away; first U106 and DF100, then DF27 and U152, and then L21?

We should not confuse current population frequencies with ancient branching so the little brothers within P312: DF19, DF99, L238 and ZZ37 might also be helpful.

ArmandoR1b
04-08-2015, 04:55 PM
That's true if equivalent SNPs only occur singly and sequentially. That's the hypothesis behind the currently fashionable estimation of age of a clade by counting SNPs. If that hypothesis is incorrect, and multiple mutations can actually occur in an individual organism at once (yet be stable in descendants of that organism), then the numerous equivalent SNPs one finds in all of L21 -- and at many other phylogenetic branching points -- need not be younger than the one chosen by modern researchers to represent the parent.
I agree that there are likely times that multiple SNPs appear in a single generation but since we don't know the exact situation with the L21 clade I used the average situation http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3312576/ and used the word should instead of do and are.

Megalophias
04-08-2015, 05:17 PM
That's true if equivalent SNPs only occur singly and sequentially. That's the hypothesis behind the currently fashionable estimation of age of a clade by counting SNPs. If that hypothesis is incorrect, and multiple mutations can actually occur in an individual organism at once (yet be stable in descendants of that organism), then the numerous equivalent SNPs one finds in all of L21 -- and at many other phylogenetic branching points -- need not be younger than the one chosen by modern researchers to represent the parent.

Obviously we don't know whether the SNP chosen to label the clade is younger or older than the equivalent SNPs, if we knew they wouldn't be equivalent.

Multiple mutations *can* occur in one generation, there's nothing preventing it. SNP counting does not rely on SNPs occurring singly, it just takes an average rate, assuming that they occur more or less randomly. The number of mutations that has accumulated in different branches of the same age often varies wildly, which is why there are large confidence intervals around the estimated dates.

Or are you suggesting that mutations occur in batches *as a rule*?

Hector
04-08-2015, 05:53 PM
Multiple mutations *can* occur in one generation, there's nothing preventing it. SNP counting does not rely on SNPs occurring singly, it just takes an average rate, assuming that they occur more or less randomly. The number of mutations that has accumulated in different branches of the same age often varies wildly, which is why there are large confidence intervals around the estimated dates.


SNPs are assumed to accumulate in a Poisson process. Ins/Del may depend on the type but on average they too are assumed to be Poisson-distributed. Micro and mini satellites are modeled in terms of Markov process but random nature of it is still largely Poisson.

Large confidence intervals are more due to the uncertainty concerning the true rate of mutation than variation among branches. How much the rate varies is hotly debated but it should not be too much; otherwise it would be impossible to date the age of a clade based on the number of SNPs it accumulates.

P.S. A small number of SNPs utilized will also increase the confidence interval. Ascertainment bias theoretically does the same but it is difficult to quantify this so age estimate is done only with a pool of SNPs that are assumed to have been collected randomly like within certain segments etc.

razyn
04-09-2015, 03:02 AM
Or are you suggesting that mutations occur in batches *as a rule*?

To the extent that I'm suggesting anything, it's that mutations can be (and are) caused by background radiation; and such an event (e.g. a nearby supernova), though not necessarily fatal or sterilizing, could cause multiple mutations (on any chromosome, including the Y) in an individual at the same time and place -- not more or less, pretty close to, kind of the same average time -- retained in his lineally descendant family.

As long as such mutations don't decrease the survival chances of the bearer, they will all transmit to all the offspring, whether they occurred serially or in bursts. Seeing seven (or seventeen) at one branching point -- but only one at many other branching points -- suggests to me that the branching points at which we see multiple equivalent SNPs may have resulted from interference of some kind. Whether that has to have been cosmic, or (though I suspect it's less likely, prior to the mid-20th century) might have been terrestrial, I have no idea. There clearly have been spikes in the cosmic background radiation, recorded in carbon 14 dating of oceanic sediments. Correlation of those spikes with multiple equivalent SNPs would be a neat trick, if it could be done.

miiser
04-09-2015, 04:33 AM
To the extent that I'm suggesting anything, it's that mutations can be (and are) caused by background radiation; and such an event (e.g. a nearby supernova), though not necessarily fatal or sterilizing, could cause multiple mutations (on any chromosome, including the Y) in an individual at the same time and place -- not more or less, pretty close to, kind of the same average time -- retained in his lineally descendant family.

As long as such mutations don't decrease the survival chances of the bearer, they will all transmit to all the offspring, whether they occurred serially or in bursts. Seeing seven (or seventeen) at one branching point -- but only one at many other branching points -- suggests to me that the branching points at which we see multiple equivalent SNPs may have resulted from interference of some kind. Whether that has to have been cosmic, or (though I suspect it's less likely, prior to the mid-20th century) might have been terrestrial, I have no idea. There clearly have been spikes in the cosmic background radiation, recorded in carbon 14 dating of oceanic sediments. Correlation of those spikes with multiple equivalent SNPs would be a neat trick, if it could be done.

I think there is significant evidence that SNP mutations do, at least in some cases, occur in runs.

In addition to environmental factors, there is also the possibility of mutation runs occurring through structural instability of the DNA molecule. DNA is, after all, a physical structure. The nucleotide sequence affects the physical shape of the molecule. Some physical configurations are more stable than others from a mechanical perspective. For example, it has been documented that DNA tends to maintain a constant ratio of each of the four nucleotides. I suppose this is because each nucleotide is not structurally identical, and getting too far out of balance with nucleotide ratios will cause physical stresses which increase the probability of mutations which restore the physical structure of the molecule to a more neutral state.

So we can suppose that an initial random mutation could generate a run of subsequent mutations over a small number of generations until the molecule attains a more stable configuration.

If SNP mutations are solitary and independent, then we should expect to see them randomly distributed throughout the chromosomes. However, if multiple SNP mutations can occur as a run due to a physical instability, then we should expect to see some SNPs occurring adjacent to or near to each other, more so than would be expected from a random distribution. In fact, a quick review of a few test kits shows that there are some SNP mutations which are near to each other. (These SNPs are phylogenetically consistent and therefore real data, not misreads.) This suggests that some SNPs do have a causal relationship with other SNPs, and occur in runs.

We also know that there are disease states which can cause damage to DNA (oxidative stress, etc.), which might hypothetically lead to an increased mutation rate within certain individuals.

In addition to the large blocks of phylogenetically equivalent SNPs being indicative of mutation runs, there is also the large number of branches per node seen in some subclades. For example, the large number of branches per node seen throughout L21 (averaging around 4, and in some cases as great as 7 or 8) is more than should be expected from a uniform distribution of singular SNPs. The tree structure is more consistent with multiple SNPs occurring within a short time span, and thus being concentrated together as phylogentically equivalent blocks, with each run occurring less frequently than if assumed to be singular.

The increasing mutation rate with older fathers should also tend to cluster SNPs together within a branch.

The singular, uniformly distributed SNP mutation model is not a good fit to the data. Mutation runs, at least in some cases, fit the data better. I suspect all of the above mentioned effects contribute to mutations occurring in runs.

Hector
04-09-2015, 04:53 AM
Multiple mutations in a single generation do not usually mess up the age estimate in the long run unless the number is huge like in hundreds. No such case has been reported so far. Most Poisson processes are not really Poisson when all the information is known but still the model works in the long run.

miiser
04-09-2015, 05:58 AM
Multiple mutations in a single generation do not usually mess up the age estimate in the long run unless the number is huge like in hundreds. No such case has been reported so far. Most Poisson processes are not really Poisson when all the information is known but still the model works in the long run.

I agree that, given enough time, the rate will smooth out and approach an average. I'm not convinced L21 is old enough for this to be the case. In order for the rate to approach the long term average, the age of the subclade needs to be at least several multiples of the mutation run period. A mutation run of length much less than 100 is capable of disrupting age estimates for L21 and its peers. At an average of ~130 years per SNP, a run of 10 SNPs would affect an age estimate by +/- 1300 years, which is a significant fraction of L21's age. If the typical run length is, for example, 10 SNPs, then we are looking at probably just a few dice rolls since the appearance of L21.

Megalophias
04-09-2015, 06:54 AM
There are definitely various effects (environmental or biological) that could lead to significant mutation rate heterogeneity (it certainly varies across species, and even within modern humans A seems to have accumulated significantly fewer mutations than CDEF according to Scozzari et al).

It's better to use as many branches as possible descending from the same node and take the average (or some fancy Bayesian equivalent of the average) branch length to calculate the age of the node, rather than just one branch. That way the outliers tend to cancel out. Of course higher up in the tree you have fewer branches per node, and systemic changes in mutation rate could still throw it off.

In this particular case, though, do we know the duration of the branch in question since its parent node? And does the number of mutations exceed the expected value when you consider all the nodes you are comparing?

miiser
04-09-2015, 07:13 AM
In this particular case, though, do we know the duration of the branch in question since its parent node? And does the number of mutations exceed the expected value when you consider all the nodes you are comparing?

I'm not aware of any other method to independently establish the duration of the branch apart from mutation counting. So I think the answer is: no, we don't know the duration, and there is no other source of data on which to base an "expected value" to verify that the data is consistent.

Hector
04-09-2015, 06:45 PM
... even within modern humans A seems to have accumulated significantly fewer mutations than CDEF according to Scozzari et al).

I doubt that. My guess is that the error comes from the fact that a higher proportion of SNPs for a person belonging to A are considered private for obvious reasons.
To accuse a researcher of this kind of error will be considered rude in academic circles but I have seen far better known scientists make even more elementary errors.

lgmayka
04-09-2015, 11:56 PM
This suggests that some SNPs do have a causal relationship with other SNPs, and occur in runs.
Correlative, but not necessarily causal.

Back in 2008, Thomas Krahn enigmatically wrote (http://archiver.rootsweb.ancestry.com/th/read/genealogy-dna/2008-06/1213770183):
---
I even believe that a large share of all measured mutations can be connected to some kind of recombination events.
---
Recombination can, of course, affect an entire segment of the Y chromosome, not just one location.

He elaborated on this a little at the 2010 FTDNA conference (http://www.dna-fingerprint.com/static/FTDNA-Conference-2010-WalkThroughY.pdf):
---
Many of the derived SNPs on the existing Y tree have their derived allele from the ChrX or other autosomes.
...
The term NRY (non recombining part of the Y chromosome) is completely obsolete!
---

miiser
04-10-2015, 03:24 AM
Correlative, but not necessarily causal.

Back in 2008, Thomas Krahn enigmatically wrote (http://archiver.rootsweb.ancestry.com/th/read/genealogy-dna/2008-06/1213770183):
---
I even believe that a large share of all measured mutations can be connected to some kind of recombination events.
---
Recombination can, of course, affect an entire segment of the Y chromosome, not just one location.

He elaborated on this a little at the 2010 FTDNA conference (http://www.dna-fingerprint.com/static/FTDNA-Conference-2010-WalkThroughY.pdf):
---
Many of the derived SNPs on the existing Y tree have their derived allele from the ChrX or other autosomes.
...
The term NRY (non recombining part of the Y chromosome) is completely obsolete!
---

Fair enough, close proximity SNPs could conceivably be the result of a single recombination event. But even if this were the case, they would still represent a single mutation event which would be counted as multiple SNPs under the currently prevailing SNP counting models. In other words, they would be a mutation run rather than isolated, random, singular mutations. The point is that SNPs may occur in runs and be unevenly distributed through time.

If such runs do occur, then it would dramatically increase the error range of SNP count based age estimates, as the current models all assume single isolated mutations as far as I can tell.

If this is the case, then it is less reasonable to state with confidence that a particular subclade is younger or older than its sibling.

Megalophias
04-10-2015, 05:15 AM
I doubt that. My guess is that the error comes from the fact that a higher proportion of SNPs for a person belonging to A are considered private for obvious reasons.
To accuse a researcher of this kind of error will be considered rude in academic circles but I have seen far better known scientists make even more elementary errors.

Even professionals can make elementary errors, to be sure, but it is customary to actually read the paper before assuming they did :D. Of course they did not.

Also relevant to this thread: "Fig. 1 shows the distribution of the variant positions across the five selected MSY regions. The apparent uneven density of variant positions can be explained by the different occurrence of repetitive elements, which were largely excluded from targeted sequences.... When the occurrence of the 2,386 variants in each of the 5,274 sequenced DNA fragments (see Methods) was considered, no evidence of any uneven distribution was obtained. The linear slope (0.001591) closely matched the overall rate of occurrence (2,386/1,495,512 = 0.001595) and all points but two fell within the 0.999 confidence interval estimated according to the Poisson distribution (Supplemental Fig. S1)." (They sequenced 1.5 Mb at ~50x, choosing the least X-analogous regions and avoiding repetition-prone areas.)

Hector
04-10-2015, 06:02 AM
...If this is the case, then it is less reasonable to state with confidence that a particular subclade is younger or older than its sibling.

I think by definition sibling clades have the same age. A clade may be older or younger than a "cousin clade" though.
X-Y recombination mechanism is not yet understood except for pseudo autosomal regions.
If it happens frequently it simply means that pseudo autosomal regions should be expanded to include other regions than the ends.
There are already "NRY" regions of Y chromosome that are avoided in research, X-transposed and ampliconic(palindromic segments are included in the latter).

And if X-Y recombination occurs it must run by a mechanism hitherto unknown since recombination on autosomes usually involves only 2-3 crossover points and each segment is massive in size.
Even if a segment of 1Mbp is exchanged you should expect r x 8 x (10 ** -10) x 10000000 = r x 8 x 10**-4.
Assuming a separation of at least 1 million years between X and Y this yields > 800 SNPs introduced.(the separation time is tricky since the segments may have been exchanged earlier. It depends on the size of the exchanged area to the whole area susceptible to recombination and the frequency of recombination)

Anyway it can be treated the same as any UEP of Y-Y kinds etc. and we already know a lot about this process.

If the "SNP runs" are fewer than by 10's, SNP counting is still a viable method of age estimate for older clades even if such are very frequent. The clock becomes coarse but not useless.
We should also remember we still have good old STRs for age estimates of shallow time depths.


Even professionals can make elementary errors, to be sure, but it is customary to actually read the paper before assuming they did :D. Of course they did not.

Link to the paper you mentioned?
I was going to ask this earlier but at that time I planned on quitting this forum(something that I declared I was going to do 5 months ago anyway).

Net Down G5L
04-10-2015, 06:47 AM
I'm still up in the air about it, but if Remedello I is not R1b, then neither were the stelae people (or at least does not look compatible with), since the earlier stelae in Italy have what are clearly Remedello I daggers carved into them. These are very different from the later Italian stelae that have Bell Beaker daggers, which in in Iberia and elsewhere, resemble Yamnaya ones.

What about CTS7822 being associated with stelae? And if so, could it have been in Maritime Bell Beaker (as an alternative R1b)?
Not much of it in the ht35 project but....
Generally lots of interesting data in that project that will take some explaining.

Megalophias
04-10-2015, 07:06 AM
The Scozzari et al paper is here: http://genome.cshlp.org/content/early/2014/02/02/gr.160788.113


I'm not aware of any other method to independently establish the duration of the branch apart from mutation counting. So I think the answer is: no, we don't know the duration, and there is no other source of data on which to base an "expected value" to verify that the data is consistent.

Well, you surely must have some way of estimating it or you would not consider the number of mutations anomalous.

I was thinking of the difference between the estimated age of the parent node (based on the sister branches) and the estimated age of the node itself (based on daughter branches). That should give you an expected number of mutations in the particular branch you are looking at - provided you have an adequate number of sister and daughter branches, of course!

miiser
04-10-2015, 07:19 AM
The Scozzari et al paper is here: http://genome.cshlp.org/content/early/2014/02/02/gr.160788.113



Well, you surely must have some way of estimating it or you would not consider the number of mutations anomalous.

I was thinking of the difference between the estimated age of the parent node (based on the sister branches) and the estimated age of the node itself (based on daughter branches). That should give you an expected number of mutations in the particular branch you are looking at - provided you have an adequate number of sister and daughter branches, of course!

I am not arguing that L21 ISN'T younger than its siblings, but just challenging the soundness of the argument used by others to assert that it is ALMOST CERTAINLY younger than its siblings.

I am not asserting that the number of mutations in L21 IS anomalous, but just supporting the comment made by Razyn, suggesting that it could POSSIBLY be anomalous - that mutation runs may have occurred, so we cannot assume that SNPs are evenly distributed through time and that a large block of phylogenetically equivalent SNPs must therefore represent a proportionally long time period. If mutation runs are a real thing, then a large number of SNPs may occupy a small number of generations or even a single generation.

I agree that counting SNPs in multiple branches will tighten up the confidence interval for the parent's age estimate. But if mutation runs do occur, then a differing number of phylogenetically equivalent SNPs from the parent down to each of its descendants may not be a reliable indicator of the relative age of each descendant.

The argument claiming that L21 must be younger than its siblings (which started the discussion down this path) relied on counting SNPs within a single branch relative to its siblings. This reasoning is less certain if the large block of phylogenetically equivalent SNPs could represent a single mutation run event rather than a series of unrelated solitary mutations.

Averaging out the SNP count to a parent by using counts from multiple branches is effective at reducing the error from such effects. But this is not the approach that was utilized in asserting that L21 must be younger than its siblings. That argument relied on the number of phylogenetically equivalent SNPs in L21 compared to its siblings.

Counting the number of SNPs from present time up to L21 and its siblings, using data from every available descendant branch, should be less vulnerable to such effects, as it is less dependent on the count within a single branch.

razyn
04-10-2015, 01:28 PM
I am not asserting that the number of mutations in L21 IS anomalous, but just supporting the comment made by Razyn, suggesting that it could POSSIBLY be anomalous - that mutation runs may have occurred, so we cannot assume that SNPs are evenly distributed through time and that a large block of phylogenetically equivalent SNPs must therefore represent a proportionally long time period. If mutation runs are a real thing, then a large number of SNPs may occupy a small number of generations or even a single generation.

Quod erat demonstrandum. Thank you for noticing the actual argument. I have not argued against the use of statistics, Poisson distributions, Bayesian corrections of them, bell curves... or smart people, in general. I think the guys at YFull have it almost right, and will get closer as we feed them better data. The quibble (with something Armando said, involving the word "should") relates to a fly in the ointment -- namely multiple equivalent SNPs that IMO may be better considered as one mutational event than as a string of sequential mutations following the same, hypothetically "constant" rate for such mutations. We know they exist, and so far in insufficiently large numbers to expect them to conform to a law of large numbers. It seems to me that the number of such multiply-equivalent SNP clusters is steeply on the rise, as more NextGen sequencing results arrive and are compared. And that seems to me a cautionary tale, in itself.

Hector
04-10-2015, 04:25 PM
The Scozzari et al paper is here: http://genome.cshlp.org/content/early/2014/02/02/gr.160788.113


I read the paper and I don't think the discrepancy is due to "SNP runs" because there is a certain trend running from A to B to C-T with increasing number of mutations per lineage.

I hope they just made counting errors. Or they missed quite a lot of SNPs for A haplogroup and then as they typed more individuals per HG when they moved to B to C-T the missed SNPs decreased resulting in overall increase in SNPs per lineage.

Megalophias
04-11-2015, 01:32 AM
Sorry guys, I wasn't trying to give you the third degree, I am just interested in cases of rate heterogeneity and thought you might have a good example. I do agree with you that there are many things we don't know that could totally throw off our calculations.

BTW this is something the professionals are concerned with as well - from the same paper again "Major improvements in dating may derive from taking into account the complexity of the mutational process (Michaelson et al. 2012). For example, the possibility that clusters of mutations may hit the MSY seriously challenges the concept of linear accumulation with time. We highly recommend that, in future, mutation rates be worked according to the local features of the MSY sequences, and that they then be used on appropriately partitioned datasets as previously suggested (Fu et al. 2013a). We see deep-rooted pedigrees (Xue et al. 2009) as the material that should be chosen in order to work out robust estimates of these rates, given the low chances of observing mutational events in such a small portion of the genome in a single generation."

Hector:

Possibly it could be the number of missed SNPs when fewer samples were used, I couldn't find anything addressing this.

R.Rocca
04-11-2015, 02:06 AM
Guys, please stay on topic. There are plenty of posts about SNP/STR counting, mutation rates, etc.

rms2
04-11-2015, 07:31 PM
It damned unlucky for P312 folks that their yDNA has to be linked to beaker - one of the most complex and baffling archaeological phenomenons . . .

I wanted to react to that when Alan first wrote it, but I guess I got distracted. Now it gives me the chance to get this thread back to the Beaker Folk.

I agree with Alan's sentiment, not because I think the Beaker Folk are not interesting or don't make cool y-dna progenitors or anything like that. I agree because, as Alan indicated, Beaker is not simple and straightforward. It doesn't give us the smoking gun right to Yamnaya or some other easily identified group. Full-blown Beaker is very Kurgan-looking (in Gimbutas' sense of the term) and thus far is connected to R1b (and P312), which has an east-to-west phylogeography. But the archaeological orthodoxy says Beaker began in Portugal and moved east. This paradox gives those who oppose connecting Beaker and/or R1b to the Indo-Europeans, including those who still cling to the old R1b-out-of-Iberia hypothesis, something to quibble about. And we all know and lately have really seen, I think, that there is plenty of genuinely visceral hatred out there for the idea that R1b has any connection to the spread of Indo-European languages into peninsular Europe.

So, until we get that smoking gun that establishes the connection of R1b-L51 to the Indo-Europeans beyond a reasonable doubt, or at least with an overwhelming preponderance of the evidence, I agree with the sentiment expressed by Alan.

David Mc
04-11-2015, 08:13 PM
So, until we get that smoking gun that establishes the connection of R1b-L51 to the Indo-Europeans beyond a reasonable doubt, or at least with an overwhelming preponderance of the evidence, I agree with the sentiment expressed by Alan.

Well said. I suspect the fury and the denials will continue even when there is an "overwhelming preponderance of the evidence," but it will be great to see the migration map filled in step-by-step.

alan
04-12-2015, 12:37 AM
Way I look at it is that in Samara we have one branch that is L23xXZ2103 and several that are Z2103. So we have TWO BRANCHES OF L23 IN SAMARA which by definition split before the Z2103. By 3100-2900BC at Samara those branches had split over 1000 years earlier but yet remained in the same area. I think that is significant.

Agamemnon
04-12-2015, 01:32 AM
I wanted to react to that when Alan first wrote it, but I guess I got distracted. Now it gives me the chance to get this thread back to the Beaker Folk.

I agree with Alan's sentiment, not because I think the Beaker Folk are not interesting or don't make cool y-dna progenitors or anything like that. I agree because, as Alan indicated, Beaker is not simple and straightforward. It doesn't give us the smoking gun right to Yamnaya or some other easily identified group. Full-blown Beaker is very Kurgan-looking (in Gimbutas' sense of the term) and thus far is connected to R1b (and P312), which has an east-to-west phylogeography. But the archaeological orthodoxy says Beaker began in Portugal and moved east. This paradox gives those who oppose connecting Beaker and/or R1b to the Indo-Europeans, including those who still cling to the old R1b-out-of-Iberia hypothesis, something to quibble about. And we all know and lately have really seen, I think, that there is plenty of genuinely visceral hatred out there for the idea that R1b has any connection to the spread of Indo-European languages into peninsular Europe.

So, until we get that smoking gun that establishes the connection of R1b-L51 to the Indo-Europeans beyond a reasonable doubt, or at least with an overwhelming preponderance of the evidence, I agree with the sentiment expressed by Alan.

Quite so, it's completely irrational and seems to be quite trendy... Experienced it myself for years now, as I'm sure you have as well!
Even if we get the "smoking gun" linking R1b-L51 to the IEs (and, IMHO, the Haak et al. 2015 results were but another nail in the coffin of idiosyncratic theories revolving around R1b since everything we know about L51 points towards an association with the break up and dispersal of IE), we'll still have to deal with that irrational amount of obnoxious anti-R1b hatred methinks.

vettor
05-27-2015, 09:11 PM
I find it very interesting that the Haak ydna findings in his paper in regards to central Germany , that is LBK_EN ones, are all within 10 kms of the Goseck circles of Neolithic Germany
http://en.wikipedia.org/wiki/Goseck_circle

some time later Rossen culture moved into the area...........maybe bringing the R1 people

yxc
05-27-2015, 09:58 PM
those had not moved from the east . CW did so post 3,000 bc.

Globular Amphora was before though in proximity and they were 'ANE' and 'Steppe' I'd say

alan
05-30-2015, 08:54 PM
Interesting posts on things that may effect mutations SNPs in L21 etc above that I missed. A couple of things that spring to mind are heavy use of Arsenical Ross Island Copper which was the only mined copper in the isles in the early pre-tin-Bronze beaker phase. Cannot help thinking that might cause a bit of genetic damage. There was an intense period of its use in the isles starting in the 2400sBC and tailing off as tin alloying allowed harder metals to be created from non-Arsenical Bronze c. 2200BC. So, it could be argued there is a 2-300 year window of lots of Arsenic poisoning before it reduced. If L21 dates to c. 2500BC or just after then the first 300 years after it saw a unique period of Arsenic exposure in the isles and adjacent areas of the continent who also used this arsenical metal. Could this be linked to the odd run of SNPs between L21 and DF13? It does seem to work in terms of date, geography etc.

Another factor of course could be average age a child is fathered. There seems clear evidence of elite hogging of reproduction in the structure of L21. This will almost certainly raise the average age of a son's father at the time of their birth. If this pattern was established very early when the numbers of L21 people controlling the metals was very small then the effect of increase mutations could have been magnified. Again, in terms of L21 and beaker, there was that short period where basically all the locally mined copper used in early isles beaker metalwork was coming from one source - the Ross Island arsenical copper mine in SW Ireland. So again the timeslot that stands out is between some starting point c. 2500-2400BC and an ending point around 2200BC when tin bronze made the unusually hard arsenical copper of Ross Island a whole lot less important.

I realise the implication of this is that L21 arrived from the continent but DF13 may have arisen in the isles. I still firmly believe L21 is continental in origin but I have an open mind on DF13. It could have arisen on the isles.

By the way a book on bell beaker is due out in 4 weeks -

The Bell Beaker Transition in Europe: Mobility and local evolution during the 3rd millennium by Maria Pilar Prieto Martínez and Laure Salanova