Age of Z253?

[alan trowel hands.]

...in light of it being present in at least three clusters which are very different in STR terms?

[MHammers]

I think the "unassigned" category members will give us the best estimate for age.  This is where the largest variance will likely be found.  I did notice 4 of the Spanish Z253 "unassigned" had a very low variance at 67 markers with .06.  A founder effect perhaps?  That's not much to go on  though.  Unfortunately, it's still too early to make heads or tails.  Some of these identified categories may become a downstream SNP and they'll have to be calculated accordingly.

[alan trowel hands.]

I think the "unassigned" category members will give us the best estimate for age.  This is where the largest variance will likely be found.  I did notice 4 of the Spanish Z253 "unassigned" had a very low variance at 67 markers with .06.  A founder effect perhaps?  That's not much to go on  though.  Unfortunately, it's still too early to make heads or tails.  Some of these identified categories may become a downstream SNP and they'll have to be calculated accordingly.

What I really meant was if you look at Iberian Z253, Irish types 3 and 4 and other outliers (all of whom are Z253) then what sort of variance does that give.  Does it suggest that the Z253 SNP is old.  I would have thought if the common Z253 ancestor of Irish 3 and 4 was a complete surprise to those who look at the STRs, then the Z253 SNP shared by the two Irish clusters, the Spanish folk and others must be old. 

[Mikewww]

Unfortunately, since you don't need long haplotypes to identify Irish IV people so we don't have as many 67 STR ht's as we should.  I think this is enough to have a good idea though.  The TMRCA for Z253 may be (probably is) older than this, but the TMRCA for Irish III and IV is about 2.8k ybp.

Even though neither cluster is that old, they are different enough that they separated quite a while ago.


IrishIII&IrishIV Interclade Z253** MRCA Age _ 2.8 (3.5-2.0)  N=209
IrishIII Clade MRCA Age _____________________ 1.2 (1.4-1.1)  N=183
IrishIII Clade Coalescence Age ______________ 1.2 (1.3-1.1)  N=183
IrishIV Clade MRCA Age ______________________ 1.3 (1.4-1.1)  N=26
IrishIV Clade Coalescence Age _______________ 1.1 (1.3-1.0)  N=26

[alan trowel hands.]

Unfortunately, since you don't need long haplotypes to identify Irish IV people so we don't have as many 67 STR ht's as we should.  I think this is enough to have a good idea though.  The TMRCA for Z253 may be (probably is) older than this, but the TMRCA for Irish III and IV is about 2.8k ybp.

Even though neither cluster is that old, they are different enough that they separated quite a while ago.


IrishIII&IrishIV Interclade Z253** MRCA Age _ 2.8 (3.5-2.0)  N=209
IrishIII Clade MRCA Age _____________________ 1.2 (1.4-1.1)  N=183
IrishIII Clade Coalescence Age ______________ 1.2 (1.3-1.1)  N=183
IrishIV Clade MRCA Age ______________________ 1.3 (1.4-1.1)  N=26
IrishIV Clade Coalescence Age _______________ 1.1 (1.3-1.0)  N=26

I would agree that I suspect the interclade is probably in reality older than that.  Would adding the continental Z253 guys help? 

[Mikewww]

...

IrishIII&IrishIV Interclade Z253** MRCA Age _ 2.8 (3.5-2.0)  N=209..

I would agree that I suspect the interclade is probably in reality older than that.  Would adding the continental Z253 guys help?  

The 2.8k ybp is the interclade age between Irish III(L226) and Irish IV.
This age should be fairly accurate, however, by definition the true MRCA for all of Z253 can not be any younger than 2.8 (actually the low end of the range or 2.0).

If you calculate the intraclade age for all of Z253 you actually get a number less than 2.8k. This is because the youthfulness and high number of L226 is dragging the average down. The value of the interclade age is that it is more accurate.

In an effort to corner in the upper age of Z253 I ran interclades between all of Z253 and L513  (11-13 Combo) as well as Z255 (Irish Sea L159.2 is.)

L513&Z253 Interclade L21* MRCA Age _ 3.2 (3.7-2.8)  N=856
Z255&Z253 Interclade L21* MRCA Age _ 2.5 (3.1-1.9)  N=497

The upper range of the younger pair is 3.1k ybp for Z255 and Z253's interclade MRCA man.

So that's it. Fairly reliably we have that the Z253 can be no younger than 2.0k ybp and no older than 3.1k ybp.

[MHammers]

This is much younger than the Beaker period and more like Iron age.  What connects Scandinavia, Iberia, British Isles, and Switzerland in that time-frame?  Celtic speakers seems like an obvious choice, except for Scandinavia.

[alan trowel hands.]

...

IrishIII&IrishIV Interclade Z253** MRCA Age _ 2.8 (3.5-2.0)  N=209..

I would agree that I suspect the interclade is probably in reality older than that.  Would adding the continental Z253 guys help?  

The 2.8k ybp is the interclade age between Irish III(L226) and Irish IV.
This age should be fairly accurate, however, by definition the true MRCA for all of Z253 can not be any younger than 2.8 (actually the low end of the range or 2.0).

If you calculate the intraclade age for all of Z253 you actually get a number less than 2.8k. This is because the youthfulness and high number of L226 is dragging the average down. The value of the interclade age is that it is more accurate.

In an effort to corner in the upper age of Z253 I ran interclades between all of Z253 and L513  (11-13 Combo) as well as Z255 (Irish Sea L159.2 is.)

L513&Z253 Interclade L21* MRCA Age _ 3.2 (3.7-2.8)  N=856
Z255&Z253 Interclade L21* MRCA Age _ 2.5 (3.1-1.9)  N=497

The upper range of the younger pair is 3.1k ybp for Z255 and Z253's interclade MRCA man.

So that's it. Fairly reliably we have that the Z253 can be no younger than 2.0k ybp and no older than 3.1k ybp.

well if Z253 is between 1100BC and 0BC/AD then the obvious candidate is the Atlantic Bronze Age of c. 1300-700BC.  In the period after that, the Iron Age, Ireland features no Iberian material or indications of contacts.  The Bronze Age is a far more likely phase for some modest Irish-Iberian contact.  However, we should not forget that a major reason there is a hotspot of Z253 in Ireland and to a lesser degree Iberia is because of two Medieval guys who produced a cluster. Without them Z253 might not be considered especially strong in those areas.  So, we have to be cautious about this.  It is quite possible that Z253 sprung up in in neither Iberia or Ireland. 

[MHammers]

Here are a couple of more interclades.  This time with DF21*/** and DF23*/** compared with Z253*/**.  All Locations

DF 21 and Z253
  Interclade MRCA Age __ 3.8 (4.2-3.4)  N=29         
DF21 Clade MRCA Age ________ 3.4 (3.6-3.1)  N=15         
  Clade Coalescence Age _ 3.0 (3.3-2.7)  N=15         
Z253 Clade MRCA Age ________ 4.1 (4.6-3.6)  N=14         
  Clade Coalescence Age _ 3.1 (3.6-2.6)  N=14

DF 23 and Z253
  Interclade MRCA Age __ 5.2 (6.0-4.4)  N=23         
DF23 Clade MRCA Age ________ 4.7 (5.2-4.2)  N=9         
 Clade Coalescence Age _ 3.7 (4.2-3.2)  N=9         
Z253 Clade MRCA Age ________ 4.1 (4.6-3.6)  N=14         
 Clade Coalescence Age _ 3.1 (3.6-2.6)  N=14   

It is interesting that the confidence intervals don't even overlap on the interclade estimates.  I think the oldest Z253 may be around 4.4-4.2 kya when the downstream snps are excluded, which is right in line with the Beaker period in Britain/Ireland.      
   

      

[alan trowel hands.]

Here are a couple of more interclades.  This time with DF21*/** and DF23*/** compared with Z253*/**.  All Locations

DF 21 and Z253
  Interclade MRCA Age __ 3.8 (4.2-3.4)  N=29         
DF21 Clade MRCA Age ________ 3.4 (3.6-3.1)  N=15         
  Clade Coalescence Age _ 3.0 (3.3-2.7)  N=15         
Z253 Clade MRCA Age ________ 4.1 (4.6-3.6)  N=14         
  Clade Coalescence Age _ 3.1 (3.6-2.6)  N=14

DF 23 and Z253
  Interclade MRCA Age __ 5.2 (6.0-4.4)  N=23         
DF23 Clade MRCA Age ________ 4.7 (5.2-4.2)  N=9         
 Clade Coalescence Age _ 3.7 (4.2-3.2)  N=9         
Z253 Clade MRCA Age ________ 4.1 (4.6-3.6)  N=14         
 Clade Coalescence Age _ 3.1 (3.6-2.6)  N=14   

It is interesting that the confidence intervals don't even overlap on the interclade estimates.  I think the oldest Z253 may be around 4.4-4.2 kya when the downstream snps are excluded, which is right in line with the Beaker period in Britain/Ireland.      
   

      

Interesting.  Well that would open up the entire copper and bronze age as possibilities.  One thing of interest I have read in Iberian papers is that Atlantic Iberia has a couple of phases of stronger contacts northwards in the early beaker period c. 2500BC and in the Atlantic Bronze Age period c. 1300-800BC with a gap in between (perhaps c. 800 years long) when contact with the north Atlantic was less and Atlantic Iberia was more linked to the west Med. as far as Italy.  That to me seems to be a scenario which would fit the non-Celtic, perhaps Italic-like dialects such as Lusitanian and Ligurian which seem overlain by Celtic.  However, I would feel that rather than the traditional Urnfield-Hallstatt-La Tene explanation for the spread of Celtic (which frankly doesnt really work in western Europe) that the second phase of contact of Atlantic Iberia with the areas to the north known as the Atlantic Bronze Age is the best scenario for the emergence of Celtic. In general I see the spread and establishment of the clade pattern as something that happened before the emergence of a distinctive Celtic dialect, something I think was to do with elite/trade networks in the later Bronze Age. 

[Mikewww]

Here are a couple of more interclades.  This time with DF21*/** and DF23*/** compared with Z253*/**.  All Locations

DF 21 and Z253
  Interclade MRCA Age __ 3.8 (4.2-3.4)  N=29         
.....
DF 23 and Z253
  Interclade MRCA Age __ 5.2 (6.0-4.4)  N=23
.....

Using the fencing in technique, L21 can't be any younger than the lowest range of its oldest interclade nodal pair, DF23 and Z253 it looks like, here.  Hence L21, has a 68% chance of being 4.4k years old or older.

I think we want to the number of haplotypes up, though. These are fairly small sample sizes. It looks like you are using the DF23* (M222-) only and Z253* (L226-) only.

I guess I should point out that in the numbers I calculated below, I went beyond using only confirmed SNP tested folks. I used all the 67 STR haplotypes that fit into the varieties (i.e. IrishIII and IrishIV) regardless of SNP testing. I could have introduced some error there.

[MHammers]

Using the fencing in technique, L21 can't be any younger than the lowest range of its oldest interclade nodal pair, DF23 and Z253 it looks like, here.  Hence L21, has a 68% chance of being 4.4k years old or older.

I think we want to the number of haplotypes up, though. These are fairly small sample sizes. It looks like you are using the DF23* (M222-) only and Z253* (L226-) only.

I guess I should point out that in the numbers I calculated below, I went beyond using only confirmed SNP tested folks. I used all the 67 STR haplotypes that fit into the varieties (i.e. IrishIII and IrishIV) regardless of SNP testing. I could have introduced some error there.

I agree about the sample sizes.  Is DF23 coming up older than its brothers?  

I was looking again at the intraclade MRCA ages and found something of a pattern for DF23, L513, and Z253.  Note that these are */** types except in L513 I included 3 undiffentiated continentals.  These are the oldest layers of the L21 "onion" with continental members.
 
DF23 Clade MRCA Age ________ 4.7 (5.2-4.2)  N=9 (France)
L513 Clade MRCA Age ________ 4.2 (4.6-3.9)  N=39 (undifferentiated France, Germany, Netherlands)
Z253 Clade MRCA Age ________ 4.1 (4.6-3.6)  N=14 (Spain, Switzerland, Norway)

Although, we still don't have alot of continentals tested on these, it looks like the earliest L21's may have been close to the Rhine, or more likely France.  Then after that a starlike radiation into the Isles, Norway, Spain, and Switzerland at various times.   

This reminds me more of the secondary Beaker movement coming down the Rhine and into the Isles and Scandinavia.  Of course, many would have settled in France.  The earlier Beaker period of Iberia looks slightly more associated with L11, P312, and Z196.

[alan trowel hands.]

Using the fencing in technique, L21 can't be any younger than the lowest range of its oldest interclade nodal pair, DF23 and Z253 it looks like, here.  Hence L21, has a 68% chance of being 4.4k years old or older.

I think we want to the number of haplotypes up, though. These are fairly small sample sizes. It looks like you are using the DF23* (M222-) only and Z253* (L226-) only.

I guess I should point out that in the numbers I calculated below, I went beyond using only confirmed SNP tested folks. I used all the 67 STR haplotypes that fit into the varieties (i.e. IrishIII and IrishIV) regardless of SNP testing. I could have introduced some error there.

I agree about the sample sizes.  Is DF23 coming up older than its brothers?  

I was looking again at the intraclade MRCA ages and found something of a pattern for DF23, L513, and Z253.  Note that these are */** types except in L513 I included 3 undiffentiated continentals.  These are the oldest layers of the L21 "onion" with continental members.
 
DF23 Clade MRCA Age ________ 4.7 (5.2-4.2)  N=9 (France)
L513 Clade MRCA Age ________ 4.2 (4.6-3.9)  N=39 (undifferentiated France, Germany, Netherlands)
Z253 Clade MRCA Age ________ 4.1 (4.6-3.6)  N=14 (Spain, Switzerland, Norway)

Although, we still don't have alot of continentals tested on these, it looks like the earliest L21's may have been close to the Rhine, or more likely France.  Then after that a starlike radiation into the Isles, Norway, Spain, and Switzerland at various times.   

This reminds me more of the secondary Beaker movement coming down the Rhine and into the Isles and Scandinavia.  Of course, many would have settled in France.  The earlier Beaker period of Iberia looks slightly more associated with L11, P312, and Z196.

VERY VERY interesting.  I would like to hear comments from other SNP/variance folks like Mike etc on that observation.  I do like the idea that L21 was some sort of star-like secondary takeoff of p312 from somewhere in France or east-adjacent to France.

[GoldenHind]

Although, we still don't have alot of continentals tested on these, it looks like the earliest L21's may have been close to the Rhine, or more likely France.  Then after that a starlike radiation into the Isles, Norway, Spain, and Switzerland at various times.   

A couple of years ago I suggested that a likely location for an expansion point (as opposed to an origin) for L21 was somewhere in the vicinity of the upper Rhine or Danube, perhaps southwest Germany. At the time, I was attacked by the L21-out- of-Britain devotees as someone who was just seeking a personal origin among the Germanics. I thought this was a bit odd, as I am L21- and also aware that there no Germanics anywhere in the area at the time. Nonetheless, the charge was recently repeated on the DNA Forum.
I am glad to see that I may not have been too far off.

[Mikewww]

Although, we still don't have alot of continentals tested on these, it looks like the earliest L21's may have been close to the Rhine, or more likely France.  Then after that a starlike radiation into the Isles, Norway, Spain, and Switzerland at various times.

A couple of years ago I suggested that a likely location for an expansion point (as opposed to an origin) for L21 was somewhere in the vicinity of the upper Rhine or Danube, perhaps southwest Germany. At the time, I was attacked by the L21-out- of-Britain devotees as someone who was just seeking a personal origin among the Germanics. I thought this was a bit odd, as I am L21- and also aware that there no Germanics anywhere in the area at the time. Nonetheless, the charge was recently repeated on the DNA Forum.
I am glad to see that I may not have been too far off.

I think the Old Norway project results that get down to the L21* and M222 level put a nail in that coffin. There is too much L21, even as you go into Sweden to explain away as all thralls and Norwegian traders. Heck, we even have a few Finns. The M222 ratio to L21* is much lower than in Ireland.

I'm watching these two guys closely. One is L705.2+ and the other I've had to resort to writing a letter to. They're both in my L513+ cluster labeled B-2.

N29541 Håkan Sunesson 1656-1745, Tibbhult, Kalmar län, Sw
128145 Henric Holmberg, Tammisaari, Finland

http://www.familytreedna.com/public/scandinavianydna/default.aspx?vgroup=scandinavianydna&section=yresults
http://www.familytreedna.com/public/Finland/default.aspx?vgroup=Finland&section=yresults

How do these guys fit with Welsh and West English people?  Were they coming or going?  I don't know. It would seem like they must be going (from Wales) but I have no good evidence.

I am curious. Isn't Holmberg more of German name? I know the Swede's name changed through the generations but I don't know if the same pertains to Holmberg.

Well, we have the Frenchman, Bergeron, as well. Maybe this what just part of the star-like expansion from the Rhine and our TMRCAs are just totally out of whack?  I think the key is understanding the ages (which means diversity) in conjunction with the phylogenetic tree and geographic spread.

[GoldenHind]

Holmberg might be German, but is more likely Swedish in origin. I don't know a great deal about Finland, but I do know that immigration from Sweden was quite common there, and also that many ethnic Finns used Swedish surnames. Holmberg, unlike Svensson for example, is not a patronymic name, and so would not necessarily have changed with every generation. You would have to contact the person in question and ask him if the surname remained unchanged over the generations. Surnames in Scandinavia are somewhat unusual, compared to usage in the rest of Europe. Some of my Danish ancestors used two different surnames, one patronymic and one from the name of their farm.

You and I and most others may think the L21 out of Britain idea is dead, but I'm not sure it doesn't still have its die hard adherents.

I wouldn't care to speculate about the how the ancestors of these two reached Sweden and Finland. I think an answer might possibly be suggested by the age of the SNP or cluster in question.

[NealtheRed]

Holmberg might be German, but is more likely Swedish in origin. I don't know a great deal about Finland, but I do know that immigration from Sweden was quite common there, and also that many ethnic Finns used Swedish surnames. Holmberg, unlike Svensson for example, is not a patronymic name, and so would not necessarily have changed with every generation. You would have to contact the person in question and ask him if the surname remained unchanged over the generations. Surnames in Scandinavia are somewhat unusual, compared to usage in the rest of Europe. Some of my Danish ancestors used two different surnames, one patronymic and one from the name of their farm.

You and I and most others may think the L21 out of Britain idea is dead, but I'm not sure it doesn't still have its die hard adherents.

I wouldn't care to speculate about the how the ancestors of these two reached Sweden and Finland. I think an answer might possibly be suggested by the age of the SNP or cluster in question.

L705.2 in Sweden and Finland definitely makes it harder to associate the SNP there with recent migration from the Isles. I suspect it is really old in Scandinavia, but have you compared the Swedish/Finnish STRs with folks from the Isles?

[OConnor]

I suspect Scandinavia may be under-represented because of the plague in the 1300's

Maybe M222 was hatched in a village in Scandinavia and the majority went to the Isles as fortune seekers, leaving a few in Scandinavia...who's future offspring had to deal with the plague.

half the population/or so.. of Norway died.  Maybe an M222 hot-spot was decimated?