R-L21 in Ireland not from the Bronze Age?

Mmiikkii

Regular Member
Messages
440
Reaction score
67
Points
28
Ethnic group
southern EUROPEAN
I've been watching maps from this sitehttps://indo-european.eu/maps/early-bronze-age/And I noticed that R1b-L21 is not dominant in the British Isles throughout most of the Bronze and Iron Ages.Only in the Late Iron Age comes back in prominence. Before that time, R1b-P312(the ancestor of DF-27 & U-152) and subclades(S461 in particular) are the most prevalent.P312 subclades are the dominant now in the continent.
 
I know this post was made long ago, but I’m new to this forum and just now came across it.

I often encountered this same confusion on Anthrogenica, and have noticed that there is a prevalent misconception among the community of how genetic selection works. People think, “because Bell Beaker derived British Isles population is dominated by L21 in later times, there must have been a lot of L21 in early Bell Beaker that migrated to the Isles.” People expect L21 to show up with high frequency in early Britain, Ireland, or continental Bell Beaker ancient DNA samples. It irks me that this misconception rears its head so frequently, even among preeminent forum contributors who otherwise know what they’re talking about.

They imagine a Bell Beaker population full of L21ers coming over to England en masse from the continent, as a full fledged community. The reason I bring it up now is that I still see people pushing this idea in other forums, mystified as to why we‘re not finding loads of L21 in continental early Bell Beaker sampling, and presupposing this notion of a herd of L21ers coming over from the continent to the Isles.

That‘s just not how genetic selection works.

L21 expanded within Bell Beaker and later Isles cultures, and it mostly did that expansion after arriving in the Isles. It started with a single person within a much larger Bell Beaker population. Due to advantageous genes well suited to the setting, it gradually out-competed other lineages. So we shouldn’t expect to find L21 in high frequency on the continent, or even in early Isles sampling. Supposing the MRCA is around 2600 BC (precise date not important to this topic), what we should expect to see for the Isles is something like (just a rough example, not intending to propose precise numbers):

2600 BC: 1% of Bell Beaker population
2500 BC: 5% of Bell Beaker population
2400 BC: 20% of Bell Beaker population
2300 BC: 40% of Bell Beaker population
2200 BC: 50% of Bell Beaker population
2100 BC: 53% of Bell Beaker population
etc... you get the idea

So the odds of finding L21 in early Bell Beaker is not great. This is exactly what we should expect.

There certainly was an initial founder’s effect after entry to the Isles. But subsequently it would have been a slow burn, continuously diluted by inflow from the continent. Flood’s publication fits the ancient DNA data very well, with L21 growth occurring in several waves after its initial appearance.

People who start from the fact that the modern Bell Beaker derived Isles population is largely L21, and from that deduce that a big tribe of L21ers must have migrated from the continent to the Isles as a homogeneous Bell Beaker tribe - they need to go study up on how genetic selection works.

You don’t end up with hundreds of millions of modern L21ers because of a solitary founder effect that happened 4500 years ago. This only happens because there is a genetic advantage within the lineage which leads to gradual dominance within a population.

Based on this, while the L21 MRCA’s recent ancestors certainly came from the continent, it’s not at all unreasonable to suppose that the L21 MRCA individual could have been an Isles native, and L21 expansion (amidst other extinct lineages) could possibly have occurred from within the Isles shortly following a migration from the continent and settlement in the Isles.


(As an interesting side note, members of other forums habitually conflate this notion with a theory of Isles origin (versus continental) for Isles Bell Beaker derived populations, and pointing out this misconception, thus offending the egos of prominent members, is what got me banned from Anthrogenica, LOL)
 
Last edited:
  • Like
Reactions: Syd
P312 in itself cannot be taken as an example favouring U152 and DF27 at the cost of L21
 
P312 in itself cannot be taken as an example favouring U152 and DF27 at the cost of L21
The mistake I see people frequently make is in assuming that, because L21, U152, and DF27 are the branches still extant, those are the same branches that would have been dominant in early Bell Beaker, failing to recognize that there would have been many other branches at the time that were out-competed by those subclades, and are now extinct. Those three guys just happened to have advantageous genes, eventually resulting in flourishing populations expanding from three different places.

The earlier we go back in Bell Beaker, the more we should see P312 XL21 XDF27 XU152.
 
L21 or the just upstream clade were the dominant ones among Britain BB's.
I don't see too clearly what is your reasoning. I suppose that as a whole the other NORTHERN BB's were dominantly descendants of the same male ancestors, even if some of "brother" rare subclades(to be found!) were lost in the subsequent times. Being Y-R1b-P310 in a previous time doesn't excude that you could be the ancestros of the subsequent L21, DF27 and U152.
What seems true to me is that the clade L21 was more typical of the Irleand Britain BB's at Chalco/Early Bronze).
Besides, I don't believe too much in physical selection for these haplo's, rather in a political/cultural selection (sexually unbalanced mating).
 
“Besides, I don't believe too much in physical selection for these haplo's, rather in a political/cultural selection (sexually unbalanced mating”

By iron age there were already thousands of descendants of L21, competing against thousands of xL21, yet L21 continued to expand favorably. So how did the later L21s know who to allow to reproduce? Haplogroup tattoos to keep track of who was important?

“L21 or the just upstream clade were the dominant ones among Britain BB's.“

Going by dating range centers, we really only have ancient L21 DNA from later beaker in Britain. We don’t have DNA from the earliest centuries, because there weren’t that many of them yet so archaeologists are less likely to dig them up. It is during those earlier, more obscure centuries of British bell beaker that it became prominent.
 
Last edited:
"could be the ancestros of the subsequent L21, DF27 and U152."

Almost certainly NOT!

The paternal ancestor of L21 is one single man per generation. Archaeologists dig up a small random sample of a much larger population. The odds that you happen to get lucky and randomly put a shovel in the ground that hits that single man? Very, very low. If I pick a nearby graveyard, and go start digging randomly, it's very improbable that I'll hit my grandfather.

Most of the ancient DNA we find is from extinct lineages. We just don't observe the SNPs of their extinct branches, because the DNA tests aren't designed to look for them. When we find an ancient DNA sample in Britain, and the paternal haplogroup is reported as P312*, it doesn't mean his terminal SNP is P312. It means his actual, unknown haplogroup is P312->X->X->X->X…, which has no modern descendants.
 
Last edited:
“Besides, I don't believe too much in physical selection for these haplo's, rather in a political/cultural selection (sexually unbalanced mating”

By iron age there were already thousands of descendants of L21, competing against thousands of xL21, yet L21 continued to expand favorably. So how did the later L21s know who to allow to reproduce? Haplogroup tattoos to keep track of who was important?
That's basically incorrect, by Iron-Age the R-L21 expansion rate is negligible (the e-fold is of the order of ~0.001 per year), passed the ~2100 BCE expansion peak, the average expansion rate of R-L21 becomes negligible.

From Y-FULL v11.01 tree, R-L21 diversification rate e-fold (~number of newly generated clades per year and per already existing clades) :

R-L21-diversification-e-fold.png


Few descendant clades might have had a succesfull history, but then this is better explained by "random" (politically/culturally related in fact) selection among the existing population at this time. If the population is heavily R-L21, a future expansion is more likely to contain some R-L21 rather than another obscur lineages. However, post BB-expansion, the R-L21 familly as a whole stop beeing succesfull for its diffusion.

The main argument to say that haplo-diffusion is political/culturally related comes mainly from that, the mass diffusion of an haplogroup is usually a one/two-shot event/cultural horizon.
Some descendents might get involved in other cultural expansions, but passed a given cultural expansion the average diffusion rate drops significantly for haplogroup families, whereas non-diffusive haplogroups might suddently become massively diffusive (a good exemple is I1).
Which poorly fits with a natural selection driven diffusion rate.
Are haplogroup passing from badly-adapted to super-adapted (or the opposite) in a one single mutation event ? That would be very unlikely, the chaotic and short lived high diffusion rates of haplogroup are instead completely supporting that diffusion is not driven by natural selection but by political/cultural criteria.

Another argument against a natural selection based Y-DNA is the relatively constent fixing rate for Y-DNA. If Y-DNA was significantly affected by biological natural selection, the fixing rate would show huge time-dependances when migration or change of environnement occurs.
We don't observe that significantly (which make Y-DNA a descent "clock" when looking at the number of mutations). Contrary to that, we do observe huge variation of the mtDNA fixing rate during the Out-of-Africa event for exemple.

Your point here is nearly proven wrong by all the data we have ... to prove your natural selection Y-DNA diffusion claim, a serious work published in peer reviewed journal would be needed.
 
"could be the ancestros of the subsequent L21, DF27 and U152."

Almost certainly NOT!

The paternal ancestor of L21 is one single man per generation. Archaeologists dig up a small random sample of a much larger population. The odds that you happen to get lucky and randomly put a shovel in the ground that hits that single man? Very, very low. If I pick a nearby graveyard, and go start digging randomly, it's very improbable that I'll hit my grandfather.

Most of the ancient DNA we find is from extinct lineages. We just don't observe the SNPs of their extinct branches, because the DNA tests aren't designed to look for them. When we find an ancient DNA sample in Britain, and the paternal haplogroup is reported as P312*, it doesn't mean his terminal SNP is P312. It means his actual, unknown haplogroup is P312->X->X->X->X…, which has no modern descendants.
When I said "ancestors" I should have said: "mother population where the mutation occurs before devleopping - surely ALL P312 cannot be fathers of L21, DF27, U152!
I don' think ONE man suddenly migrated elsewhere, giving birth to thousands of sons in two or three generations! So it's a part of the P312 that moved, among them first L21 (maybe, or rather, surely, the mutated male gave birth immediatly to a lot of sons, bastards or not depending on social statute). You're correct if you say that the success of an haplogroup is often at the cost of other haplogroups, but here we speak nevertheless of very near "cousins" SNP's if not "brothers", so a very close population at autosome level, and culturaly for some of them at least.
you wrote: "The paternal ancestor of L21 is one single man per generation" !?! What meaning? ONE at the first generation but after? Surely ore than one... Or it could not have had some success in posterity.
 
That's basically incorrect, by Iron-Age the R-L21 expansion rate is negligible (the e-fold is of the order of ~0.001 per year), passed the ~2100 BCE expansion peak, the average expansion rate of R-L21 becomes negligible.

From Y-FULL v11.01 tree, R-L21 diversification rate e-fold (~number of newly generated clades per year and per already existing clades) :

R-L21-diversification-e-fold.png


Few descendant clades might have had a succesfull history, but then this is better explained by "random" (politically/culturally related in fact) selection among the existing population at this time. If the population is heavily R-L21, a future expansion is more likely to contain some R-L21 rather than another obscur lineages. However, post BB-expansion, the R-L21 familly as a whole stop beeing succesfull for its diffusion.

The main argument to say that haplo-diffusion is political/culturally related comes mainly from that, the mass diffusion of an haplogroup is usually a one/two-shot event/cultural horizon.
Some descendents might get involved in other cultural expansions, but passed a given cultural expansion the average diffusion rate drops significantly for haplogroup families, whereas non-diffusive haplogroups might suddently become massively diffusive (a good exemple is I1).
Which poorly fits with a natural selection driven diffusion rate.
Are haplogroup passing from badly-adapted to super-adapted (or the opposite) in a one single mutation event ? That would be very unlikely, the chaotic and short lived high diffusion rates of haplogroup are instead completely supporting that diffusion is not driven by natural selection but by political/cultural criteria.

Another argument against a natural selection based Y-DNA is the relatively constent fixing rate for Y-DNA. If Y-DNA was significantly affected by biological natural selection, the fixing rate would show huge time-dependances when migration or change of environnement occurs.
We don't observe that significantly (which make Y-DNA a descent "clock" when looking at the number of mutations). Contrary to that, we do observe huge variation of the mtDNA fixing rate during the Out-of-Africa event for exemple.

Your point here is nearly proven wrong by all the data we have ... to prove your natural selection Y-DNA diffusion claim, a serious work published in peer reviewed journal would be needed.
Very good point.
 
That's basically incorrect, by Iron-Age the R-L21 expansion rate is negligible (the e-fold is of the order of ~0.001 per year), passed the ~2100 BCE expansion peak, the average expansion rate of R-L21 becomes negligible.

From Y-FULL v11.01 tree, R-L21 diversification rate e-fold (~number of newly generated clades per year and per already existing clades) :

R-L21-diversification-e-fold.png


Few descendant clades might have had a succesfull history, but then this is better explained by "random" (politically/culturally related in fact) selection among the existing population at this time. If the population is heavily R-L21, a future expansion is more likely to contain some R-L21 rather than another obscur lineages. However, post BB-expansion, the R-L21 familly as a whole stop beeing succesfull for its diffusion.

The main argument to say that haplo-diffusion is political/culturally related comes mainly from that, the mass diffusion of an haplogroup is usually a one/two-shot event/cultural horizon.
Some descendents might get involved in other cultural expansions, but passed a given cultural expansion the average diffusion rate drops significantly for haplogroup families, whereas non-diffusive haplogroups might suddently become massively diffusive (a good exemple is I1).
Which poorly fits with a natural selection driven diffusion rate.
Are haplogroup passing from badly-adapted to super-adapted (or the opposite) in a one single mutation event ? That would be very unlikely, the chaotic and short lived high diffusion rates of haplogroup are instead completely supporting that diffusion is not driven by natural selection but by political/cultural criteria.

Another argument against a natural selection based Y-DNA is the relatively constent fixing rate for Y-DNA. If Y-DNA was significantly affected by biological natural selection, the fixing rate would show huge time-dependances when migration or change of environnement occurs.
We don't observe that significantly (which make Y-DNA a descent "clock" when looking at the number of mutations). Contrary to that, we do observe huge variation of the mtDNA fixing rate during the Out-of-Africa event for exemple.

Your point here is nearly proven wrong by all the data we have ... to prove your natural selection Y-DNA diffusion claim, a serious work published in peer reviewed journal would be needed.
You make some good arguments, but I think you misunderstand the point I'm making.

I'm not saying that L21 is exceptional among modern extant lineages. I'm saying that L21 is exceptional among ALL lineages, including all those that are now extinct.

Many extant lineages have a dominant initial burst, due to starting within a small effective population. So in terms of branch numbers, the initial expansion dominates over later expansion. It is naturally much easier for early random variation of the growth rate to have a significant impact than it is for later random variation, which occurs as a much smaller fraction of a much larger population.

But that doesn't mean that an exponential factor just slightly greater than 1 is insignificant.

An exponential factor of 1.001 sounds small, but sustained over a long period it is very significant, and very exceptional among a data set including both extant and extinct lineages. For example, supposing that, after 2000 BC, L21 had instead had an exponential factor of 0.999. Then, starting with around 270 branches in 2000 BC, after 4000 years it would be down to about 5 surviving branches. Change that rate to 0.990, and it would now be extinct.

The problem is that our data set is inherently biased. We mostly just have data for extant lineages, so we are comparing the best to the best. Most surviving lineages are the cream of the crop. So making comparisons among other extant lineages gives a biased perspective. The growth rate will appear average, when in fact it is exceptional.

For most advantageous genes, we should not expect an exponential factor dramatically greater than 1. A sustained value of 1.001 is probably fairly typical of a moderately advantageous mutation that comes to dominate over time.

And even L21's huge initial growth spurt is sustained over quite a long time, close to a millennium. I am not convinced that this can be explained through randomness or a political/cultural basis alone.

I really don't think it's necessary to publish such a theory, because I think most professionals already take these facts for granted. It is only within the "citizen science" community of the forums that this is so poorly understood. Most of that community is quite knowledgeable in history, archaeology, and so on, but quite dismal at math and science. This is why Dr. Flood's perspective is so confusing to the cliques that dominate online forums. Within the community, his grasp of the math is an exception rather than the norm, so his views are dramatically at odds with the opinions that dominate within the online community. You are the first person I've ever encountered on any forum who is even capable of making a sensible argument about such things.
 
Last edited:
When I said "ancestors" I should have said: "mother population where the mutation occurs before devleopping - surely ALL P312 cannot be fathers of L21, DF27, U152!
I don' think ONE man suddenly migrated elsewhere, giving birth to thousands of sons in two or three generations! So it's a part of the P312 that moved, among them first L21 (maybe, or rather, surely, the mutated male gave birth immediatly to a lot of sons, bastards or not depending on social statute). You're correct if you say that the success of an haplogroup is often at the cost of other haplogroups, but here we speak nevertheless of very near "cousins" SNP's if not "brothers", so a very close population at autosome level, and culturaly for some of them at least.
you wrote: "The paternal ancestor of L21 is one single man per generation" !?! What meaning? ONE at the first generation but after? Surely ore than one... Or it could not have had some success in posterity.
"
surely ALL P312 cannot be fathers of L21, DF27, U152!
I don' think ONE man suddenly migrated elsewhere, giving birth to thousands of sons in two or three generations! So it's a part of the P312 that moved, among them first L21 (maybe, or rather, surely, the mutated male gave birth immediately to a lot of sons, bastards or not depending on social statute). "

Then I strongly agree with you in this regard. This is precisely what I think much of the online community fails to understand. The misconception apparently expressed in the opening comment of this thread is that L21 experienced a later growth, so there must have been a later flow from a latent L21 population on the Continent:

"L21 in Ireland not from the Bronze Age?"

Many people have the idea that early P312 population was dominated exclusively by DF27, L21, and U152, and the reason L21 dominates the Isles in later times is that a herd of L21 migrated to the Isles. And any subsequent changes in percentage must be due to people coming and going between these populations.

But based on your most recent comment, you don't appear to be one of them, so I don't think our positions are at odds with each other.

L21 no doubt experienced a rapid expansion early on, to thousands within a small number of generations. And this is probably attributable to random variation within a small effective population, and founder effects, as Dr. Flood explains.

What's clear is that there WAS a single man or his small clan that migrated to the Isles (within a larger population), and that soon after arrival his lineage expanded rapidly within the Isles. We should not be surprised to see that it continued to expand at a diminished rate in later periods.

You seem to grasp this, yet it is apparent that many in the online community do not.

In one of the other forums, I see prominent members arguing that L21 could not have been a minority within a larger P312 population, or else we would see other extant lineages in the Isles adjacent to L21. This position is founded on the incorrect assumption that later dominant subclades must be the same ones that dominated early P312.

This is the notion that my original post was directed at. Perhaps the idea isn't pervasive among Eupedia. I am new to the community. But the view is obviously entrenched in other forums. I’m not trying to rewrite genetic studies or anything like that.
 
Last edited:
Then, WHEN did L21 become the majority among Celtic peoples?

I think that it may have to do with the Irish expansion during the Early Middle Ages
 
Then, WHEN did L21 become the majority among Celtic peoples?

I think that it may have to do with the Irish expansion during the Early Middle Ages
You should read Dr. Flood's publication on L21 phylogeny if you haven't already. Although dominated by the initial burst in early Isles Bell Beaker, expansion did occur in multiple waves of bottlenecks and growth.

The main point I was trying to make is that many people are apparently puzzled as to why 1) we don't see lots of L21 in early P312 before Isles expansion, and 2) we don't see a constant percentage of L21 following the initial expansion.

If one understands that extant L21 branches competed among other extinct P312 branches, these observations are not the least bit surprising or confusing.
 


Relevant publications. We know that Bell Beakers tend to be tall. It is also known that there are genes for height in yDNA. There is probably at least some small degree of natural selection occurring within Bell Beaker paternal DNA, although I don’t mean to suggest that success of particular lineages is entirely attributable to this.
 
Last edited:

Here’s another publication that claims a correlation between forensic (and consequently genealogical) STR counts and medically relevant gene expression, supporting that selection pressure is probably a significant factor for yDNA lineage survival.
 
Last edited:
It's just founder effects, nothing to do with selection. There's nothing special about L21 as opposed to the myriad other Y chromosome founder effects we've seen in global populations. It's just been over-analysed to death because Anglo-Americans + Irish-Americans are the largest testing cohort globally and they've been active since the beginning of DNA testing.

The Y DNA height papers are pseudo-scientific nonsense, populations with tall men also have tall women and they don't have a Y chromosome.
 
  • Like
Reactions: Syd
It's just founder effects, nothing to do with selection. There's nothing special about L21 as opposed to the myriad other Y chromosome founder effects we've seen in global populations. It's just been over-analysed to death because Anglo-Americans + Irish-Americans are the largest testing cohort globally and they've been active since the beginning of DNA testing.

The Y DNA height papers are pseudo-scientific nonsense, populations with tall men also have tall women and they don't have a Y chromosome.
already old surveys had shown that Y chromosome has more autosomal effects than believed previously but I agree that it isn't sufficient to make a pop extraordinary because a lot of autosomal traits are under multiloci influences, the big most of them out the Y chromosome.
 

This thread has been viewed 2685 times.

Back
Top