Archiver > GENEALOGY-DNA > 2006-05 > 1147000619

From: "John McEwan" <>
Subject: RE: [DNA] R1b page updated
Date: Sun, 7 May 2006 23:16:59 +1200
In-Reply-To: <>

Dear Ellen

You said
My point has been that R1b1c* is predominant today in
these population. I don't think the evidence is there
to support the contention that it was predominant in
the same region in the prehistoric past, or that it
was even present post-LGM (ie, pre-Neolithic) in this
same region. I not saying it was not there, only that
there is no evidence that it was either. Given the
ancient mtDNA evidence, it would appear that quite a
bit has changed genetically between prehistoric (even
late Bronze Age) and modern European populations, so
one should probably not assume that the Y haplogroups
have not suffered the same fate.

You are correct that we have no ancient Y DNA SNPs or profiles. We can
only infer history based on what we see in current populations. With
minor differences what we see in these populations agrees between mtDNA,
Y DNA and autosomal DNA. The current commonly accepted hypothesis is
based on SNP distributions and clines and their associated STR
diversity. It suggests R1b1c(xSRY2627,S21,S28,M222) was in Iberia during
the LGM and expanded northwards after that time and settled Britain and

You often refer to aMTDNA evidence as urging caution on this

However, whatever alternative proposal is put forward also has to
recreate the observed distribution and do so without contrived
complexity, be consistent with the linguistic and archaeological record,
plus the limits imposed by biology.

Could you please provide your hypothesis as to the last 10,000 years in
Europe and the British Isles and how the current observed Y
distributions and diversity arose?

Later you said
We discussed SRY2627 (R1b1c6) just a few weeks ago on
the list. Discussion of the origin and age of this
clade appears in McEvoy's "The Longue Duree" article.
He dates it to between 2000-3000 years ago. I posted
a question as to whether anyone had found evidence of
the presence sub-clade in the British Isles which
would support a more recent migration from Spain into
the British Isles.

McEvoy's Longue Duree article does mention SRY2627, to quote:

" It is difficult to distinguish
genetically between a common Paleolithic origin
and more recent contacts. However, haplogroup R1b3f
Y chromosomes, which have a recent origin in Iberia
(Hurles et al. 1999), have not been found in Ireland (Hill
et al. 2000), arguing against the migration of very large
numbers of men by this route, at least, in the past 2,000-
3,000 years. This would be consistent with the suggestion
that most contacts over this period would have been
small scale, rather in the manner of the Kula ring in the
western Pacific (Cunliffe 2001)."

For new readers of this list R1b3f = R1b1c6 = SRY2627+ = M167+

My April 20th e-mail provided figures and suggested perhaps 8% of the
current Irish population could have came from later migration than the
initial expansion after the LGM. It was based on Irish frequency of
SRY2627 from the more recent Moore et al 2006 paper, but the figures are
not that precise and it could be much lower. However, the SRY2627+
frequency presented for southern England by Rosser et al 2000 and also
from my R1b SNP webpage are not that different from the adjacent region
in coastal Europe and this is consistent with a potentially much higher
level of migration to southern England.

I am not a fan for wholesale replacement of the Irish population by
Iberian Celtic invaders ~100BC.

Please note McEvoy et al do not date SRY2627, they quote Hurles et al
and their estimate of date of origin (1650-3450 yrs bp) uses a mutation
rate of 2.1 x10-3 (range 0.6-4.9 x 10-3) and a 25 yr generation interval
provided from an earlier reference.

In my summary of M167/SRY2627 at I
explicitly comment on this estimate and provide other published evidence
of frequency and diversity of R1b1c6. I think Hurles et al. estimate is
rather recent, in part because of the generation interval and the
"effective" mutation rate used, but also in part because its diversity
relative to R1b1c.

Perhaps more importantly, we also have to account for biology, I cannot
see how a SRY2627 mutation 2000 years ago could realistically reach the
current frequency observed in Iberia. For example it appears that the
Spanish and Portugese now have ~2% SRY2627+. This region has a current
population of 50 Million people, therefore 500,000 million males are
SRY2627+. In fact the number is much larger than that, because it is
spread with appreciable frequency through France, parts of Germany and
the Atlantic Seaboard. I estimate perhaps 1 Million SRY2627 in total. A
mutation 2000 years ago in a single individual and a generation length
of 30 years would require more than a 23% growth in numbers per
generation. During that same time the population of France and Spain
increased perhaps 8 fold or 3% per generation.

I think it extremely unlikely that this mutation conferred a 20% benefit
in male reproductive success for the last 70 generations. Random chance
and genetic drift are also not an explanation.

To quote the robot in the "Lost In Space" TV programme of my youth:

"It does not compute" :-)

If you use a growth rate of 3%, the same as the balance of the
population, then the mutation would have had to have occurred 14,700 yrs
bp to increase to its present numbers!

Perhaps more explicitly a mutation origin ~2000 years ago would have
been at such low frequency at that time that just about the entire male
population could invade Ireland during the Celtic "invasions" and there
would be no trace in the current Irish population, because literally the
handful of SRY2627 all remained in Spain. In fact viewed critically both
your and McEvoy et al comments cannot be sustained by Hurles et al
estimate of the age of SRY2627. It is dangerous to speculate, but upon
rereading the McEvoy et al statement and given the Irish group have also
recently used 0.0007 as a mutation rate estimate I suspect the authors
have "delicately phrased" this paragraph.

However, I do not want to restart the mutation rate thread yet again. I
simply provide these numbers to get people to reflect on what they are
implying when they suggest impossibly young estimates for age of R1b

In summary, I feel SRY2627 frequency in Ireland is not consistent with
large scale migration in the last several thousand years between Spain
and Ireland, but this also depends on a SRY2627 being much older than
Hurles et al estimate. The case for southern England and adjacent
regions of Europe is less clear cut.


John McEwan

This thread: