November 12, 2010

Y chromosomes in Brabant

From the paper:
The Duchy of Brabant was a historical region in the Low Countries between the 12th and 18th century and consisted of a present-day Dutch province and three contemporary Belgian provinces together with the Brussels-Capital Region. The total area is 14.425km2 with approximately 150 km between the two most remote places in Brabant. The main reason for selecting this region was the ability to obtain reliable genealogical data of the patrilineal line for each of the numerous donors living together on a small geographical scale.

The authors typed 37 Y-STRs, and 103 Y-SNPs. They write:
All individuals were correctly assigned to the main haplogroups using the Whit Atheys’ Haplogroup Predictor. In total, eight main haplogroups were observed with almost 85% of the samples belonging to haplogroup R(63%) and I(21%)(Table 1). On the lowest observed level of the phylogenetic tree 32subhaplogroups were found in the dataset, whereby nearly 70% of all samples belonged to only four subhaplogroups: R1b1b2a1(R-U106), R1b1b2a2* (R-P312*), R1b1b2a2g(R-U152) andI1*(I-M253*)

They found star-patterns in all their subhaplogroups, but uncovered some structure in their J2a* (J-M410*) chromosomes. Youngest expansion ages "were observed for E1b1b1a2(E-V13) and I1*(I-M253*), respectively 4182–5855 and 4531–6344 years ago."

a strong downward trend in the frequency of haplogroup R was observed from North to South (Table 1; Fig. S5). The difference in the frequency of R haplogroups was circa 10% between the most northern and southern part, mainly due to the downward frequency of R1b1b2a1(R-U106).
The European-wide distribution of R-U106 suggests to me that it was a Germanic lineage.

Moreover, it was even possible to detect further substructuring within subha-
plogroup J2a*(J-M410*)based on the network analysis of all single-allele Y-STR haplotypes. Nevertheless, it was remarkable that the network analyses could not differentiate all observed subhaplogroups within R1b1b2(R-M269) and I2b(I-M223). This might be due to the relatively young age of these specific subhaplogroups making it impossible to differentiate these groups based on the Y-STRs.

The extraordinary success of these subhaplogroups is one of the most interesting questions: natural selection, or demographic dominance of a recently formed population group storming Western Europe by force of numbers? Ancient Y-DNA urgently needed...

The occurrence of haplogroup Q1 in 2.6% at Kempen, and 1.59% at Mechelen is an oddity of the findings that might merit further study.

In short this might be called a "model study" of Y-chromosome variation, due to the large number of individuals (477) and markers tested.

Forensic Sci Int Genet. 2010 Oct 29. [Epub ahead of print]

Micro-geographic distribution of Y-chromosomal variation in the central-western European region Brabant.

Larmuseau MH, Vanderheyden N, Jacobs M, Coomans M, Larno L, Decorte R.


One of the future issues in the forensic application of the haploid Y-chromosome (Y-chr) is surveying the distribution of the Y-chr variation on a micro-geographical scale. Studies on such a scale require observing Y-chr variation on a high resolution, high sampling efforts and reliable genealogical data of all DNA-donors. In the current study we optimised this framework by surveying the micro-geographical distribution of the Y-chr variation in the central-western European region named Brabant. The Duchy of Brabant was a historical region in the Low Countries containing three contemporary Belgian provinces and one Dutch province (Noord-Brabant). 477 males from five a priori defined regions within Brabant were selected based on their genealogical ancestry (known pedigree at least before 1800). The Y-haplotypes were determined based on 37 Y-STR loci and the finest possible level of substructuring was defined according to the latest published Y-chr phylogenetic tree. In total, eight Y-haplogroups and 32 different subhaplogroups were observed, whereby 70% of all participants belonged to only four subhaplogroups: R1b1b2a1 (R-U106), R1b1b2a2* (R-P312*), R1b1b2a2g (R-U152) and I1* (I-M253*). Significant micro-geographical differentiation within Brabant was detected between the Dutch (Noord-Brabant) vs. the Flemish regions based on the differences in (sub)haplogroup frequencies but not based on Y-STR variation within the main subhaplogroups. A clear gradient was found with higher frequencies of R1b1b2 (R-M269) chromosomes in the northern vs. southern regions, mainly related to a trend in the frequency of R1b1b2a1 (R-U106).



Andrew Oh-Willeke said...

Given the quality of the genelogical records, it ought to be possible to estimate ancient DNA proportions several hundreds of years back in time with reasonable accuracy and a well chosen algorhythm to estimate missing data based on neighbor frequencies, etc.

Andrew Oh-Willeke said...

Your comments on R-U106 are worth restating here:

"As noted in the other recent paper, and shown in the above Figure from the current one, R-U106 peaks in northern Europe. Its frequency (including the R-U198 sublineage) is 36.8% in the Netherlands, 20.9% in Germany and Austria, 18.2% in Denmark, 18.2% in England, 12.6% in Switzerland, 7.5% in France, 6.1% in Ireland, 5.9% in Poland, 5.6% in north Italy 4.4% in Czech Republic and Slovakia, 3.5% in Hungary, 4.8% in Estonia, 4.3% in south Sweden, 2.5% in Spain and Portugal, 1.3% in eastern Slavs, 0.8% in south Italy, 0.6% in Balkan Slavs, 0.5% in Greeks (i.e. 2 of 193 Cretans, and no mainland Greeks), 0.4% in Turks, 0% in Middle East.

The age of R-U106 is estimated by the authors as 8.7ky BP, which translates to about 2.5ky BP with the germline rate. The existence of R-U106 as a major lineage within the Germanic group is self-evident, as Germanic populations have a higher frequency against all their neighbors (Romance, Irish, Slavs, Finns). Indeed, highest frequencies are attained in the Germanic countries, followed by countries where Germanic speakers are known to have settled in large numbers but to have ultimately been absorbed or fled (such as Ireland, north Italy, and the lands of the Austro-Hungarian empire). South Italy, the Balkans, and West Asia are areas of the world where no Germanic settlement of any importance is attested, and correspondingly R-U106 shrinks to near-zero."

Your suggestion would imply that Germanic expansion ca. 500 BCE has provided about 20% of the Germanic gene pool, twice as much in the Netherlands. This is not a full population replacement, but this could be enough to account, for example, for the differences in mtDNA H frequency found today in Germany and that found in the LBK era. (One would need a different explanation to explain the demise of N1a).

The 8.7 ky BP date would be, within the margin of error, the arrival of the Neolithic in these areas. This can't be ruled out by the three Y-DNA we have from then, but can hardly be established very definitively either, since there is no Y-DNA hg R of any kind that old in aDNA yet.

Average Joe said...

demographic dominance of a recently formed population group storming Western Europe by force of numbers?

It is beginning to look like genetic research is validating the truth behind the old Aryan theory of European colonization.

AWood said...

This one hits close to home as I've yet to find any haplotype closer than 16 gd at 67 markers - in particular because P312* was highlighted rather than L21 (perhaps an oversight, or Belgium truly is a P312* hotspot).
Would love to see the haplotypes but don't feel like dropping $40 USD for electronic media. Should be a good paper....

Kepler said...

Interesting. Now, take a look at this:

Could this influx not be part of a more recent phenomenon? One starting with the Germanic groups settling in the borders of the Roman empire?

Fanty said...


The population estaminates on those ages, that I know of, are quiet low. Specially for the Germanics.

Those estaminate a total number of 500.000 - 1 Million Germanic people in continental Germania and Scandinavia combinded, during the 1th century AD. And a population of the Roman Empire of 60 Million humans.

With a Germanic population of 2.5 Million at the time, the Germanix tribes push into the Roman Empire.

More of those estaminates where:

10.000-20.000 Angles and Saxons impact on 250.000 Britons.

100.000 Franks impact on 5 Million citicens of "Gallia".

200.000 Goth on 9 Million Spanish.

Its claimed, the fall of Rome was not by Germanic numbers. But because

1. Rome heavyly decreased the money for defense, compared to the 1th. Century (at that time, Rome spend 80% of the Empires GDP on defence. Modern European states spend 1%. The USA spends 4%)

2. Roman citicens buyed their sons out of military service.
About 50% of the Roman Army at the time of its fall, had been Germanic migrants. With Germanic migrants as beeing the only people "at arms". And those migrants not even fully accepted "citicens", but, while serving in Romes army, pledging loyality to their "King" (who settles on Roman turf with all his subjects...) its was quiet easy to overthrow the Power and establish themselfs as the ruling class in Rome.

No superior numbers needed.

Jack said...

More than 2% haplogroup Q1? What the...?
Something happned in the lab we should know about?

Anonymous said...

certainly the Franks had an important influence (Franchi Salii), but we must not forget that the region, at least in the North, was already inhabited by Germanic peoples, like the Batavians and Cananefates.
Cesare already described the Belgians as mostly Germans and the northern part of Brabant was in the province of Germania Inferior and not in Gallia Comata.

Jan said...


More than 2% haplogroup Q1? What the...?
Something happned in the lab we should know about?[/q]

2.6% = 2 persons of 77 in one area
1.59% = 1 person of 63 in another area

To sum up: 3 of 477 sample size, "white noise", irrelevant for this study.

Also, defining mutation for Q1a or Q1b were not tested. I may only speculate that those are Q1b of Ashkenazi origin or Q1a of Sephardic.
Unfortunately, haplotypes are not freely available to compare with commercial tested ones from open databases.

eurologist said...

Large regions west of the Rhine were Germanic, and some where mixed Germanic/Gallic long before the Romans arrived.

Initially, the Roman terminology of the Roman provinces of Germania Inferior and Superior reflects this. However, later, especially under Caesar, for political reasons the Germanic tribes often where only identified with the "bad Barbarians" on the east side of the Rhine. However, the people on the east side never lost their connection to and alliance with those on the west, and used the weakness of the late Roman empire to reunite.

Likewise, Germanic peoples also likely were living in at least parts of Noricum, but again, the late Romans only called the tribes outside their consolidated boarder Germanic (e.g., on the north bank of the Inn river in the Alps).

"Mass migrations" are usually hugely exaggerated: Germanic people lived in those regions (Netherlands/Frisia, part of Belgium, left bank of the middle and upper Rhine, Alsace/ Black Forest, and the northern part of Raetia and Noricum) all along; often, leadership changed or some people settled in less populated areas, but more often than not, there was population continuity.

In the north, the story of the Batavi provides a good example, in the south, that of the Alemanni.

And yes, the fact that many of the late Roman armies where provided by Germanic tribes helped consolidate power - but again, that does not mean mass migrations.

eurologist said...

"More than 2% haplogroup Q1? What the...?
Something happned in the lab we should know about?"

Supposedly, Norway is 4% Q - it isn't really all that uncommon in parts of Europe.

Perhaps the Bromme and peri-glacial Hamburgian and Ahrensburg cultures already had some Q... would definitely be interesting to have a higher-resolution tree for this. I know it's more difficult for the small haplogroups - but they seem to get neglected compared to R1a and R1b, for example.