The Coming of the Anglo-Saxons to Britain

[Angela]

Another new paper on British genomics:
Iron Age and Anglo-Saxon genomes from East England reveal British migration history

http://www.nature.com/ncomms/2016/16...omms10408.html

"British population history has been shaped by a series of immigrations, including the early Anglo-Saxon migrations after 400 CE. It remains an open question how these events affected the genetic composition of the current British population. Here, we present whole-genome sequences from 10 individuals excavated close to Cambridge in the East of England, ranging from the late Iron Age to the middle Anglo-Saxon period. By analysing shared rare variants with hundreds of modern samples from Britain and Europe, we estimate that on average the contemporary East English population derives 38% of its ancestry from Anglo-Saxon migrations. We gain further insight with a new method, rarecoal, which infers population history and identifies fine-scale genetic ancestry from rare variants. Using rarecoal we find that the Anglo-Saxon samples are closely related to modern Dutch and Danish populations, while the Iron Age samples share ancestors with multiple Northern European populations including Britain."

This is the paper and the samples that resulted in quite a kerfuffle when we tried to analyze them.

They give an estimate of 38% for Anglo-Saxon admixture into eastern England. Leslie et al gave 35%, yes? So, it's virtually the same.

For more nuance:
"By this measure the East England samples are consistent with 38% Anglo-Saxon ancestry on average, with a large spread from 25 to 50%, and the Welsh and Scottish samples are consistent with 30% Anglo-Saxon ancestry on average, again with a large spread (Supplementary Table 4). These numbers are lower on average if we exclude the low-coverage individual HS3 from the Anglo-Saxon group (35% for East English samples). A similar result is obtained when we analyse modern British samples from the 1,000 Genomes Project, which exhibit a strong substructure (Supplementary Note 4, Supplementary Fig. 4). We find that samples from Kent show a similar Anglo-Saxon component of 37% when compared against Finnish and Spanish outgroups, with a lower value for samples from Cornwall (Supplementary Fig. 5a, Supplementary Table 4)."

"An alternative and potentially more direct approach to estimate these fractions is to measure rare allele sharing directly between the modern British and the ancient samples. While being much noisier than the analysis using Dutch and Spanish outgroups, this yields consistent results (Supplementary Fig. 5b, Supplementary Note 3). In summary, this analysis suggests that on average 25–40% of the ancestry of modern Britons was contributed by Anglo-Saxon immigrants, with the higher number in East England closer to the immigrant source. The difference between groups within Britain is surprisingly small compared with the large differences seen in the ancient samples. This is true for both the UK10K samples and for the British samples from the 1,000 Genomes project, although we note that the UK10K sample locations may not fully reflect historical geographical population structure because of recent population mixing."


"There are striking differences in the sharing patterns of the samples, illustrated by the ratio of the number of rare alleles shared with Dutch individuals to the number shared with Spanish individuals (Fig. 2a). The middle Anglo-Saxon samples from Hinxton (HS1, HS2 and HS3) share relatively more rare variants with modern Dutch than the Iron Age samples from Hinxton (HI1 and HI2) and Linton (L). The early Anglo-Saxon samples from Oakington are more diverse with O1 and O2 being closer to the middle Anglo-Saxon samples, O4 exhibiting the same pattern as the Iron Age samples, and O3 showing an intermediate level of allele sharing, suggesting mixed ancestry."

"The ancient samples fall within the range of modern English and Scottish samples, with the Iron Age samples from Hinxton and Linton falling closer to modern English and French samples, whereas most Anglo-Saxon era samples are closer to modern Scottish and Norwegian samples. Overall, though, population genetic differences between these samples at common alleles are small."

[Angela]

Some broader conclusions for European population history in general:

"The common feature in all three trees is a first split between Southern and Northern Europe with a median time ~7,000 years ago, followed by three more separations close in time ~5,000 years ago between Netherlands, Denmark, Finland and Britain."

Well, 5,000 years ago is 3000 BC, so nice correlation between that and the movement from the East, but what to make of the split between Southern and Northern Europe 5000 BC?

There's also this:

"The relatively recent estimate for the split time between Italy and Spain, ~2,600 years ago, may be a consequence of migration following an earlier separation; the population size of the Italian-Spanish ancestral population was estimated to be extremely large and an upper bound could not be determined, which could be an artifact of ancestral substructure or admixture. Another explanation would be a common source of admixture into both the Spanish and the Italian population, resulting in relatively recent common ancestry."

Here we go again. The programs never seem to produce dates or admixtures that totally work for Southern Europe. 2600 years ago is 600 BC. What common source of migration? In 600 BC northern and central Italy were being hit by the "Celtic" migrations. Does that apply to Spain as well? There was Greek migration to southern Italy but no such large colonization in Spain to my knowledge.

Maybe large population sizes mean just that...large population sizes. That might explain a lot of the difference between what happened in Southern versus Northern Europe.

That 2600 hundred years ago date does square with Ralph and Coop, who give a date of 2500 years ago as the time after which neither Spain nor Italy experienced really significant gene flow from outside.
http://journals.plos.org/plosbiology...l.pbio.1001555

[Vukodav]

Anglo Saxon ancestry is 30% in Scotland and Wales and 38% in East Anglia. IMHO the only explainatiom for these strange results is that the "Anglo Saxon" remains were already mixed with native Celts.

[Maciamo]

By analysing shared rare variants with hundreds of modern samples from Britain and Europe, we estimate that on average the contemporary East English population derives 38% of its ancestry from Anglo-Saxon migrations.

In my Genetic history of the British and the Irish, I estimated that today most English counties have between 55% and 65% of Germanic Y-DNA haplogroups, while Cornwall has 45% and Wales has much as 25%. Obviously, Germanic does not mean only Anglo-Saxon, but also Viking, Norman and even Frankish (through immigration from the continent + descendants of English kings from the Houses of Anjou and Plantagenet). Then, Y-DNA does not equal whole genomic inheritance. It's obvious that imported monarchs like the Normans tend to have much more children than average, passing their Y-DNA unchanged, but diluting their autosomes with local women at each generation. So overall I am inclined to agree with a figure of 35-38% of Anglo-Saxon genes in modern English people. This may represent about 45% of Anglo-Saxon Y-DNA - the remaining 15% of Germanic Y-DNA being either Viking or Norman (Frankish Y-DNA is probably very low, around 1%).

[Fire Haired14]

The Anglo-Saxon estimates shouldn't be taken literally. Anglo Saxons and Britons are basically identical when plotted into pan-European genetic diversity. The authors knew this so they used a bunch of analysis which look at rare-alleles, because rare-alleles track recent ancestry where Anglo Saxons and Britons were different. Anglo Saxons share more recent ancestry with Dutch/Danish and also Finnish than Britons did.

The results they got using many different methods make a lot of sense. East England in every analysis shares more rare alleles with Dutch/Danish/Finnish/Anglo Saxon and certainly do have more Anglo Saxon or at least Germanic-ancestry than Wales/Scotland/Cornwall. We should consider the 35-40% estimates for East England and 25-30% for Scotland/Wales/Cornwall, but that might not be very true because it is IMPOSSIBLE to estimate Anglo Saxon/Briton ancestry using DNA.

All I take from this paper is: All modern British have Anglo Saxon ancestry, East England has the most Anglo Saxon ancestry, British are something close to 100% North European(Briton, German, Unetice, East Bell Beaker, etc) but it's hard to track which North Europeans they descend from(How much is native Celtic-Briton and Gealic vs German, etc). There could also be Italian, Gaulish/French, Scandinavian, and SouthWest Asian ancestry in Britain. My guess is all British are mostly from Celtic-Britons/Picts/Gealics/whoever else.

[Fire Haired14]

Spain and Italy(Tuscany) coming from the same population just 2600yo is impossible. This could mean there's some recent common ancestry though.

[Greying Wanderer]

Some broader conclusions for European population history in general:

"The common feature in all three trees is a first split between Southern and Northern Europe with a median time ~7,000 years ago, followed by three more separations close in time ~5,000 years ago between Netherlands, Denmark, Finland and Britain."

Well, 5,000 years ago is 3000 BC, so nice correlation between that and the movement from the East, but what to make of the split between Southern and Northern Europe 5000 BC?

There's also this:

"The relatively recent estimate for the split time between Italy and Spain, ~2,600 years ago, may be a consequence of migration following an earlier separation; the population size of the Italian-Spanish ancestral population was estimated to be extremely large and an upper bound could not be determined, which could be an artifact of ancestral substructure or admixture. Another explanation would be a common source of admixture into both the Spanish and the Italian population, resulting in relatively recent common ancestry."

Here we go again. The programs never seem to produce dates or admixtures that totally work for Southern Europe. 2600 years ago is 600 BC. What common source of migration? In 600 BC northern and central Italy were being hit by the "Celtic" migrations. Does that apply to Spain as well? There was Greek migration to southern Italy but no such large colonization in Spain to my knowledge.

Maybe large population sizes mean just that...large population sizes. That might explain a lot of the difference between what happened in Southern versus Northern Europe.

That 2600 hundred years ago date does square with Ralph and Coop, who give a date of 2500 years ago as the time after which neither Spain nor Italy experienced really significant gene flow from outside.
http://journals.plos.org/plosbiology...l.pbio.1001555

dunno about the first one but 2600 BP is around the right time for the Celtic thing

http://historyofenglishpodcast.com/1...century-BC.jpg

[Northener]

@Angela "The middle Anglo-Saxon samples from Hinxton (HS1, HS2 and HS3) share relatively more rare variants with modern Dutch than the Iron Age samples from Hinxton (HI1 and HI2) and Linton (L)." Yes it were the same kind of Angles, Saxons and Jutes who filled the gap on the so called Frisians terp which were left behind after the severe population. Some autors state (not without phatos) "you could only hear the seagulls cry" ;)

[RobertColumbia]

In my Genetic history of the British and the Irish, I estimated that today most English counties have between 55% and 65% of Germanic Y-DNA haplogroups, while Cornwall has 45% and Wales has much as 25%. Obviously, Germanic does not mean only Anglo-Saxon, but also Viking, Norman and even Frankish (through immigration from the continent + descendants of English kings from the Houses of Anjou and Plantagenet)....overall I am inclined to agree with a figure of 35-38% of Anglo-Saxon genes in modern English people. This may represent about 45% of Anglo-Saxon Y-DNA - the remaining 15% of Germanic Y-DNA being either Viking or Norman (Frankish Y-DNA is probably very low, around 1%).

Good point. There have been a few recent threads on Eupedia from people asking something close to the same question - "I tested positive for <typically Germanic haplogroup such as I1 or R-U106>. What tribe am I from?". The answer, at least today, is "We don't know".

[MOESAN]

Some broader conclusions for European population history in general:

"The common feature in all three trees is a first split between Southern and Northern Europe with a median time ~7,000 years ago, followed by three more separations close in time ~5,000 years ago between Netherlands, Denmark, Finland and Britain."

Well, 5,000 years ago is 3000 BC, so nice correlation between that and the movement from the East, but what to make of the split between Southern and Northern Europe 5000 BC?

There's also this:

"The relatively recent estimate for the split time between Italy and Spain, ~2,600 years ago, may be a consequence of migration following an earlier separation; the population size of the Italian-Spanish ancestral population was estimated to be extremely large and an upper bound could not be determined, which could be an artifact of ancestral substructure or admixture. Another explanation would be a common source of admixture into both the Spanish and the Italian population, resulting in relatively recent common ancestry."

Here we go again. The programs never seem to produce dates or admixtures that totally work for Southern Europe. 2600 years ago is 600 BC. What common source of migration? In 600 BC northern and central Italy were being hit by the "Celtic" migrations. Does that apply to Spain as well? There was Greek migration to southern Italy but no such large colonization in Spain to my knowledge.

Maybe large population sizes mean just that...large population sizes. That might explain a lot of the difference between what happened in Southern versus Northern Europe.

That 2600 hundred years ago date does square with Ralph and Coop, who give a date of 2500 years ago as the time after which neither Spain nor Italy experienced really significant gene flow from outside.
http://journals.plos.org/plosbiology...l.pbio.1001555

Surprising conclusions of the study (for me at least). My feelings are that in fact stronger difference exists between Welsh people and Eastern England people, even with towns so "OGM"Welshmen; I don't rely completely upon these auDNA analysis when coming to precise distribution of ancestry. Sure I'm not amazed that today differences are less marked than in the V°Century. I 'll come back to this if I 'can read and understand well the whole concerned paper.
Concerning Italy and Spain and these 600 BC common imputs: apparently Belgae tribes, complete or not, and even a little bunch of Germanic tribes neighbouring these Celts, descended to Iberia until Southern Portugal. For I believe I know, genuine celtic Belgae came from Bavaria-Czechia surroundings and were eastern continental Celts a bit different from earlier Atlantic Celts; they were cousins of the Boians by example, for I can guess. Some Celts in Italy came from the Boians country if I don't mistake. So???
The problem is the relative lack of Y-R1b-U152 in Iberia except some rare "pale" hotspots. But traditions said too some of the Northern Italy Celts were come fromWest (so: Southern France): no solution here again concerning Y-haplos!

&: tribes from North Gallia?Belgia went to Iberia:
sets more than full tribes:
Boians (from Bohemia) >> Gallia >> NW Iberia
Volques (Hessen) >> Gallia >> North Tage, E Castile
Bellovaques (Belgia) >> Meseta in Spain
Veliocasses >> Cantabricas
Sefes/Saefs >> Numantia Soria NE SPain
Pelendones (?) from Colonia/Köl >> Numantia
Cempsi (Germanics?) >> S Portugal S Spain
Eburones (half Celts hals Germanics?) Belgia >> S Portugal Evora
Suessiones, Belgia >> Alava, Navarra N Spain

[MOESAN]

I wonder if rare alleles is a so reliable method: they could be submitted to more drift than others and can also on another side represent sometimes elite heritage which doesn't reflect the basic overruned population?

[Angela]

I wonder if rare alleles is a so reliable method: they could be submitted to more drift than others and can also on another side represent sometimes elite heritage which doesn't reflect the basic overruned population?

Exactly my thinking. I don't understand why they don't think that's a problem. As to your post about the migrations 600 BC into Italy and Spain, the problem is indeed the yDna, unless the different tribes were still more or less one yDna clade although autosomally they were different.

In terms of Italy, you mentioned the Boii. I don't think there's agreement on their origin.
https://en.wikipedia.org/wiki/Boii#S...in_north_Italy

These are some of the other "Gallic" or "Celtic" groups.
Lingones:
https://en.wikipedia.org/wiki/Lingones

Sennones:
https://en.wikipedia.org/wiki/Senones

Cenomani:
https://en.wikipedia.org/wiki/Cenomani

(The French and Italian Wiki articles are more comprehensive.)

If the Cenomani did enter Italy from generally southern France but originated further north, might that explain the L21+ we find in western northern Italy?

Just in order to visualize it, this is a pretty good map. The only group they miss is the Anares who were around Parma. Also, they don't show the movement south into the Appennini. Of course, it used to be dogma that after being pacified, they left Italy. That doesn't seem to be the case, although I want to see samples tested pre and post these invasions. It's important to know the yDna as well as the autosomal dna. If these men were L21 then they didn't have much impact. If they were some form of U152, then they had a great deal of impact.



This is another map. It shows which Celtic tribes migrated into Liguria to mix with the Ligures.