Variations of R1b Ydna in Europe: Distribution and Origins.
A.A.Foster. 13 March, 2005.

Based on the  differences and diversity of the alleles of R1b's DYS390 locus, there is evidence that there are four regional variants of the R1b sub-haplogroup in Europe.    These are:
(i) Baltic-Russian. (ii) NorthSea-Baltic. (iii) Alpine-South German. (iv) Atlantic. 
In Central and Western Europe, north of its great mountain ranges - The Pyrenees, Alps, and others - the major rivers flow northwest and northwards to the Atlantic, the North Sea and the Baltic. Only the Danube, which flows eastwards from the Northern Alpine regions to the Black sea, follows a different pattern.  Extrapolating from data available within the online "YHRD"  database (?, see below)  suggested all variants of R1b in Europe, as pre-historic hunter-gatherers, entered Europe from the east, and migrated and expanded along rivers and  coastlines, and across the ridgeways of high ground, eventually to reach the Baltic, North Sea, Mediterranean and Atlantic coastlines.
The mean frequency for DYS390=24 within the whole of the "YHRD" European database is about 59% of the R1b DYS390 population. In Iberia and France, and in the more remote areas of the British Isles, it averages almost 70% and reaches 80%. But in the Baltic regions the frequency is consistently low: it averages only 33% throughout the Baltic States, about  43% in the Netherlands, and 47%  in Baltic Germany. The lowest European percentage (29%) is to be found in Moscow, Russia. An even lower frequency, of 22%, can be found in  Asian Khazakstan. 
Complete R1b data from the "YHRD" database, indicated that, after an earlier existence in Asian Khazakstan, all European variants of R1b shared an existence in Russia ( in the region of Kazan, on the Volga river at about 55° North and 50° East), and that, later  they separated and expanded  into two major migrations ( a westward  migration to the Russian-Baltic region, and a south-western migration to the Black Sea area and then further, westwards, to the Alpine-South German region). Eventually, a North Sea-Baltic migration evolved from the  Russian-Baltic expansion; and  an Atlantic migration  evolved from the  Alpine-South German variant.
Baltic--Russian R1b:
Research showed that the greatest diversity of  R1b's DYS 390 locus is within the Russian-Baltic region. The data suggested that the Russian-Baltic variant migrated/expanded from the Kazan region of Russia westwards to Moscow, and then to the Baltic States of  Finland, Estonia, Lithuania, Latvia & Poland.
In this Baltic-Russian area, a sample of  159 haplotypes showed the R1b DYS390 percentages to be:

DYS 390=25.  28.9%;
DYS 390=24.  32.7%;
DYS 390=23.  32.1%;
DYS 390=22.    3.1%
Diversity:        68.6% (?)

North Sea-Baltic R1b:
Within the North Sea-Baltic area (Northern  Germany, Denmark, Netherlands and Norway) a sample of 1,227 haplotypes showed the R1b DYS390 percentages to be:   
DYS 390=25... 10.1%.
DYS 390=24... 46.6%.
DYS 390=23... 38.1%.
DYS 390=22...   3.7%.

Diversity:        61.5% (?)
These percentages  were less diverse than in the Russian-Baltic area - supporting the likelihood of  an R1b migration/expansion from east to west along the Baltic coast.  The coastal parts of the North Sea-Baltic region had more R1b diversity than  in  Norway and in the (German) Elbe river cities, indicating a further migration - from "Greater Frisia"(?) northwards to Norway and southwards into the Saxon lands alongside the  Elbe.(4)   Ultimately, North Sea-Baltic R1bs invaded  England and other parts of British Isles during the period 450 to 1,000AD as part of the Germanic-speaking Anglo-Saxon and Danish Viking invasion forces.
Alpine-South German R1b:
Analysis of the Yhrd data for this region indicates a migration/expansion path from Kiev (Ukraine - Russia), westwards along the River Danube (2,850 km), and north/westwards along the Rhine (1,320 km) to the North Sea. Politically, this whole region includes today's Ukraine,  Romania,  Hungary, Austria and Switzerland, Rhineland Germany, and Southern Holland. Except for the exception given below, a sample of 1,296 haplotypes revealed the R1b percentages for this region were uniform at:

DYS 390=25...   8.3%.
DYS 390=24... 57.9%.
DYS 390=23... 30.1%.
DYS 390=22...  2.6%.
Diversity:       55.7% (?)

With the Alpine-South German group, a small sub-sample of 122 haplotypes in the eastern Danube area, showed only 53% DYS390 =24, and 13% for DYS 390=25. This higher diversity supported the notion of a migration path of the Alpine-South German group from the east, and its heightened DYS390=25 in the eastern Danube area suggests that this variant may well have split  from the Russian-Baltic variant near to its source in central Russia.

Atlantic R1b:
This variant is found on the Atlantic coast, in Iberia, France and in the more remote parts of  Ireland and Scotland. In order to obtain more accurate data on the
 aboriginal/indigenous Scots/Irish, data was extracted from Capelli et al, (5) for Pitlochry and Oban in the Scots Highlands, and from Castlereigh in Central Ireland.

In the Atlantic region, R1b's DYS=390  showed the least diversity.  A sample of 1,516 haplotypes showed its R1b's DYS390 percentages to be:
DYS 390=25... 10.4%.
DYS 390=24... 69.7%.
DYS 390=23... 17.8%.
DYS 390=22...   1.1%.
Diversity:        46.1% (?)

The origin of this sub-population is unclear, but its lack of DYS390 diversity makes it the "youngest" R1b in  Europe. Some data suggested that it may have  split from the Alpine-South German variant in the region of Albania, and  then subsequently expanded, westwards, along the Mediterranean coast to  Iberia.

The YHRD  R1b sub-populations were identified by carrying out a geographic search  based on the selection of DYS 392=13, within Europe.  A repeat geographical selection, selecting DYS392=13 and combining it, alternatively, with DYS390=25,24,23 & 22 revealed the frequencies of each of DYS390's alleles. After analysis, these were aggregated into the four variant groups. The frequencies of DYS390=26 and 21 were so low that they could be ignored as being statistically irrelevant to this study.

The age of R1b?  
If the allele DYS390=24 was the original modal value for all four R1b variants, then the Russian-Baltic group has been mutating either at least  twice-as-long or twice-as-fast as the Atlantic one. About 30% of the  Atlantic group's DYS390 does not have an allele of 24, while within the Russian-Baltic group this figure increases to 68%.  Perhaps the Russian-Baltic variant never did have the well-known Atlantic Modal Haplotype where DYS390=24. But in either case, the difference in allele frquencies highlights that the populations are not homogenous. Differing alleles at the same locus position can be measured to show how diverse is the locus, and such increased locus diversity is a sign of a population's increased age (since its foundation or since it was isolated with a reduced amount of genetic diversity).
A simple application of the different levels of diversity of the four variants to the known archaeology of the Atlantic countries suggests that the ages of the variants, since separation from an earlier, parent type,  may, approximately,  be as follows:
Atlantic group c.14-18,000 ybp;
Alpine-South German  c.18,000 - 22,000 ybp;
NorthSea-Baltic c. 21,000- 25,000 ybp, and
Russian-Baltic possibly c. 24,000 - 28,000 ybp.
More work needs to be done on this aspect, and on the question of where the variant R1bs may have existed during the Last Glacial Maximum.

*1*  The YHRD database can be found at www.yhrd.org  and it is maintained by the Institute of Legal Medicine, Charite' - University Medicine Berlin.
*2*  Diversity has been calculated  using Simpson's Index of Diversity, 1-D, expressed as a percentage where 100% represents complete diversity and 0% represents complete homogenity.  The maximum diversity of 4 alleles, comprising a total of 100% of those occurring at a locus, cannot exceed 75%. This would be achieved when all four alleles have the same frequency of occurrence, i.e 25%.
*3* "Greater Frisia" was coined by Dr Ken Nordtvedt, during 2004, to describe the North Sea coastal region of the Northern Netherlands and Southern Denmark, after he detected that the frequency of  the R1b combination DYS390=23 and DYS391=11, was unexpectedly high in this region.  See, Ken Nordtvedt's R1b Sub-Clade   at www.worldfamilies.net/Tools/R1b.html
*4* Data from Sweden  was excluded both Baltic groups. Some of its data accords with them, but other data suggests that Sweden and Polish Gdansk may have  received a later input of  Central European R1bs from Bohemia. These R1bs might have been intermingled with the later inruption of R1as  northwards across the Baltic.
*5*  Capelli et al,  A Y-Chromosome Census of the British Isles,   Current Biology, Elsevier Science Ltd. 2003

The R1b patterns that have verifiable geographic links include DYS390=23, DYS391=11 which point to a Germanic or Scandinavian origin. Even here, one must be careful because a recent mutation of DYS390=24 to 23 could explain the pattern better than geographic attribution. This Y-DNA signature includes DYS390=23 and DYS391=11. This R1b Y-DNA signature is clearly different from most Celts.

The type of Y-chromosome markers measured by genealogical genetic testing labs are known as STRs. Genealogists are interested in finding connections between families on a time scale of centuries, and the mutation rate of STRs is such that they are a good choice for that kind of work.
Population geneticists are interested in tracking the movements of groups of humans over time scales of 1000's or 10,000's of years.  Therefore their studies usually involve a different type of Y-chromosome marker known as UEPs that have a much slower mutation rate than STRs.  Haplogroups are defined by patterns seen in the values of these slowly mutating markers.  Identification of your Y-chromosome haplogroup can provide an interesting glimpse into the deep ancestry of your paternal line.
It seems like every author has used different terminology for haplogroups.  The Hg1 naming system is from Wilson, who used DYS19, 388, 390, 391, 392, and 393.  He gave the name Atlantic Modal Haplotype (AMH) to the most common haplotype in Haplogroup 1, where Hg1 was defined by results of a different type of test using slowly mutating markers.
The table below includes the Cohen Modal Haplotype:
             HG1 HG2   HG3   CMH
DYS019   14    14      16      14
DYS388   12    14      12      16
DYS390   24    22      25      23
DYS391   11    10      11      10
DYS392   13    11      11      11
DYS393   13    13      13      12

Some conclusions can be drawn about haplogroup classification by looking just at the STR marker value patterns.  A good rule of thumb for determining haplogroup is the following:

1) If you have a value of "12" at DYS426 and DON'T have a value of "11" at DYS392 then you belong to haplogroup 1 (HG1).

2) If you have a value of "11" at DYS426 then you belong to haplogroup 2 (HG2).

3) If you have a value of "12" at DYS426 and a value of "11" at DYS392 then you are a member of haplogroup 3 (HG3).

The value in Y-chromosome testing lies in comparing results.   Currently, comparing the most common values for haplotypes for each of the markers in our Study there are:

Twenty-eight donors with a “12” at DYS426 that don’t have an “11” at DYS392, placing them in HG1

Twenty-nine donors with an “11” at DYS426, placing them in HG2

One donor has a “12” at DYS426 and an “11” at DYS392, placing him in HG3

The members of HG1 are thought to be the descendants of the Paleolithic hunter-gatherers who arrived in Europe before the last Ice Age about 40,000 years ago (Aurignacian culture).  That pattern is most common in Western Europe, but is also found in all other parts of Europe.

The members of HG2 are believed to be the descendants of two later waves of humans into Europe.  The last of these waves arrived about 8,000 years ago and is credited with introducing agriculture into Europe.  HG2 is most common in Southern and Central Europe, but thathaplogroup is also often seen in those of Anglo-Saxon and Scandinavian descent.  HG3 is seen more frequently on the eastern side of Europe (9% of the population of Turkey is HG3). But it is also common in Scandinavia, and is said by some to be indicative of "Viking blood" when seen in paternal lines originating in the British Isles. The forefather of all HG3's may have been born in the Ukraine during the last Ice Age about 15,000 years ago.

Keep in mind that haplogroup classification is fairly useless for locating the place of origin of your paternal line.  While each haplogroup has general areas in which it is more common, there has been enough mixing of people on the European continent to prevent using these classifications to pinpoint any single place of origin.

A fascinating map of the distribution of haplogroups in Europe is given on page 1156 of Semino's 2000 paper "A Genetic Legacy of Homo Sapiens Sapiens in Extant Europeans: a Y Chromosome Perspective".  HG1 is shown in green, HG2 is shown in red, and HG3 is shown in purple.  Other haplogroups shown on the map but not discussed here are haplogroup HG16 and haplogroup HG21.

In any case, the point of all of this is not for you to look up your haplotype in the list and say either, "Aha! I am HGx because there is my haplotype in the list," or "Uh oh! I must be a freak because I can't find my haplotype in the list."  If you are within one or two steps of one of the listed patterns, then you are PROBABLY in the corresponding haplogroup.

When it comes to Europe, the haplogroups observed can be broadly split into two groups, Palaeolithic and Neolithic.

The first image (Map 1) shows Palaeolithic Europe 18,000 years ago in the grip of the last ice age. Glacial ice 2km thick covers much of Northern Europe and the Alps. Sea levels are approx. 125m lower than today and the coastline differs slightly from the present day. For example, Britain and Ireland would have been connected to continental Europe (not shown on map).

Map 1 - Ice age Europe (18,000 years ago)

The air would have been on average 10-12 degrees cooler and much more arid. In between the ice and the tree line, drought-tolerant grasses and dunes would have dominated the landscape.

The Neanderthals would have died out around 14,000 years ago leaving the nomadic hunter-gatherer Cro-Magnon (modern man) to pursue the animals of the time. Due to the cold and the need for food, the populations of the day waited the ice age out in the three locations shown on the map. These were the Iberian Peninsula, the Balkans and the Ukraine.

These people were skilled in flint-knapping techniques and various tools such as end-scrapers for animal skins and burins for working wood and engraving were common. Cave painting using charcoal had been around for a couple of thousand years although at this time they were now more subtle than mere outline drawings. These artistic expressions are significant as it shows that people are able to obtain some leisure time. Whether this is ‘art for art’s sake’ or objects of ritual is not known.

If we fast forward to 12,000 years ago (Map 2), the ice has retreated and the land has become much more supportive to life. Many animal species have returned to inhabit the land, although the snake, harvest mouse and mole never made it as far as Ireland before the land bridges re-flooded (ever wondered why there are no snakes in Ireland?).

Map 2- spread of Haplogroups R1b, I and R1a (12,000 years ago)

The three groups of humans had taken refuge for so long that their DNA had naturally picked up mutations, and consequently can be defined into different haplogroups. As they spread from these refuges, Haplogroups R1b, I and R1a propagated across Europe.

- Haplogroup R1b is common on the western Atlantic coast as far as Scotland.
- Haplogroup I is common across central Europe and up into Scandinavia.
- Haplogroup R1a is common in eastern Europe and has also spread across into central Asia and as far as India and Pakistan.

These three major haplogroups account for approx 80% of Europe's present-day population.

Around 8,000 years ago (Map 3), the Neolithic peoples of the Middle East that had developed the new technology of agriculture began moving into Europe. There were several haplogroups involved, mainly E3b, F, J2 and G2.

Map 3 - spread of Neolithic haplogroups (from 8,000 years ago)

These Neolithic haplogroups came in several waves over time and are found predominantly along the Mediterranean coast. Around 20% of the present-day population are from these Neolithic haplogroups. What is interesting to note is that the agricultural technology spread much further than the people who first 'invented' it.

A little later, around 4,500 years ago, Haplogroup N3 began moving across from west of the Ural mountains. Haplogroup N3 follows closely the spread of the Finno-Ugric languages.

Source: http://www.dnaheritage.com/masterclass2.asp

If these results represented only a single modal haplotype at 24,11 (the Atlantic Modal Haplotype) which had come into existence in the past and had been accumulating mutations in 390 and 391 since its founding, the fall off in population would tend to be more symmetric than seen. One of the first things to note is that 390 mutates faster than 391, in any case. There is excess population at 23/11 and 24/10 beyond what one would expect by one-step mutations from the AMH. If independent modal haplotypes came into existence at 23/11 and 24/10 at their individual times in the past, and we superimpose these populations with their mutations in 390 and 391 which have accumulated since their founding, then the asymmetric distribution of populations can be understood significantly better. But there is still an apparent excess population at 23/10 beyond what one would expect from one-step mutations from the two new modal haplotypes at 23/11 and 24/10. So an even smaller population of descendants of a modal haplotype 23/10 completes the picture.

The numbers in the Figure are Total Europe. Although I see the evidence I have explained above for the need for a hierarchy of four modal haplotype descendant populations, that by itself would not be that useful, and perhaps not that convincing to some. But the YHRD database is divided up into about 100 regional databases spread throughout Europe. One can find the populations of these four modal haplotypes by region and determine if the geographical distributions differ by modal haplotype population? If they do in a clear measured way, this is strong confirmation that we are seeing in today's product of haplotype distribution the results of four populations of peoples who have to some degree spread out and migrated in Europe differently.

Very briefly, some of the highlights of geographic patterns seen are:

24/11 (ATM) is very high population with the Basques and surrounding Iberian areas and then up the Atlantic coast to France and Britain; well-present along the Rhine and into southern Germany.

23/11 is highest in "Greater Frisia" (Belgium, Netherlands, NW Germany, Denmark); solid throughout Germany and falls off back in Iberia

24/10 is elevated back in Iberia but not like 24/11; it hardly gets to Norway/Sweden.

23/10 is present most strongly toward SE Germany; it also hardly got to Norway/Sweden.

Detailed geographic tabulations for these modal haplotype populations are posted to the List.

Note: The 809 haplotypes of the 23/11 variety surely include many one-step mutations of haplotypes which are descendants of the 24/11 founder, and vice-versa. Similarly, some fraction of the 23/10 haplotypes are not descendants of the hypothesized founder of that modal haplotype, but they are instead one-step mutations of originally 24/10 haplotypes, and vice-versa. The population flow up and down the boxes due to single-step mutations at DYS 391 is significantly less due to the evidently lower mutation rate of this marker. But from the overall pattern of populations in the Figure, one makes more sense of this pattern by adding the four sources of founding modal haplotypes with their separate growths of descendant populations.

Look at the one-step mutational exchange between 23/11 and 24/11. If we want to understand the ratio of descendants of the modal haplotype founders, then mutations between the (larger) 24/11 population into the (smaller) 23/11 population will tend to enlarge the 23/11 population and diminish the 24/11 population --- unless there is substantial difference in "up" versus "down" mutational rates. These corrections can be done. Also, populations like 25/11 and 25/10 can be assumed to be one-step mutated variations of the 24/11 and 24/10 descendant populations, respectively. These corrections can also be made for purposes of improving the geographical statistics and better determining the sub-clade populations. What do I mean by "true sub-clade populations"? Unique SNP mutation tags will probably be found eventually for each of these identified modal haplotype descendant populations; they will then become haplogroup sub-clades or sub-sub-clades .... of R1b.

One might ask, "why don't you go on to assume another very small sized modal haplotype 25/11 to "explain" the 319 found haplotypes"? First reason for not doing so is that 319 does not seem to be too unusual a number of mutations from the large 24/11 population; and secondly, there would simply not be the statistical power from the geographically divided database to check whether this additional hypothesized modal haplotype had its own unique geographical distribution in Europe.

My purpose for doing this kind of study is not to give every R1b haplotype another label, but to try to understand the pre-historic history of the spread of peoples in Europe. The "history" will be written in the haplotype and haplogroup distributions. For purposes of the haplotypes of individual people, this structure may tilt the odds as to the deep origins of their haplotype.


Various Notes:

Anglo-Saxon Sub-Clade of R1b:
This sub-clade will have values of 23/11 at DYS markers 390 and 391.
If one's known ancestry is in the British Isles and one has R1b of this sub-clade, the odds are tilted against that being an "indigenous" R1b and toward being a NW European continental R1b brought to the British Isles by one of the historic invader/immigrant groups from the Low Countries, NW Germany, and Denmark. This tilt should be incorporated into all the other surname and related information you have about origins of your R1b. This represents Anglo/Saxon England populations after the Roman occupation ended in 410 AD but before the Norman/Viking populations in the early 1000's BC.

The combined Weale/Capelli data also reveals that the proportion of the combination, 23/11 for DYS390 and DYS 391, is about 33% of the Germanic R1b y-dna found in Friesland.

North Sea-Baltic Group

PURPOSE: The group is an informal, voluntary contact group for anyone wishing to contribute their knowledge, generally participate, or share results of their Y-Chromosome DNA testing, as applied to genealogical research into the origins of the R1b, Y-Chromosome sub-clade borne by the Germanic-speaking populations of the North Sea countries.

RATIONALE: There is historic evidence that, during the first century, BC, 'Celtic' R1b groups in the 'Germanic' part of continental, North Western Europe began to combine (and intermarry) with other groups, such as the Teutons, Slavs and Wends ( who possessed R1a and I1a y-dna) to form homogenous, Germanic speaking populations, to defend their territories against the Roman Empire. Recent research has shown that, later, within England, and to a lesser extent within the Scottish border areas, when the Anglo-Saxon and Danish Viking incursions took place the indigenous, ancient British R1b (Celtic) population was largely dispersed and replaced by the invaders, whose Germanic R1b (North Sea version), R1a, and I1a y-dna became the dominant strains in those areas. In England, the invaders y-dna eventually accounted for about 67% of the newly-established 'English' male population, with the ancient British accounting for about 33%. Within the now mixed Germanic/British R1b population, of England, about 55% of the males have Germanic, 'Celtic' paternal ancestry descent, rather than ancient British. In particular, there is evidence that the R1b values of 23/11 for DYS 390 and DYS391 markers within England, and in some parts of Scotland, are especially likely to have deeper Germanic paternal ancestry: These marker values occur much more frequently in the coastal areas of Denmark and the Netherlands, and within England, but remain much less frequent within the remaining Celtic R1b areas of the British Isles.

A.A. Foster, 20th December. 2004

Various Notes:

Further research, in 2004, into R1b haplotypes, by Ken Nordtvedt , revealed, inter alia, that the alleles of 23 for DYS 390 and 11 for DYS 391, appear much more commonly in the Friesland and Danish Jutland areas of the North Sea than elsewhere in Europe. A re-examination of Weale's and Capelli's data reveals equally high frequencies of the 23/11 combination within the English R1b population - and at very low frequencies within the more remote Ancient British regions of Ireland, Scotland and Wales. This confirms that the Anglo-Saxon/Danish migrations to England, and Southern Scotland, had indeed originated in the coastal North Sea area north western Europe, and that they had replaced a large percentage of England's indigenous Celtic population.

Searching the YHRD database Alan Foster found that the North Sea Baltic Group is most commonally found in Northwest Europe, from Belgium to Southern Norway with its highest frequency in Denmark!
Denmark 15.9% 
Friesland 13.5%  
Netherlands 12.3% 
Belgium 12%
Britain 6.2%
Germany 6.2%
Switzerland 6%               
France 4.8% 
Portugal 4.8% 
Spain 4.4%