Monday, March 14, 2011

Y-DNA and Viking Migration

Y-DNA and Viking Migration

Y-DNA analysis of STR information can provide clues to Viking migration in recent time frames. The Y-DNA STR Genetic Distance suggests that the Norse (Vikings) could have originated from seafaring peoples of the Mediterranean Sea.


Ever since the National Genographic Project began, DNA has been a useful tool to analyze migration patterns of various indigenous peoples. The National Genographic Project typically uses a combination of Single Nucleotide Polymorphisms (SNP) and a limited number of Short Tandem Repeats (STR). The analysis of a larger number of STR's can be used to provide clues to Viking migration in recent times.

The technique has been used successfully in the past by the HAM DNA Project in order to provide clues to the ancestral origins of the participants. An extension of the technique can be used to help understand the migration patterns of the Norse peoples, often referred to as Vikings.

It would appear, however, that nobody has attempted to analyze the the migration patterns of the Norse groups up to this time. In this study, Y-Search data for the HAM DNA Group #1 was used. This group has been previously shown to have an ancestral haplotype that corresponds to the modal for the I1 (M253) haplotype group. Therefore, the results may apply to a number of I1 groups. For example, HAM DNA Group #7 has German origins, and the findings of this analysis can also be applied to this separate I1 group.

Further, HAM DNA Group #1 contains an individual with known ancestors from County Somerset, England. The HAM surname for this area has been documented back to Taunton circa 1250 A.D. There is an ancient Norman castle in Taunton, and other Norman strongholds in that area of County Somerset. Additionally, County Somerset was not settled by the Danes, providing further support that the arrival of haplotype group I1 in County Somerset, England was with the Norman invasion of 1066. It is important to note this, as the time scale (below) has been dated to the time of the Norman Conquest.

Previous Background: Scientific Studies

Siiri Rootsi, et. al., (2004) was perhaps one of the first to suggest haplotype I1 origins in Norway. More recently, Alexander Shtrunov has written about the I1 (M253) origins in Eastern Europe. They did use a limited number of STR values in order to infer population frequencies. Regarding haplotype I1a (now reclassified as 'I1'), Siiri Rootsi, et. al. noted:

"Subclade I1a accounts for most of Hg I in Scandinavia, with a rapidly decreasing frequency toward both the East European Plain and the Atlantic fringe, but microsatellite diversity reveals that France could be the source region of the early spread of both I1a and the less common I1c. Also, I1b*, which extends from the eastern Adriatic to eastern Europe and declines noticeably toward the southern Balkans and abruptly toward the periphery of northern Italy, probably diffused after the Last Glacial Maximum from a homeland in eastern Europe or the Balkans. In contrast, I1b2 most likely arose in southern France/Iberia. Similarly to the other subclades, it underwent a postglacial expansion and marked the human colonization of Sardinia ∼9,000 years ago...."

"Previous studies revealed that Hg I reached a frequency of ∼40%–50% in two distinct regions—in Nordic populations of Scandinavia and, in southern Europe, around the Dinaric Alps—each showing different background STR modal haplotypes (Semino et al. 2000; Passarino et al. 2002; Barac´ et al. 2003)."

I1 density map, Rootsi et. al., 2004

Rootsi et. al. had recognized that I1 was observed in high frequencies among Norse (or Scandinavian) populations. Here, they speculate early origins in France, and mentions the locations of I1 in France, Scandinavia, and the Dinaric Alps. They also mention I1b in the Balkans and northern Italy.

Ken Nordvedt is a widely known researcher of the Norse populations. He has categorized the various DNA patterns into a number of different Norse groups, largely based upon STR patterns. It should be noted that Ken has recently commented on the Genealogy-DNA email list (Dec 26, 2010) that after examination of I1 interclade ages:
      "TMRCA for AS1 is 70 generations; TMRCA for AS2 is 64 generations, and the interclade node age between them is 180 generations estimated.
   So perhaps you can see why I conclude that the overall I1 TMRCA lived 4000 to 4500 years ago?"

For the purposes of this study, it is useful to be aware that dating Y-DNA samples using STR values is still under development, and that Y-DNA SNP dating is, at least, arguable.
David Faux has written separately on the Cimbri people, an element of the Danish Vikings who lived between the 9th to 11th centuries (2007).

Alexander Shtrunov has written (2010) about the Vikings in Eastern Europe and Russia, and has some discussion of various tribes, such as the Goths, the Varangians, the Sami people, the Arensburgian people, Swiderian, etc. He appears to have analyzed the DNA for SNP population frequencies, combining genetics, archeology, linguistics, anthropology. For I1 in Eastern Europe, he suggests:

"Roots of haplogroup I1 evidently came from such Paleolithic cultures as Ahrensburgian and Swiderian; its carriers represented were the part of autochthonous population of Northern and Eastern Europe."

The National Genographic Project has current information about I1 (M253) migration paths online, and should be publishing the final results of the Project in the future. Past migration maps of the I1 haplotype had shown an overland route through the mountains of Europe, north of the Mediterranean Sea upwards to Norway, then back down the western coast of Europe to France and England. This has traditionally been displayed as an overland route to Norway.

The current National Genographic web page for M253 is out of date, as the web page for M253 is still using the old "I1a" nomenclature (Family Tree DNA reclassified this as "I1" in May, 2008):

National Genographic map

This (Viking Migration Y-DNA STR) study suggests that the I1 haplotype possibly took a sea route from the Mediterranean Sea (perhaps Sicily), first landing in the Netherlands.

Regarding the HAM DNA Project, the National Genographic Project shows

For I1, the SNP's are:

"Adam" -> M168 -> M89 -> M170 -> M258 ( -> M253 -> M307) Migrated from the middle east along the northern coast of the Mediterranean Sea, then finally migrating north along the east coast of Europe. (Entered the area of Norway)

For I1b, the SNP's are:

"Adam" -> M168 -> M89 -> M170 -> M258 (Negative for M253 and M307 ) Migrated from the middle east along the northern coast of the Mediterranean Sea, then finally migrating north along the east coast of Europe. (did not enter Norway)

The focus of the data was on the Genetic Distance for the I1 haplotype STR values. Other haplotypes that may have been included in the data would be the I1d haplotype group.

There is an I1 Project at Family Tree DNA that summarizes a number of the Nordvedt and various I haploype group (SNP) categories. This web site presents the raw data from FTDNA participants.

The Procedure

Using the I1 ancestral haplotype of the HAM DNA Project (Group #1, the ancestral haplotype previously determined), the Y-Search database was utilized to obtain matches on 35 to 42 markers. This returned about 1,000 kits that matched the haplotype given.

The number of STR markers used typically numbered 67, with a maximum at about 96 STR markers.
Genetic Distance typically ranged from 15 to 25
There are about 39 mutating markers for the I1 haplotype in this study.

The output was then sorted by Country of origin, ignoring the results from the United States and Canada. For each Country of origin, the STR marker values were provided as input into Dean McGee's Y-DNA Utility. A modal value was obtained for each country of origin. These modal values (of countries of origin) were then provided as input into Dean McGee's Utility again, this time without computing the modal. The output was then run through the "Kitsch" program of the PHYLIP software package for proper rooting of the resulting tree. The phylogenetic tree was then examined graphically for time scale. The Time to Most Recent Common Ancestor of these modals (as given by McGee's Utility) was suspect, having known the time of the Norman Conquest (i.e., 1066). Therefore, the resulting phylogenetic tree was then scaled, using the Norman Conquest as a basis for the time frames on display.

I1 geographic modal values by country of origin
[ click on image to enlarge ]

The modal table (above) shows some 39 mutating markers by country of origin. Generated by Dean McGee's Y-DNA Utility, it color codes some of the markers by Genetic Distance from the reference (row #1).

The mutating markers that appear to be useful for geographic identification include DYS464d, DYS576, CDYb, DYS413a, DYS635, DYS494, DYS522, and DYS549.


The resulting tree shows a common entry into England. This resulted from using the I1 modals for the participants that indicated separate origins from England and Scotland. At the time of the Conquest, England and Scotland were united into one country, so the DNA results appear to reflect that properly.

Phylogenetic Tree of modals for various countries
[ click on image to enlarge ]

What the Y-DNA modals (per region) revealed was that each country typically displayed separate times of origin. That is, the genetic distance was found to correspond to region.

An overland route through Switzerland might be suspected if the Genetic Distance had reflected that. However, the Y-DNA suggests that Switzerland was a migration point much later than that of that of Scandinavia. That is to say, the Genetic Distance to the Netherlands suggests that the Netherlands as one of the first (if not the first) point of entry into northern Europe. An alternative overland route would be through Austria. Austria does appear to have early origins. However, it ia hard to explain an overland route to Austria with the data available for this study. The Y-DNA STR values suggest that Austria began to be settled at about the same time as Finland, Spain, and France. During this period (roughly the 4th century), it is apparent that the countries were being settled by I1 via sea routes. Settlement of the I1 haplotype in Austria is perhaps, better explained by a sea route as well.

Also, it is well known that the Norse (Vikings) were very skilled with ships. This is reinforced by the Genetic Distance from the Y-DNA. It makes sense that that a path of migration could have been by sea, with a departure from the Mediterranean Sea and arrival in the Netherlands.

Problem discussion

There are some basic problems with the analysis:

a) The data was limited to less than 100 STR markers per individual.

b) There is limited data available from the Y-Search database, notably limited information from I1 participants in Sicilly. The early arrival in Portugal may support the theory of seafaring people from the Mediterranean. This data could be questioned given the scope of the Roman Empire, and not knowing how long the participants have actually lived in Sicily. On the other hand, Norse settlement in Sicily has been dated by some to the 11th century. The Norman Conquest timeline is supported by the Tune Runestone in Oestfold, Norway, which has been dated to the 5th century. More archeological evidence should show whether or not the theory is valid.

c) This study presumes previous SNP testing for I1, and does no real analysis of SNP data.

d) The timeline reported from these utilities can change, depending upon the data in use.
Proper scaling of Y-DNA phylogenetic trees along a timeline is not currently reliable, due to the relatively low number of markers currently being tested. Carbon dating is typically used instead.

e) This study makes no attempt to include other Norse haplotype, such as R1b or R1a.

f) Finally, this study ignored DNA results for I1 in North America. Which is to say, no attempt was made to locate Native American I1 modal values. Following the M170 migration route would suggest that Viking migration was not from the Americas, but no attempt was made to confirm this with haplotype group I1 STR values from Native Americans.


When the Y-DNA STR marker data is separated by location, a modal can be derived for each location. Each modal for each country then appear to indicate marker differences that may reflect the location. When the modals are then examined for Time to Most Recent Common Ancestor, a continuous migration trail is implied.

The data suggests that the Norse Vikings perhaps originated as seafaring peoples from the area of the Mediterranean Sea (perhaps from Sicily). These people appear to date from the very early Roman period. (The city of Rome is believed to have been founded in 753 B.C., but archeological evidence suggests a slightly earlier date.) The DNA suggests that the initial point of entry of the I1 haplotype into Scandinavia appears to be from the Netherlands.

The phylogenetic tree suggests a possible Norse migration in this order:

Sicily --> Netherlands --> Finland -> France -> Denmark -> Norway -> England, Scotland & Germany

2500 <--- 2010 <--- 1700 <--- 1500 <--- 1300 <--- 1300 <--- 1000 <--- (rough est. of years ago)

It is important to note that the entrance of the I1 haplotype into France is indicated to have happened much earlier than the settlement of Normandy (in the 9th century). The phylogenetic tree suggests that the Vikings should have entered France some 400 years earlier (circa 375 A.D.). The Y-DNA STR data regarding France suggests a correlation to the quote from Rootsi (2004) that the early spread of I1 could have been through France. Here, we find France and Denmark to be one of the early locations, but the entrance into Eastern Europe appears to have been earlier than through France. This study shows that the I1 entrance into Europe began in the Netherlands, Russia, and Belgium circa 100 A.D.

The Time to Most Recent Common Ancestor roughly estimated by using McGee's Utility on the modals for STR markers, then scaling to the time of the Norman Conquest. Carbon dating of the archeological evidence from these locations may provide a more robust timeline. In particular, better evidence may require more Y-DNA STR markers, or a better study of M170 in the area of the Black Sea. More Y-DNA testing in the area of the Mediterranean may provide better evidence for an earlier point of origin. Alternatively, Y-Search could be used with a greater Genetic Distance to search to search for M170 matches near the Black Sea.

note: edits added 03/16/2011


update 03/21/2011:

I have been able to use Y-Search to find a minimal amount of (I1) Y-DNA samples for 9 additional countries. Of interest to this study is the timeline for Sicily appears to hold using minimal data from Romania, Turkey, and the Ukraine. Using the data for the area around the Black Sea, the sequence appears to be more like this:

232 B.C. Croatia
324 B.C. Romania
624 B.C. Turkey
662 B.C. Sicily
887 B.C. Ukraine
1075 B.C. Slovenia
1300 B.C. Czech Republic

The DNA information from Croatia suggests that Croatia may have been the departure point to Portugal and the Netherlands.

Which suggests that the time line for Sicily is supported by data from Romania, Turkey, and the Ukraine. What is interesting is the data from Slovenia and the Czech Republic. This suggests a period of migration of this I1 group in and around the Black Sea prior to the beginning of the (more bold) Mediterranean adventures.

update September, 2011:

The results of the STR modal TMRCA estimates have been graphically mapped here:



Siiri Rootsi, et. al., "Phylogeography of Y-Chromosome Haplogroup I Reveals Distinct Domains of Prehistoric Gene Flow in Europe" Am. J. Hum. Genet. 75:128–137, 2004

Alexander Shtrunov, "The origin of haplogroup I1-M253 in Eastern Europe" The Russian Journal of Genetic Genealogy: Vol 1, No. 2, 2010

Ken Nordvedt

Ken Nordvedt: Genealogy-DNA email list at rootsweb ( ),
Subject: "[DNA] I1 L22- interclade ages," Sun, 26 Dec 2010.

David Faux "The Cimbri Nation of Jutland, Denmark and the Danelaw, England: A Chronological Approach Based on Diverse Data Sources" Sep., 2007

The National Genographic Project

The I1 Project at FTDNA


Dean McGee's Y-DNA Utility

The PHYLIP software package

HAM DNA I1 (Group #1) Ancestral haplotype:

Wikipedia page on Ancient Rome:

Wikipedia page on the Tune Runestone from Oestfold, Norway:

'How to Read HAM DNA Phylograms' You Tube video

to comment, click on the title and scroll to the bottom

No comments: