July 14, 2012

Y chromosomes and mtDNA from late antique Bavaria

One of the papers in the aforementioned volume includes Y-STR and mtDNA data on a burial cemetery from Bavaria dating to the Imperial Roman age. I reproduce the DNA results below; the haplogroup assignments in red are my own and have been estimated with Whit Athey's haplogroup predictor using both Northwest European and Equal priors.

The number of Y-STRs is not sufficient to make very strong haplogroup assignments in some cases. Still, we can probably say that R1b, E1b1b, and I1 were present in the population. I1 might seem more likely than G2a in a few cases, but remember that a couple of G2a men were found in 7th c. Bavaria. E1b1b, another non-typical German haplogroup has also been found in Usedom from the medieval period.

Christina Sofeso, Marina Vohberger, Annika Wisnowsky, Bernd Päffgen, Michaela Harbeck, *

Verifying archaeological hypotheses: Investigations on origin and genealogical lineages of a privileged society in Upper Bavaria from Imperial Roman times (Erding, Kletthamer Feld)

During the years 2005 and 2006 approximately 2000 archaeological
finds ranging from the Neolithic Period to Late Antiquity 
were found on the Kletthamer Feld (Erding, Upper Bavaria). 
Out of this context a burial site was examined comprising 
13 individuals, some of them rich in precious grave goods. The 
inhumations were dated to the second half of the 4th to the first 
half of the 5th century – a time of upheavals in relation to the 
demographic structure of the former Roman province Raetia (today southern Bavaria).

The high proportion of male individuals within the skeletal population as well as the finding of a Roman fibula, which is seen as part of Roman military clothing, led to distinct hypotheses which we have attempted to support in this study. The hypothesis that the skeletal remains reflect a founder population from a Germanic region north of the Danube River could be rejected on the basis of stable isotope analyses. The theory of a buried family clan had to be dismissed as well, or rather, be extended to the scenario of several families being buried there with their servants. The results obtained fit the third presumption best, namely that the buried individuals were the members of a military unit interred with their families.


Project "Magnus Ducatus Lituaniae" said...

Hg I haplotypes are rather I2b2/I2a2b than I1
Sample Y-Haplogroup Probability
EKF1662 R1b1b2-M269 87%
EKF1664 I2b2-S154 82%
EKF1665 F-M89 53% (uncertain, could be R1b)
EKF1700 E1b1b1a1-M78 67%
EKF1704 I2b2-S154 82%
EKF1717 E1b1b1a1*-M78(xV12,V13,V22,V65) 93%
EKF1719 F-M89 97% (uncertain, could be R1b)

truth said...

E1b1b is not a non-typical german haplogroup, specially in Southern Germany/Austria area, where it reaches about 8-10%, that's about higher than many parts of Iberia.

Dienekes said...


I presume you are using a different haplogroup estimator?

terryt said...

"The results obtained fit the third presumption best, namely that the buried individuals were the members of a military unit interred with their families".

A military unit? The E haplogroups could be from mercenaries rather than from natives.

Project "Magnus Ducatus Lituaniae" said...

Yes, i'm using the YPredictor

Abbat said...


Typical I1, no value is extreme.

Anonymous said...

Whit's method is probably more accurate with fewer markers. Using something else I got the following:

1662: R1b-U106/S21, R1b-Ub (ubiquitous cluster that's not yet confirmed as purely P312*)
1664: I-M284, I-S23, I-S24, J2-M102
1665: I-L22
1699: R1b (Northwestern likely), I-M253, T
1703: E probably V22+
1704: I-M284, I-S23, I-S24, J2-M102
1717: E-M34
1719: I-M253

AWood said...

Was anyone able to determine from the text which corpse wore the Roman pin? I didn't see it explicitly stated anywhere.

Anonymous said...

Danish archaeologists reopened grave of Iron Age warriors: http://www.reddit.com/r/Archaeology/comments/w0crm/archaeologists_dig_up_bog_army_bones_in_denmark/

They state Roman expansion may have been a possible cause. The remains are preserved well, so they’re going to analyze DNA.

Dienekes said...

Whit's method is probably more accurate with fewer markers. Using something else I got the following:

Yes, YPredictor misclassifies my own Y chromosome, so I am rather inclined to put my money on Whit's predictor.

But what is the "something else"?

Anonymous said...

Cullen's predictor: http://members.bex.net/jtcullen515/haplotest.htm

Resubmitting the marker inputs changes the probabilities. EFK 1664 and 1704 values listed I-M253 but at a slightly lower probability.