July 02, 2009

Genetic discontinuities between Etruscans and modern Tuscans

It is great that such a large medieval sample was assembled for this study, and hopefully it will be possible to type it for either Y chromosome or autosomal markers to determine whether continuity between medieval and modern Tuscans was only matrilineal.

I am not very surprised by the inferred genetic discontinuity. Gene pools maintain their separateness if their bearers have some sort of distinction (linguistic, political, religious, or cultural) from their neighbors. For Etruscans, such distinctions were rapidly dissolved when they were annexed by the Romans. In Imperial times, their language was still known by some, but this, too, passed into oblivion.

As distinctions disappear, so do impediments to bidirectional gene flow. The genetic characteristics of the original people do not so much disappear (genetic genealogists will surely soon scour the databases for ancient Etruscan matches, if they haven't done it already), but are diffused in the larger pool of now undifferentiated neighbors, who, in their turn, diffuse into the territory of the old ethnic entity.

The "Etruscans" label of this post points to many studies in this unfolding story of Etruscan origins. Etruscans remain, until now, the only ancient Mediterranean population for which a substantial mtDNA characterization exists.

PS: Interestingly, the conference abstract which I pointed to earlier seemed to suggest that the genetic discontinuity occurred after 1,500AD rather than before 1,000AD, as the published paper does.

(More on the details of the study to follow after I read the paper)

UPDATE I (Jul 2)

From my reading of Table 2, the medieval Tuscan sequences are:

10 of CRS
2 of 16311C
2 of 16294T 16296T 16304C

and 1 of the following:

16224C 16311C 16355T
16126C 16193T
16126C 16193T 16294T 16296T 16304C
16126C 16294T 16296T 16304C

which seems to indicate a mix of haplogroups H, HV, T, and K in the population according to the Genographic project tool.


From the paper:
Analyses of mtDNA diversity in the British Isles (Töpf et al. 2007), and Iceland (Helgason et al. 2009), also showed sharp differences between historical and current populations. In addition, a large fraction (up to 80%, depending on the region considered) of the Dutch surnames were displaced from the areas in which their frequency was highest three centuries ago (Manni et al. 2005). Nobody can tell whether the Netherlands represent an exception or the rule, until similar studies are carried out elsewhere, and there is no comparable information on previous centuries. However, the point here is that a genetic discontinuity between present and past populations seems rather common in the few European countries studied so far. Deep demographic changes in the last two millennia are both suggested by the analysis of ancient DNA in Tuscany, Iceland and Britain, and empirically demonstrated in the Netherlands. Our failure to reproduce by simulation the observed haplotype number of the contemporary Tuscan samples may mean that such changes involved multiple immigration processes, too complex to model at present.
The paper by Töpf et al. in turn points to this study of ancient British mtDNA which I had forgotten about. That study shows an increase of haplogroup H (as most of the OTHER probably is) in modern times compared to the past, and the drastic reduction of some haplogroups as U5a1 and U5a1a. Other cases of apparent drastic change over time, involves the Central Europeans (reduced haplogroup N1a) compared to early Central European farmers., and medieval vs. modern Danes (reduced haplogroup I).

So, the picture does seem to suggest substantial changes in mtDNA gene pools over time across many parts of Europe and time frames. Whether this reflects population movements or selection, remains to be seen. In the paper on the Netherlands, for examples (Manni et al.) cited in this paper shows that the original surnames in a region can be rapidly replaced over a genealogical time frame.

Studies such as these put into question the widely held assumption that modern gene pools reflect prehistorical events, such as the repopulation of Europe after the glacial age, or the advent of farming. If genetic change is so substantial over 100 generations, we are rather foolish, I believe, to attempt prehistoric reconstructions about events that took place 300 or even 600 generations ago.

UPDATE (July 13): An additional factor that may explain why ancient gene pools look different than modern ones may be of course due to post-mortem damage of the DNA, which makes it look different when it is not so. However, in the case of the Etruscan data of Vernesi et al., the question of DNA degradation was addressed in an independent study by Mateiu et al. which found no evidence for it.

Thus, while one can't be too cautious, the evidence seems strong in this case that we are dealing with an authentic snapshot of ancient Etruscan mtDNA.

Molecular Biology and Evolution, doi:10.1093/molbev/msp126

Genealogical discontinuities among Etruscan, Medieval and contemporary Tuscans

Silvia Guimaraes et al.

The available mitochondrial DNA (mtDNA) data do not point to clear genetic relationships between current Tuscans and the Bronze-Age inhabitants of Tuscany, the Etruscans. To understand how and when such a genetic discontinuity may have arisen, we extracted and typed the mtDNAs of 27 medieval Tuscans from an initial sample of 61, spanning a period between the 10th and 15th centuries A.D.. We then tested by serial coalescent simulation various models describing the genealogical relationships among past and current inhabitants of Tuscany, the latter including three samples (from Murlo, Volterra, Casentino) which were recently claimed to be of Etruscan descent. Etruscans and medieval Tuscans share three mitochondrial haplotypes, but fall in distinct branches of the mitochondrial genealogy in the only model that proved compatible with the data. Under that model, contemporary people of Tuscany show clear genetic relationships with Medieval people, but not with the Etruscans, along the female lines. No evidence of excess mutation was found in the Etruscan DNAs by a Bayesian test, and so there is no reason to suspect that these results be biased by systematic contamination of the ancient sequences or laboratory artefacts. Extensive demographic changes before 1000 A.D. are thus the simplest explanation for the differences between the contemporary and the Bronze-Age mitochondrial DNAs of Tuscany. Accordingly, genealogical continuity between ancient and modern populations of the same area does not seem a safe general assumption, but rather a hypothesis that, when possible, should be tested using ancient DNA analysis.



Gioiello said...

I have already discussed similar studies in the past: Etruscans were probably an upper class come from Aegean Sea which superimposed themselves on Italian peoples (Villanovans and Osco-Umbrians). The ancient Etruscan mtDNA was probably of this upper class, which was, as usual, limited in number, and, of course, destined to fall in number as time passed. There has been found some restricted Etruscan mtDNA linked to Asia Minor, like U7 you discussed on your forum, but also some haplogroup, which, though thought of Eastern origin, are present only in Tuscany, like my relative R0a of whom I spoke in the past on other forums. Being Tuscans one of the seven donor people of all Europe, with Italians, Sardinians, etc., I think that this does mean that Tuscans are a people genetically characterized almost from Paleolithic or Younger Dryas.

Kepler said...

"In addition, a large fraction (up to 80%, depending on the region considered) of the Dutch surnames were displaced from the areas in which their frequency was highest three centuries ago (Manni et al. 2005)."
OK, I wonder if this Manni tried to find things out in the Netherlands.
As far as I know what happened was that Napoleon annexed the Low Lands in 1810 and new rules for registration of family names took places. Before that a lot Dutch used some form of patronymic, as in other Germanic groups, like Willemsen -> Willemszoon. They were not registered but for the rich. Only with the introduction of the code civil people were forced to do it and they often took something else.
There is some here (in Dutch) about the Napoloen law:

I have heard from people in Flanders (without checking it out in a history book) that many in the Netherlands took just any name out of protest as they thought this was going to be something temporal. That is why - so they told me - Netherlanders often have very funny-sounding names than Flemish people.
This seems to say that as well (but I'd rather see that in a more serious source)

eurologist said...

Similar laws in Germany led to many of the current German-Jewish family names, that were chosen to reflect their professions, dreams, or goals towards the end of the 18th century and later - also first in cities, and first among rich people. Thus names like Rosenthal, Goldberg, etc. It would be utterly foolish to try to use names to look at population changes in much of Europe any time before probably the mid 19th century.

Although church registrars exists for almost a millennium in many places, the names listed usually refer to location (a particular village, part of a village, a particular named farm, a geographic location with respect to a river, forest, hill, etc.) --- they don't necessarily reflect genealogical lines.

As to the Etruscans, I mentioned this before: being relatively rich is often associated with marrying outside. Rich people travel more, and can much more easily choose from a wider segment of the population without stigmatization. MtDNA of a small group of settlers could easily be replaced in just a few tens of generations - even assuming initial isolation and growth. Especially, once the Etruscans lost their purported matriarchal structure.

cacio said...

Thanks for the posts on the Etruscans, being from Tuscany myself, I'm quite interested. What I did not get is - they didn't sample any new Etruscans, did they? They must have then used the (perhaps dubious) Vernesi et al.

Besides this, looking at the Vernesi paper, the results don't strike me as particularly different from the ones cited here, I must say. Various J/T's CRS, etc. Perhaps there's more CRS now than in Vernesi, but with a sample of 28 (Vernesi), what can you say anyway?


Kepler said...

What about the Y chromosome? Any studies after the ones claimed in the Wikipedia articles of

McG said...

A similar problem exists for R1b in the Isles. Who were the Picts and where did they go? Y-STR evidence suggests that later tribes evolved from the Picts and thus the haplotypes are quite similar. I don't think there was an invasion from some other genetic source. In this context I think that meaningful inferences about the past can be made?

Anonymous said...

Studies of Tuscan dna using SNPs have shown that the Tuscan group tested have more of an Eastern imprint (Anatolian or Levantine). Haplogroups are sex biased, and can and do change over time. The Iceman Oetzi had an mtDNA haplogroup belonging to K and has no modern European descendants, well, as yet found. It is likely many haplogroups and subgroups have become extinct due failing in the reproduction stakes. However, this may be the case for many haplogroups and subgroups but modern Europeans, Tuscans included, are the descendants of previous generations, some who were Etruscans. Autosomally or on the level of SNPs, the people may be extinct like the Etruscans but their dna lives on in modern Italians. Haplogroups are just a small part of the dna that is passed on. Similarly the Picts are no more as an ethnic group, culture and language. Their history is dead but their dna lives on in modern Scots. It is just a matter of finding SNPs which distinguish Etruscans or Picts and locating those in their modern descendants.

Anonymous said...

@ Kepler
I read Etruscans had strong J2.

Unknown said...

Can somebody help me to explain why I have Tuscany DNA? My family has been living near Agean coast and Black Sea coast of Turkey for the last 400 years?


Anonymous said...

I also group with Tuscans with the Genographic testing. My family is more from Emilia-Romagna on both sides. I guess Tuscany is close enough to northern Italian genetics in general. People from Turkey came through Northern Greece and then went into Tuscany. Geneographic shows this on their migration maps. I would assume Mr. Aydin that some people of Turkish ancestry are related to Greeks and Tuscans. Being on the Black Sea Coast of Turkey is rather close to the migration route of people from Turkey into northern Greece and then into northern Italy. I also group with Greeks, but have no Greek ancestry. I show up with a lot of marker in common with Otzi the Iceman. I also have the newer K1A that is common in northeast Italy and some think came from there. I also have the older U group mtDNA like U8 and U8b, showing that my maternal ancestors came into Europe 40,000 years ago and then down into Italy. My paternal ancestors most likely came in from Turkey or what people called Anatolia. Realize all these groups are genetically similar due to old migration routes that are in our deep ancient ancestry. Genographic gave me less than 50 mtDNA markers to look at, which is quite annoying. Once they saw K1A, they stopped looking further and stopped at 16278. The Tuscan mtDNA seems to be higher up than that. I am a woman, so I could have these markers and would not know. I'd have to test with a company that gives me all my raw data and not just what they think I need to see. In any event, some Turkish people do group with Greeks and Tuscans. They are close to each other.

Unknown said...

I have Tuscan, North Italian and italian mtDNA. My DNA test said that I am Italian. Both my parents were born in Greece. All my grandparents were born in Greece. My grandmother was born in Veni, Rethymno, Crete, Greece. Veni is in North Crete (an island in South Greece which became part of Greece in 1913). It was ruled by Venetians for 464 years before the Ottoman Empire ruled Crete. Crete is considered a pirate island. The Sea People people are possible ancestors for all Cretans. Etruscans may have been Sea People. My DNA makes sense. I was expecting more Greek DNA though than 1.6% though. My DNA came out 29.4% standard estimate to 45.7% Italian speculative estimate. I was in Italy last year and green eyes lke I have are common. I blended in very well. Now I know why....lol.