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 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.
UPDATE II (Jul 2)
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.