I’ve been looking at some European genotype data. So I have some samples from Greece. One of the things I noticed is that there seem to be two clusters of Greece. You can see it above. The Italian sample is really a southern Italian one (not Sicilian though). The Balkan sample are Serbs, Bulgarians, and Romanians. You can see that they are shifted toward the Poles. And so are the Greeks, in comparison to the Italians. This is not entirely surprising. What was surprising to me was that there were a number of Greeks who in the same cluster at the Italians.
The historical context for this are the Sclaveni migrations. These were Slavic peoples who pushed south, as far as the Peloponnese, after the Byzantine Empire ceded the Balkans to barbarian groups due to threats in the east from Persians and then Muslims. In fact the demographic basis of the Byzantine Empire between the loss of the Levant and Egypt and the Battle of Manzikert was Anatolia in 1056, though there were fortifications around major cities such as Thessaloniki. After the loss of their Anatolian heartlands to the Seljuks the Empire turned back toward the Balkans, which had been conquered by Basil II in the first decades of the 11th century.
These results, and others, indicate the impact of the Slavic migrations on the Balkans, and Greece proper. But, what they also suggest is that there is population structure within Greece. Why? I can think of two hypotheses. First, some of the islands in the Aegean were never touched by Slavs, and may have maintained endogamy until the modern period. Even if the Slavs never conquered the cities, their impact would be felt by migration from rural areas. But in a pre-modern era barriers such as water and mountains often serve as potent obstacles to continuous gene flow. The second, to me more plausible, scenario is the second cluster without much Slavic genetic impact are those who descend from Anatolian Greeks, who arrived in the early 20th century due to the population exchange with Turkey. These western Anatolian Greeks would have shielded from the Sclaveni migrations obviously.
To tease the relationships apart I decided to run TreeMix 20 times. As per reader suggestion, I won’t give you all the plots. But you can download them. Below is a representative one. The various Jewish groups form their own clade. The affinity of Cypriot Greeks with Anatolians is a function I believe of the fact that they are culturally Hellenized (the ancient Bronze Age polity of Cyprus was part of the orbit of Egypt, and was not Greek), even if that is an ancient occurrence. I separated the Greeks into two cluster, the major one being “Greece” and the minor one clustering with southern Italians as “GreeceItaly.” What is pretty obvious is that GreeceItaly has much less of the Slavic admixture. In this tree the Greeks proper are placed near the Balkan and Polish position on the graph, but with a huge migration arrow from nearly the GreeceItaly position. The Balkan node has a smaller migration parameter. The Greeks tend to flip from being near the Poles to being near the GreekItaly cluster, and swapping the migration arrow direction.