Agree. It was just one of my "generalistic" reasonings remarks. I was not speaking of the effects of mountains on highlands vs plains, but of the general effect of mountains on circulation as a whole, in ancient Balkans (valuable for ancient Italy). I agree that from a relatively homogenous mixed pop can be born by some geographic barriers some striking differences that concern only a little part of the genome of the "children" subregional pop's. Nothing new.
I agree with that. Mountain ranges can, if not block gene flow, lessen it, or direct it in certain directions. In Italy, the easiest entrance from the north is by skirting the Alps on the east, which ten feeds into the Veneto plain. Peoples coming from central/eastern Europe could flow both into Italy and the Balkans.
Gene flow from the Balkans is also possible by that route, but in addition one can see the Balkans across the Adriatic in some places, and in others there are small islands in the Adriatic which act as "stepping stones".
There's also gene flow through navigation of the Mediterranean, of course. By the time of the Romans it was easier to go from Anatolia to Greece and then on to Italy, but even in the Neolithic one could navigate along the shore. That's how the farmers made it to Spain from Italy, after all.
Still, all in all, looking at the IBD analysis in particular, it wasn't as easy for other groups to massively change the genetics of Italy, not as easy as was the case for central/Eastern Europe, where it's one flat plain for thousands of kilometers. Which military leader was it who said it was perfect for tanks? Add to that the large population numbers and one can see why Southern Europe has had a quite different genetic history than Northern Europe, as Ralph and Coop point out.
Spain is in much the same position, except that there was a deliberate policy of population re-settlements, and not as many internal barriers, and so the population isn't as structured as the Italian one. What people often fail to recognize is that there's not only the Alps in Italy to protect against massive migration, but there's the Apennines running north south, and creating differences between the eastern and western areas.
Add to that the fact that the south was separated from the north/center since the fall of Rome, and that the central north was divided into Papal states and competing city states, and one can see why there is so much variation in Italy, far more, actually, in the north/central regions than in the south. So, when other groups say they are "near" or "similar to" Italians, the question is, which Italians "specifically", and you'd better have a lot of Italian samples even from one province, because there's going to be a difference.
My Magra River Valley flows for 62 kilometers from the Apennines to the Mediterranean. I can see the difference in appearance of the locals by northern part, middle part, and Mediterranean part. Now, of course, we also have Southern migrants, so it's more confusing, but my point is that it would seem incredible that there could be that kind of stratification in such a small distance, but there is. The reason? On the sea coast, you're going to have more admixture, largely from Toscana but also further afield. Up in the foothills, people don't move very far from their home villages, and don't trust people not from their home villages. "Moglie e buoi dei paesi tuoi" is a famous rhyming saying. Cows and wives from your own villages. Added to that, this small area was divided into three separate sub-areas, ruled by three different governments, with passes required for movement in some cases: Modena, Toscana, and Genova.
Such is the Italian history which people should know before trying to draw big conclusions from genetic data.