I came across the
Maghreb Y-DNA Project and had a look at the R1b results. The presence of 1 to 5% of R1b in North Africa is still fairly mysterious due to the absence of detailed study on the subclades of R1b found in that region. This is why I was glad to find that one of the project members tested for subclades and happens to be R1b1b2a1b4c (U152+, L2+). I checked the STR values of other R1b's and they all seemed quite similar and could well be R-U152+. That is what I expected. In other words it means that Maghreban R1b is of Roman/Italian origin.
The presence of U152+ in the Maghreb also reinforce the hypothesis that the Romans carried this haplogroup, and that the U152 in Italy is not exclusively of Celtic origin.
Further evidence that R1b1b2 in Europe is neither of Paleolithic nor of Neolithic origin
If R1b1b2 had come to Europe during the Neolithic expansion, it would most likely also have reached North Africa during that time, based on the pattern of diffusion of agriculture along the Mediterranean (
see map). We would therefore expect to find a
trail from southern Turkey to Greece, Italy, the Maghreb and southern Iberia. The only common trail is the presence of E1b1b, J2 and G2a. R1b is present in all these regions, but in very different proportions and the subclades are completely different. U152+ is very rare in southern Turkey, minor in Greece, omnipresent in Italy, apparently the major type of R1b in the Maghreb but very low in frequency, and minor in Iberia despite the huge percentage of R1b.
A Neolithic spread of R1b would involve that all the
older subclades of R1b1b2 be found in the places settled earliest, and a new trail showing an northward expansion from southern to northern Iberia, southern to northern France and across the Channel to Britain. That is simply not the case.
The fact that U152+ appears to be the main subclade in the Maghreb is enough to rule out the presence of R1b among the Neolithic migrants that left Greece for Italy, North Africa and Iberia. Otherwise southern Spain and Portugal would also abound with U152+, but they don't.
If R1b1b2 originated in Western Europe during the Paleolithic or Mesolithic, the highest diversity of R1b subclades would be found in Iberia and southern France, and all subclades would radiate from there. Instead what we see is that the oldest subclades of R1b1b2 are found around Anatolia, the Caucasus and Central Asia, then all subclades from R1b1b2a1 radiate from the Alps, along with the diffusion of Indo-European languages.
The Italic language branch is the most distinctive in the Italo-Celtic group, and matches U152+. Once again, if R1b-U152 is the main type of R1b in the Maghreb, it fits the Indo-European migration model best because Tunisia and Algeria were heavily colonised by the Romans, the main carriers of U152+. The low overall percentage of R1b in the Maghreb (max. 5% + the Roman J1, J2, G2 and E1b1b, for a total of perhaps 10%) is also consistent with a colonisation pattern, rather than a massive migration or resettlement.
The Phoenician having settled in the Maghreb too, I would expect to find the typically Levantine type of R1b too (namely R1b1a, M18+).