Sara Tishkoff and company are leveraging their massive African data set again, with a paper out which surveys the diverse landscape of lactase persistence genetics across continent, Genetic Origins of Lactase Persistence and the Spread of Pastoralism in Africa. Many of the results were well known or could have been anticipated, selection for the trait in Africa, or even the fact that lactase persistence in the Khoikhoi derive from East Africa. So let me quote at length one section which I think is rather notable: In fact, the T-13910 variant in the Mozabite from Algeria occurred on the same haplotype background as observed in Middle Eastern populations, whereas the Fulani from Cameroon and the Bulala from Chad shared the same haplotype background with Europeans. The Fulani also shared a distinct T-13910 haplotype background with the Arabic Baggara. These patterns suggest that the distribution of observed haplotype variation might be due to gene flow that occurred over time from outside and within Africa possibly during key historical events, such as the settlement of the Roman Empire in parts of northern Africa and the expansion of the Arabs prior to and during the Ottoman empire within the last 2,000 years.67 These inferred migration events are also consistent with studies based on mtDNA, Y chromosome, and autosomal genetic variation.
The T-13910 variant is the one that is very widespread in Western Eurasia, from Northern Europe, all the way to Northwest India. Earlier analysis had suggested hat the Fulani people of the western Sahel carried the the European and South-Central Asia mutation for lactase persistence, as opposed to the ones more common in Eastern Africa or the Middle East. This was curious, but not totally incomprehensible in light of the fact that it seems that populations such as the Tuareg mediated gene flow across the Sahara, and the Berber polities extended to the northern Sahel. What is surprising is that upon closer inspection the flanking regions around T-13910 in the Fulani, but not the Mozabite Berbers, exhibit a signature which indicates a European origin. The Mozabite Berbers are a rather isolated group, probably representative of genetic variation in large part before the settlement of non-Berber populations along the coastal zone in antiquity. So it would not surprise me if T-13910 haplotypes in much of Morocco and Algeria are of the European, and not the Mozabite, flavor. But that still prompts the question, why are there two variants in the Maghreb when they are functionally the same?
A plausible model is that the European haplotype came along with the Vandals, a Germanic barbarian tribe with origins in southern Scandinavia and northern Germany which conquered North Africa in the early 5th century A.D. Of course the allele could have arrived in the centuries of earlier Roman rule, but the T-13910 mutation is generally found in low frequencies in Mediterranean populations like Italians. In contrast it is well over 50% among Northern Europeans. It is fashionable in studies of Late Antiquity to assume that the Vandals, and their Alan confederates, were ad hoc social constructs which somehow congealed out of the political compost heap of the late Roman Empire. An alternative view, championed by Peter Heather (see Empires and Barbarians), is that groups such as the Vandals were a genuine volk engaged in wholesale migration. The idea that the Vandals were social constructs of some sort, and of trivial numbers, suggests that they made only a marginal impact and disappeared in totality after the conquest by the Byzantines in the 6th century of the Maghreb. But the presence of these distinctive genetic markers indicates that the Vandal numbers may have been substantial enough to allow for the transfer of the T-13910 allele to the Berber tribes which assimilated the the rural estates they still held.
Much of this is speculation of course. We’ll know the truth of the matter when someone does extensive resequencing of the whole region of LCT in a variety of populations. At that point presumably one might be able to discern if the Fulani allele is closely related to German ones in particular. The point is that these sorts of bizarre scenarios which are more in keeping with late 19th and early 20th century adventure fiction might actually be closer to much of history than we would have thought.
Note: I should add that Greg Cochran speculated in this direction a few years ago.