Humans are not static organisms, they migrate quite often, and they also move together with their cultures. Great effort has been expended in archaeology to reconstruct ancient human migrations and many skeletal remains were used for such a task. Unfortunately, it became apparent that bones are not a very suitable material with which to study past migrations, or other demographic phenomena in general, as several intervening factors are incessantly in play in the human skeleton. In particular, biological adaptations to new environments or new lifestyles and admixture can modify enormously the morphology of both migrant and receiving populations.

Nevertheless, archaeology has recently attempted a new way to disentangle past migrations – archaeogenetics. One of the most important aspects of this approach is the possibility to locate and date ancient migrations through the study of uniparentally inherited genetic lineages present in contemporary human populations. DNA in mitochondria (mtDNA) or the Y chromosome is a particularly convenient tool for this kind of investigation as it is inherited through one sex only (that is, it does not recombine during reproduction) and its diversity is therefore accumulated exclusively by new mutations producing different DNA sequences (called haplotypes) in different geographic regions and over time. Archaeogenetics, based on contemporary genetic lineages, reconstructs haplogroups (groups of related haplotypes) and, according to their geographical distributions, can infer their common ancestors (founders). Thanks to a known mutation rate, age estimates can also be provided for each of the haplogroups. In fact, the spatial distribution of ancestral vs. derived genetic lineages together with age estimates mirrors the direction and chronology of past migrations.

Until a relatively recent period, all humans lived as hunter-gatherers closely dependent on environmental conditions and, therefore, experienced several demographic expansions followed by contractions associated with changes of climate and the availability of food resources. Logically, in periods when more favorable conditions prevailed, some fortunate genetic lineages increased numerically – an important prerequisite to favoring a new genetic lineage. However, as time passed the founder lineages acquired quite naturally new mutations and the process of haplogroup formation was started. The association of climatically favorable periods with genetic age estimates is a really exciting issue when linking the archaeogenetic and paleoecological results in archaeology.

The tracing of past human migrations through genetic lineages needs, undoubtedly, very careful sampling of autochthonous groups, avoiding recent migrants. Only this prerequisite can provide plausible results. Also, geographical scale is important. For example, the absence of the mtDNA haplogroup M in Western Eurasia, and conversely its high frequency and high variation in Southern Asia, indicate that modern humans must have exited Africa some 70,000 years ago via Bab el-Mandab and not via the Levant. On the other hand, the sampling of much smaller geographic regions can reveal some younger mtDNA lineages mirroring later migrations.

The L3f is a relatively common mtDNA haplogroup that is widely dispersed, especially in East Africa, with very ancient roots. However, in the Lake Chad Basin, we have recently discovered its daughter clade L3f3 with and age estimate of some 8,000 years ago. Interestingly, this specific clade was not present in all inhabitants of the area, but almost exclusively in Chadic speakers (a branch of the Afro-Asiatic linguistic family). Further analyses revealed close genetic relationships between the Chadic groups living today in the Lake Chad Basin and some East African Cushitic populations. These findings point to the past migration of a proto-Chadic pastoralists to Lake Mega-Chad (at that time the lake was more than 20 times larger than today), as has been suggested by some linguists. We suppose that later desertification of the area lead to the isolation and sedentarization of the proto-Chadic speakers and at the same time provided favorable conditions for the formation of L3f3 within their population. After all, the archaeological finds along Wadi Howar in Western Sudan can be taken as a physical proof of this ancient migration heading to the lake.

Another interesting example of Late Pleistocene human expansion, subsequent isolation and Holocene co-evolution of mtDNA and languages can be demonstrated by our study of the haplogroup R0a. This haplogroup is abundant throughout the Near East, but we have identified its daughter clades (such as R0a1a1 and R0a2f1) that are restricted to Southern Arabia where they must have diversified some 6,000 years ago. Close correlation of the above-mentioned genetic lineages with South Arabian (and non-Arabic) languages is as striking as the case of the L3f3 and Chadic speakers in Africa. Interestingly, some age estimates for the South Arabian language diversification indicate similar time periods as those we have obtained by the genetic study of these haplogroups. Moreover, the geographic location of these haplogroups coincides well with reconstructions of past population refugia reported by independent paleoenvironmental and archaeological investigations.

The above mentioned genetic and linguistic features revealed by our regional sampling of the autochthonous populations points to the fact that their ancestors must have been isolated. They probably established small-scale reproductive units in the Late Pleistocene and re-expanded from these refugial pockets when climatic conditions allowed wider expansion. In fact, we can witness a relatively rare interference of social and biological phenomena – the beginning of a fast-evolving linguistic diversification and the ending of a slow-evolving genetic one. We believe that such a co-evolution can be possible only as a result of population isolation in the past.