August 22, 2011

Tunisian Y-chromosomes and mtDNA

A couple of recent papers.

AJPA DOI: 10.1002/ajpa.21581

Genetic structure of Tunisian ethnic groups revealed by paternal lineages

Karima Fadhlaoui-Zid et al.

Tunisia has experienced a variety of human migrations that have modeled the myriad cultural groups inhabiting the area. Both Arabic and Berber-speaking populations live in Tunisia. Berbers are commonly considered as in situ descendants of peoples who settled roughly in Palaeolithic times, and posterior demographic events such as the arrival of the Neolithic, the Arab migrations, and the expulsion of the “Moors” from Spain, had a strong cultural influence. Nonetheless, the genetic structure and the population relationships of the ethnic groups living in Tunisia have been poorly assessed. In order to gain insight into the paternal genetic landscape and population structure, more than 40 Y-chromosome single nucleotide polymorphisms and 17 short tandem repeats were analyzed in five Tunisian ethnic groups (three Berber-speaking isolates, one Andalusian, and one Cosmopolitan Arab). The most common lineage was the North African haplogroup E-M81 (71%), being fixed in two Berber samples (Chenini–Douiret and Jradou), suggesting isolation and genetic drift. Differential levels of paternal gene flow from the Near East were detected in the Tunisian samples (J-M267 lineage over 30%); however, no major sub-Saharan African or European influence was found. This result contrasts with the high amount of sub-Saharan and Eurasian maternal lineages previously described in Tunisia. Overall, our results reveal a certain genetic inter-population diversity, especially among Berber groups, and sexual asymmetry, paternal lineages being mostly of autochthonous origin. In addition, Andalusians, who are supposed to be migrants from southern Spain, do not exhibit any substantial contribution of European lineages, suggesting a North African origin for this ethnic group.

Link

Journal of Human Genetics , (11 August 2011) | doi:10.1038/jhg.2011.92

Mitochondrial DNA and Y-chromosome microstructure in Tunisia

Hajer Ennafaa et al.

Mitochondrial DNA (mtDNA) and Y-chromosome variation has been studied in Bou Omrane and Bou Saâd, two Tunisian Berber populations. In spite of their close geographic proximity, genetic distances between them were high and significant with both uniparental markers. A global analysis, including all previously studied Tunisian samples, confirmed the existence of a high female and male population structure in this country. Analyses of molecular variance analysis evidenced that this differentiation was not attributable to ethnic differences. Mantel test showed that, in all cases, Y-chromosome haplotypic distances correlated poorly with geography, whereas after excluding the more isolated samples of Bou Omrane and Bou Saâd, the mtDNA pattern of variation is significantly correlated with geography. Congruently, the Nm ratio of males versus females pointed to a significant excess of female migration rate across localities, which could be explained by patrilocality, a common marriage system in rural Tunisia. In addition, it has been observed that cultural isolation in rural communities promotes, by the effect of genetic drift, stronger loss of diversity and larger genetic differentiation levels than those observed in urban areas as deduced from comparisons of their respective mean genetic diversity and their respective mean genetic distances among populations. It is likely that the permanent exodus from rural to urban areas will have important repercussions in the future genetic structure of this country.

Link

20 comments:

Andrew Oh-Willeke said...

The first paper is a bit hard to evaluate without more information on sample size.

The second paper, while it makes sense, is a bit disappointing. The dramatic historic conclusion one might draw from geographic variation in mtDNA but not Y-DNA is a conquest and marry the locals scenario. But, patrilocality certainly seems to fit the bill just as well, has good ethnographic support and involves a lot less historical drama.

Onur Dincer said...

Patrilocality also explains why different ethnic groups of Anatolia are indistinguishable from each other in terms of mtDNA according to the results of the Anatolian genetics studies of Gokcumen et al (indeed this is how Gokcumen et al. explain the Anatolian genetic results in their relevant papers).

Onur Dincer said...

Andrew, patrilocality always comes hand in hand with conquest and marry the locals, they are complementary aspects of the same phaenomenon.

Anonymous said...

I think the the female haplogroups represent the basal population. Male haplogroups seem to have a tendency to swamp the pool, an effect that is not necessarily reflected in the autosomal structure.

So the "Andalusans" may indeed be from southern Spain but overlaid by the dominant Y-haplogroup of the area.

In Britain the swamping effect does not appear to come with an autosomal contribution, so I think this is long-term patronage/bias rather than conquest. It could be the same in Tunisia

Dienekes said...

Andrew, patrilocality always comes hand in hand with conquest and marry the locals, they are complementary aspects of the same phaenomenon.

Patrilocality is no such thing.

Onur Dincer said...

Patrilocality is no such thing.

In patrilocal societies of the pre-modern times there was a strong asymmetry between the male and female sides of mating. The status of women was little higher than goods. Women in general didn't have right to choose their mates, but men did, so it was men who determined whom to mate. After all, "romantic" love is a modern concept. When soldiers (always males) conquered a land, local women didn't have any right to resist to marrying a conqueror once she had been isolated from her male local protectors (by various means, including violence). Conquerors used to see the conquered lands as their new homeland, so they had no problem staying there permanently and marrying a local woman. So even though patrilocality and conquer & marry the locals weren't directly related, they always went hand in hand in pre-modern times due to the pre-modern parameters.

Dienekes said...

Onur you are conflating a lot of different concepts. Patrilocality is NOT necessarily an aspect of the same phenomenon as "conquest" and it is NOT necessarily related to how highly women are valued.

Onur Dincer said...

Onur you are conflating a lot of different concepts. Patrilocality is NOT necessarily an aspect of the same phenomenon as "conquest" and it is NOT necessarily related to how highly women are valued.

Dieneke, by the same phaenomenon I was referring to the dominant mode of pre-modern gender relationships and its social reflections. Social concepts generally come in packages and there is a good correalation between some seemingly unrelated concepts. Are modern societies patrilocal? We can't say they are. Modern societies are, IMO, plutolocal (staying where the wealth is).

sykes.1 said...

So, whatever happened to the Goths and the Vikings? Wasn't there a major influx of Goths in the 5th Century? And weren't the Vikings all over the place in the 10th plus centuries?

Anonymous said...

Not Goths or Vikings, but Vandals.
The Vandals founded a kingdom in ancient Carthage, which lasted several decades.
But the Vandals were too few, perhaps 80,000 people, and this shows how the genes of a small group of invaders is dispersed and disappears in a few generations, if the invaded country is heavily populated.

dalouh said...

"So the "Andalusans" may indeed be from southern Spain but overlaid by the dominant Y-haplogroup of the area. "
I disagree
the Andalusian ancestry was made up largely by folks who migrated to the urban areas,and in order to hide their Berber identity and origin or to gain some kind of 'prestige' ,many invented Yemeni , Hijazi , Hilalian Arab , Andalusian and even Turkish ancestries....of the thousands of Andalusian refugees who were expelled from Iberia, perhaps a few hundred families settled in Tunisia.

@ DagoRed
I believe that the number of the Vandals was a lot smaller,and I don't think they had any significant genetic impact in the region (the French stayed a bit longer,did they ever mix with the natives? ),not in the Berber heartland in any case...
the North African region with some significant European ancestry is Kabylia, Euro lineages in Kabylians (and also other coastal groups) are overwhelmingly of pre-historic origin,after all, one important fact to know about Kabylia,the only time in history that it was conquered by foreigners was in 1857, by a technologically superior European power.

Andrew Oh-Willeke said...

"The Vandals founded a kingdom in ancient Carthage, which lasted several decades. But the Vandals were too few, perhaps 80,000 people, and this shows how the genes of a small group of invaders is dispersed and disappears in a few generations, if the invaded country is heavily populated."

80,000 isn't necessarily too few. That number would have been substantial and certainly noticable relative to the existing population if it had lasted, particularly in a socially dominant position for a generation or two. But, from the abence of much of a strong population genetic signal in the vicinity of ancient Carthage, we can infer that the historical accounts that describe a combination of mass slaughter, mass imprisonment and deportation, and mass flight of the Vandal population from Carthage at the time of the Vandal state must be substantially accurate.

"[T]hey are complementary aspects of the same phaenomenon."

Not really. It is perfectly possible to have gradual genetic impact from patrilocality over short distances between people who are at peace with each other produce extremely different patterns of Y-DNA and mtDNA diversity. Patrilocality is observed in a very large number of societies (all the way back to the Neanderthals for that matter) including many societies that were at peace for centuries. Certainly, not all patrilocal societies have engaged in wars of conquest, and some matrilocal societies have experienced war although the historical record of warlike activities and in particular wars of conquest by matrilocal societies is contested.

My point of course, is that one can get the same genetic result with either punctuated processes (like conquest) or gradual processes (like peaceful patrilocality with neighboring villages) and that the ethnography in this case seems to support the latter.

Onur Dincer said...

Andrew, peaceful or at least non-belligerent coexistence of different groups doesn't mean that they will always remain that way nor they were always that way in the past. For pre-modern post-Neolithic times, there is always conquest and marry the locals in the past of every group that is numerically dominant in a large enough territory if the relevant territory was already populated enough before its conquest, I know of no exception to this rule. Both conquest and marry the locals and patrilocality serve to shuffle the mtDNA and the autosomal DNA of different groups (whether unidirectionally or bidirectionally), so it is very difficult (and often impossible) to separate the impacts of C&ML and patrilocality on mtDNA and autosomal DNA from each other.

Anonymous said...

The result about andalusians in first paper is quite surprising. Some context:

--Wikipedia entry on moriscos (andalusians), with some quantitative details: http://en.wikipedia.org/wiki/Morisco (good enough entry).

--A link with details of the Tunisian moriscos:
http://moriscostunez.blogspot.com/2011/08/huellas-de-los-moriscos-expulsados.html.

The moriscos/andalusians tested in this paper are all individuals from Zaghouan town, and therefore according to bellow comment from Valencia kingdom.

“Els valencians arribaren a Tunísia amb una identitat més forta que els altres moriscos de la península. Mantingueren eixa identitat i es quedaren a Zaghouan, Soleiman, i una bona part a Testour», així ens ho explica Raja Yassine Bahri, catedràtica de la Universitat de Tunis qui ens acompanya en este viatge per la geografia morisca al nord de Tunísia” on this link

http://mas.levante-emv.com/especiales/9doctubre/%C2%ABde-terra-endins-de-mar-enlla%C2%BB/

The sample is 32, and resuls are:
E around 43%, most (40%) E-1b1b1b.
F almost 10%
J 47% most (43%) J1

If we consider bellow propositions to be true:

Proposition 1: Pre-islamic Y-chromosomes (i.e. I, R,..) in all Iberia were different from Pre Islamic Y-chromosomes in Maghreb (i.e. E). Or at least there was very small presence of E derived from pheonicians, cartaginians...

Proposition 2: Most moriscos were direct descendants of pre-islamic populations in Iberia, that is hispano-goths converted first into islam and then (forced) to convert into christianity. That´s what spanish historiography asserts.

Proposition 3: Most moriscos in Tunez have kept their Y-chromosome along generations and researchers has made sure that these 32 samples came from Andalusian lineages and not just from people from Zaghouan.

Then we would expect to find european Y-chromosomes (I,R) in high proportion among andalousians in Tunez. But this is not what researchers found. Even if all F turned out to be european, only 10%is still surprising.

Andrew Oh-Willeke said...

A C&ML scenario almost necessarily entails a lack of numerical dominance.

We also know that the Vandals invaded with whole family groups (it was a folk wandering), not with predominantly male armies, from historical accounts, something that is a poor fit to the C&ML scenario.

mregdna said...

It was not only Vandals kingdom but Vandals and Alans and the proportion of Alans among them is not known.
If many were deported in anatolia and on the borders of Bizance territory, some seems to have taken refuge in the mountains of west tunisia and Kabylia.

I am surprised that there is no mention of the jews who were numerous in that region (writings of Saint Augustin and other christians). If many jews were converted to christianism and later to Islam a jewish component could be non neglectable.

Onur Dincer said...

We also know that the Vandals invaded with whole family groups (it was a folk wandering), not with predominantly male armies, from historical accounts, something that is a poor fit to the C&ML scenario.

But even in the case of folk wanderings conquering groups had to assimilate the bulk of the locals in order to attain numerical supremacy in a large enough territory. Vandals pretty quickly disappeared because that they couldn't assimilate the locals in large enough numbers in the regions they conquered, and were eventually assimilated by them. In pre-modern post-Neolithic times, if, as a conquering group of a large enough territory, you couldn't assimilate the locals, the locals would eventually assimilate you unless you lived a very isolated life like the Jews and Gypsies of the pre-19th century (but you could never attain numerical supremacy through isolation).

Onur Dincer said...

In pre-modern post-Neolithic times, if, as a conquering group of a large enough territory, you couldn't assimilate the locals, the locals would eventually assimilate you unless you lived a very isolated life like the Jews and Gypsies of the pre-19th century (but you could never attain numerical supremacy through isolation).

How a conquring group would remain as the ruling class and at the same time isolate itself from the locals is another issue. Such a scenario is, even if not impossible, very implausible; it is as if an unseen force is ruling a country from afar.

frenchy said...

Another study by the same author about North African Y-DNA from Sousse (Tunisia):

Karima Fadhlaoui-Zid 2015, "Sousse: extreme genetic heterogeneity in North Africa".

http://www.nature.com/jhg/journal/v60/n1/full/jhg201499a.html#fig2

It is quite interesting as this is quite a large sample (220) from a costal city.

Total E: 56% and J: 34%

N=220

A-M91 0.45%
DE*-YAP 0.91%

E1b1b1a1*-M78 0.91%
E1b1b1a1b-V13 0.45%
E1b1b1a1c-V22 0.45%
E1b1b1a1d-V65 4.09%
E1b1b1b1a2*-M183 44.55%
E1b1b1b2a-M123 1.82%
E1b1b1-M35 1.36%
E1b1b-M215 0.91%
E-M96 2.27%

G-M201 0.45%

J1b2*-P58 0.45%
J1b2b*-L147 1.82%
J1b2b1*-L222.2 23.64%

J2a1h2a-DYS445=6 0.45%
J2a1h-L24 5.45%
J2a1-Page55 0.91%
J2a-M410 1.36%

L-M20 0.45%

R1a1a-M198 0.45%
R1b1c-V88 0.91%
R-M207 4.09%

T-M184 1.36%

raffi17 said...

If we only take into accounts non-Berbers samples analysed so far from Tunisia (
Semino 2004, Onofri 2008, Grugni 2012, Fadhlaoui-Zid 2011, Alvarez 2014, Fadhlaoui-Zid 2015)

we get regarding J1 percentages :

Tunisia (73) :30.10%
Tunisia (52) :34.60%
Tunisia (Sousse) (220) :25.90%
Tunisia (Tunis) (148) :32.40%
Tunisia (Tunis) (33) :24.20%
Tunisia (Zaghouan) (32) :43.80%
Tunisia (Sejenane) (47) :34.00%

which gives an average of 183/605 = 30.2 % J1

Berbers groups who represents not more than 10% of the Tunisian people have on average less than 5% of J1.

So we can reasonably estimate that among 11 million Tunisians, at least 25% belong to J1.