Friday, January 7, 2022

The Bishop's Sea: Islands in the Helix


(The Canaries have a great deal of volcanic tuft that makes it relatively easy to dig out a cave sanctuary or necropolis, something that old time Canarians loved to do.)

The Omicron Variant isn't just a rejected Michael Crichton manuscript. It's also eaten my weekend! But I did want to post something today, and given the rate of typos in just the paragraph I've already written, it sure better be low effort. Fortunately, there's an interesting question that has been weighing on me. It's even tangentially related to an epidemic of swabbed rapid tests! Have we caught up with the ancestral genetics of the island Atlantic now that everyone is asking 23andMe to do their genealogy homework for them? We haven't.

A bit of googling turned up a Springer map of Atlantic islands that I'm practically morally obligated to steal. It's got a bit of cruft --we're not going to be doing the genetics of the Falklands here-- but it's got the islands I do want to cover. (Translation: It's got the Faeroes.)

Let's go!

The first Atlantic, as opposed to Scottish, archipelago, to be inhabited, the Canaries have been explored extensively by DNA surveys in recent years. In fact, I'm tempted to place the Canaries on some kind of pole of extremes in what Statistics New Zealand is pleased to call "inter-ethnic mobility." Canarians are fine with being Guanches, and, perhaps more importantly, DNA surveys bear this self-identification out. Wikipedia notes a 2003 survey that showed a 42--72% continuity between pre-Plantation and current Canarian populations. The survey also claimed to find a sexual skewing, with European descent more common in the male line compared with female. This, while classically seen as evidence of sexual inequality if not outright genocide, also does not seem to bother Canarians. Nor are links between ancestral Canarian populations and Amazigh Berber populations across the Channel in Morocco. However, some authors continue to stoutly insist that this relationship must be traced to the distant past. A significant share of Sub-Saharan ancestry, however, does seem to be reported with the same degree of clarity. 

Taken together, theories about population replacement in the Canaries can now be discounted. That being said, I am not sure how seriously they were ever held. (You can still sell this shit to the Daily Mail, but come on.)

--The second Atlantic archipelago now known to have been populated, the Faeroes, appear to have received an immigrant population from the British Isles. in the immediately post-Roman period, which perhaps is another clue that the story of the decline of the Roman Empire in the west is that of the collapse of an economically parasitic or simply self-destructive regime. Here, the Faroes are found to have one of the most homogenous and isolated of North Atlantic populations, "with the highest level of asymmetry in Scandinavian vs British ancestry among female and male settlers of the archipelago."

To put numbers to the claim, "the current study suggests that only about 17% of the female settlers of the Faroe Islands were of Scandinavian descent, whereas a much larger 83% had British Isles ancestry. Previous studies1920 suggested that 87% of the male settlers were of Scandinavian descent, with only 13% having British Isles ancestry. These results may, however, have been affected by the high level of genetic drift occurring in the Faroese population."

Remarkably, although the Faroes are seen as the jumping off point for the settlement of Iceland, interpretation of this data suggests that the Faroes are more genetically isolated of the two. The article at the link  makes some entertainingly  ad hoc distinctions between islands of the Atlantic which were more homey, and thus settled by families from Scandinavia, and more frontier-like, and thus settled by lone Scandinavians who then sent out for British mail-order brides. Clearly coastal northwestern Scotland belongs in the former category, Skye in the latter!

 Well, it is a long trip. 

Icelandic archaeologists are more resistant to the idea of an early settlement of the island, and while I have argued with this position elsewhere, here it is sufficient that everyone agrees that it was settled after the Faroes. Iceland's settling population has been characterised as mixed northern British ("Gaelic") and Scandinavian, with the same sexual skewing as the former two cases, although less extreme than the Faroese case. Interestingly, the proportion of Scandinavian ancestry has risen over time in the Icelandic population, from 56% at the founding to 70% today. Or it's not interesting at all, because it reflects the place from which the small number of immigrants came from, Native North American traces apart. Although Wikipedia, bizarrely enough, wants Scandinavian-descended Icelanders to have an "increased chance of reproductive success." I suppose once you've ruled out immigration a priori (that pesky paleo-Eskimo aside), Super-Vikings is what you're left with. 


The closest of the Macaronesia islands are the odd archipelago out, since we've no scientific groups currently championing an earlier settlement with really strong evidence. (Those northern European rodents could have got there on their own.) Compared with Portuguese, Madeirans have significantly higher Sub-Saharan ancestry and also a U6 mtDNA component which, not being identical to the one found on the Canaries and thus an indicator of post-Plantation gene flow from those islands, indicates a significant degree of Northwest African ancestry.   On a related question, however, this study rejects a significant contribution on the part of Corsair raiders. Hey-ho!


A study from way back in 2003, sampling only mtDNA, as was the practice in those archaic days, found that Azoreans are mainly Portuguese with a significant northern European admixture.  This raises the question of what Y-DNA studies might show. Sure enough, there are marked differences. "Haplogroup J* is the second most frequent in Azores (13.4%), but it is modestly representedin mainland Portugal (6.8%). The other non European haplogroups – N3 and E3a –, whichare prevalent in Asia and subSahara, respectively, have been found in Azores (0.6% and1.2%, respectively) but not in mainland Portugal (Neto et al., 2007)." The medical geneticists who published this study don't seem particularly interested in the historical implications of this, so let me do the hard work of googling to tell you that sampling in Oran, Algeria, returned 27.4% of the population with a J* Y-haplogroup (paternal) lineage.

Puerto Rico

The genetic history of the Caribbean is a bit much to get into on my current deadline of "Stop dawdling, you need bread for breakfast and lunch tomorrow!" but Puerto Rico is noteworthy for the particularly high Taino contribution to the current population. 

Wikipedia: "Studies have shown that the racial ancestry mixture of the average Puerto Rican (regardless of racial self-identity) is about 64% European, 21% African, and 15% Native Taino, with European ancestry."  

"Regardless of racial self-identity." Again, there is considerable skewing to European ancestry in male lineages.


The received population history of Newfoundland is that it was settled in the 1760s by roughly 20,000 Roman Catholic populations from southern Ireland and Protestants from western England, and had previously had small indigenous populations of Algonquin Micmacs, Beothuks and Innu. Protestant and Catholic populations did not intermarry, and the indigenous populations became extinct on the island of Newfoundland at an early date, with the Beothuks in particular having no close relatives elsewhere. A detailed genetic survey shows the Irish contribution has been overstated and that there is a significant indigenous admixture, especially considering that the small expected size of the indigenous population should have been so thoroughly dominated by an in-migration of tens of thousands of Europeans.


Modern Greenlanders are Inuit with some, mainly modern male, European ancestry, which is reflected in DNA studies. The predominance of European male lineages is interesting in that we are not currently invoking a genocidal event, although that Icelandic "superior reproductive success," which I have snidely implied is being  assumed as due to superior Scandinavian reproductive fitness, is still on the table. Presumably we want to substitute social factors for Viking super-genes, and we're good to go.  

No evidence for Paleo-Norse admixture is found, and evidence for Dorset admixture at low levels of gene flow is not definitive. The long term reader may recall my thesis that the Paleo-Norse were actually Dorset culture individuals undergoing ethnogenesis as people who didn't have to live on seal blubber. Drilling down a bit, eighty percent of Inuit Greenlanders have some European ancestry, but this varies greatly from region to region, with Tasiilaq, Qaanaaq (Thule), and small villages in southern Greenland reporting a smaller level of admixture, and this is used to reconstruct prehistoric migration routes and conclude that Inuit entered Greenland from the northwest and migrated down both west and east coasts, with East Greenlanders then migrating to south Greenland. 

So the Dorsets as Paleo-Norse hypothesis is for the moment salvageable, and the east-coast warm springs town of Tasiilaq remains an intriguing anomaly, the first place where Icelanders and Greenlanders interacted, per the saga literature, and also the Greenlandic community closest to Iceland, but with the least evidence of interaction. While this is not surprising given the ice conditions off Tasiilaq, the Medieval Warm Period explanation for the settlement of Greenland would do away with those floes and put Tasiilaq in the frontlines of interaction. (The hypothesis being that coasting voyages between Iceland and the then-Norse colonies would have been the primary means of communication between the two communities, and this would seem to imply a straight crossing of the Denmark Straits.)

In conclusion, the number of demonstrably different historical cases in which European ancestry became prevalent in male lineages but not female across various Atlantic islands suggests soft-pedalling theories of overt sexual domination in favour of ones in which soft social factors are in play. My personally preferred hypothesis is that it might reflect sex sorting in immigration, with single male out-migrants more common and more likely to assimilate in their partner's communities. I suspect that this trend could be demonstrated by YDNA distributions of non-hegemonic migrant communities. It seems like  the descendants of Lascar Londoners would be a good target to support the hypothesis by analogy, and I only refuse to pursue here at length because I am running out of time, having satisfied my own curiosity at some basic level.  

1 comment:

  1. Nice to see all this genetic evidence coming out!

    This might be of interest on earlier medieval sources for the Canary Islands: De historia atlántica.