Saturday, October 5, 2013

French Canadians: The unexplained genetic diversity

Interior of a magasin général (source: photographiquement Frank). Wherever there was less competition from British or American merchants, it was easier for French Canadians to go into business. These same regions also have unusually high rates of neurological disorders, including Tay-Sachs. Coincidence?

French Canadians have a unique demographic history. From a founding population of some 10,000 settlers who came in the 17th and 18th centuries, they today number over ten million in Canada alone. This rapid growth shows what humans can do when war is absent, land plentiful, and government reasonably good.

Over those past three centuries, French Canadians have diverged from their cousins in France, not only culturally and politically but genetically as well. This is the finding of a recent study:

[…] in less than 20 generations of genetic isolation from the French population, the genetic pool of French-Canadians shows reduced levels of diversity, higher homozygosity, and an excess of rare variants with low variant sharing with Europeans. Furthermore, the French-Canadian population contains a larger proportion of putatively damaging functional variants, which could partially explain the increased incidence of genetic disease in the province. (Casals et al., 2013)

The authors try to explain this genetic divergence in terms of two causes: (1) a genetic bottleneck that allowed in only a small sample of France’s gene pool; and (2) relaxation of natural selection due to better living conditions, with a consequent accumulation of new genetic variants, i.e., people survived and reproduced who otherwise wouldn’t have back in Europe.

Clearly, there was a bottleneck. This is obvious from historical records and also from the degree of genetic homogeneity:

Compared to French individuals, French-Canadians have lower levels of heterozygosity (on average 19.2% and 11.5% of the variants per individual are heterozygous in French and French-Canadians, respectively) and have lower average nucleotide pairwise diversity. (Casals et al., 2013)

Despite this relative homogeneity, rare genetic variants are more frequent in French Canadians than in French people from France: 

The French-Canadian population also exhibits an excess of low frequency variants in comparison to the French population, and the proportion of variants with MAF [minor allele frequency] less than 5% is significantly higher in the French-Canadian population (p less than 0.01). The excess is not a consequence of different sample sizes; if we resample the same number of individuals from each population and include only sites where all individuals pass identical quality filters, we observe a similar excess of rare variants in the French-Canadian population compared to the French population. (Casals et al., 2013)

Why would French Canadians be less diverse for common alleles but more diverse for rare ones? The authors suggest relaxation of selection. As new mutations appeared in this new population, there was less selection than in France to weed out the bad ones.

Even so, the frequency of minor alleles seems to be higher than can be explained by relaxation of selection alone: “the shift observed in the empirical data, which shows an increase of 9.8% of variants with MAF less than 5% in the French-Canadian population compared to the French population when using the same sample sizes […], is larger than that generated by simulations” (Casals et al., 2013). The authors suggest that current theoretical models might underestimate the effects of relaxed selection, particularly under conditions of rapid population growth. Or perhaps this unexplained genetic diversity has some other cause. 

That other cause might be selection pressures that developed in some French Canadian subgroups and not in others. The authors themselves seem to raise this possibility when they go on to state that “the French-Canadian population is genetically stratified into subpopulations with differentiated demographic histories” (Casals et al., 2013).

Demographically and historically, there was a difference between frontier regions, where land was more available, and settled regions, where it was less so. Moreau et al. (2011) have shown that settlers on the wave front of colonization enjoyed a reproductive advantage over other French Canadians and thus contributed more to the current gene pool. Such settlers also seem to have been selected for higher fertility, according to a study of one French Canadian community. Between 1800 and 1940, the community of Île aux Coudres saw its mean age of first reproduction (AFR) fall by four years, not through a lowering of the mean age of marriage (which remained stable) but through a shortening of the mean interval between marriage and first birth. The study’s authors rule out better nutrition as a possible explanation, since infantile and juvenile survival rates should have likewise improved, yet they failed to do so during this period (Milot et al., 2011). 

There was also a demographic and historical difference between regions where French Canadians co-existed with American or British settlers and regions where they formed almost the entire population. In the first type of region, English-speakers dominated occupations that required business skills, i.e., numeracy, literacy, merchandising, bookkeeping, etc. In the second type of region, which covered most of eastern Quebec, these mercantile niches were much more open to any suitable French Canadian.

It is also in the second type of region that we see unusually high rates of neurological disorders. Of the top ten disorders in eastern Quebec, six are primarily neurological and two secondarily so. By comparison, only three of the top ten genetic disorders in the United Kingdom are primarily or secondarily neurological (Frost, 2012). These eastern French Canadian disorders also include two forms of Tay-Sachs: one on the south shore of the St. Lawrence and another on the north shore. In the town of Rimouski, the heterozygote frequency of Tay-Sachs is 7.6%, versus 4.2% among Ashkenazi Jews and 0.3% among French Canadians in Montreal (De Braekeleer et al., 1992). Relaxation of selection alone cannot account for this high incidence, nor can it explain the emergence of two different forms of Tay-Sachs within the same geographic area.

Instead of relaxation of selection, these eastern Quebec disorders may attest to intense selection for individuals who could exploit mercantile niches that were less available to French Canadians elsewhere. These niches were much more numerous than might seem from the relatively few people who identified themselves as ‘merchants’ to census takers. There was in fact a much larger number of ‘farmers’ who nonetheless had mercantile sidelines of one sort or another (note #1 and Frost, 2012). In eastern Quebec, the payoff for going into business, and the requisite mental skills, was so lucrative that it may have favored alleles, like Tay-Sachs, that seem to improve mental processing in the heterozygous state while being deleterious in the homozygous state (Cochran et al., 2006).

This last point bears repeating. What is harmful in the homozygote may still be adaptive in the far more numerous heterozygotes. A high prevalence of genetic disorders might just mean that a very strong selection pressure was acting for a very short time … and that there was nothing better to work with. Natural selection does its best with whatever genetic variants are available. Eventually, through random mutation, there will appear more suitable variants that provide the same benefits with fewer adverse effects, and these genetic solutions will replace the ‘rough-and-ready’ ones that had initially been favored. 


French Canadians are genetically unique for several reasons. First, they descend from a small founder group. Second, they had high rates of population growth over a period of some three centuries. Third, they adapted to a very different environment and a very different set of selection pressures. In France, their ancestors lived almost under steady state conditions: land scarcity, small family size, late marriage for men and women alike, rigid class distinctions, and limited geographic mobility. In Quebec, those limitations were weaker if not absent altogether. Because land was much more plentiful, the opportunities were accordingly greater for marrying younger and having larger families. This freer and less limiting environment foiled attempts to transplant feudalism to Quebec even during the French Regime, with the result that the term paysan never caught on. People called themselves habitants. Finally, in comparison to other French Canadians within the same region, business-minded individuals had more chances for economic betterment in those areas, like eastern Quebec, where competition from American or British merchants was relatively weak.


1. Gustave Papillon, the founder of one of Quebec’s leading cement plants, started off as a poultry producer and probably identified himself as such to census takers. Yet he also had two sidelines: installing elevators and escalators; and reselling pharmaceutical products to local drugstores. His decision to start up a cement plant grew out of this pre-existing entrepreneurial mindset (Gadbois and Papillon, 2013).


Casals, F., A. Hodgkinson, J. Hussin, Y. Idaghdour, V. Bruat, T. de Maillard, J-C. Grenier, E. Gbeha, F.F. Hamdan, S. Girard, J-F. Spinella, M. Larivière, V. Saillour, J. Healy, I. Fernandez, D. Sinnett, J.L. Michaud, G.A. Rouleau, E. Haddad, F. Le Deist, and P. Awadalla. (2013). Whole-exome sequencing reveals a rapid change in the frequency of rare functional variants in a founding population of humans. PLoS Genet 9(9): e1003815. doi:10.1371/journal.pgen.1003815 

Cochran, G., J. Hardy, and H. Harpending. (2006). Natural history of Ashkenazi intelligence, Journal of Biosocial Science, 38, 659-693. 

De Braekeleer, M., P. Hechtman, E. Andermann, and F. Kaplan. (1992). The French Canadian Tay-Sachs disease deletion mutation: Identification of probable founders, Human Genetics, 89, 83-87.

Frost, P. (2012). Tay-Sachs and French Canadians: A case of gene-culture co-evolution? Advances in Anthropology, 2 (3), 132-138.

Gadbois, M-E. and L. Papillon. (2013). Gustave Papillon. Building Ciment Québec, Ciment Québec.

Milot, E., F.M. Mayer, D.H. Nussey, M. Boisvert, F. Pelletier, and D. Réale. (2011). Evidence for evolution in response to natural selection in a contemporary human population, Proceedings of the National Academy of Sciences (USA), 108, 17040–17045.

Moreau, C., C. Bherer, H. Vezina, M. Jomphe, D. Labuda, and L. Excoffier. (2011). Deep human genealogies reveal a selective advantage to be on an expanding wave front, Science, 334, 1148–1150.


Sean said...

The St. Lawrence river and Saguenay were a major conduit for Coureur des bois who were traders, sort of. There was alogging boom that started during Napoleonic wars, loggers floated timber down the Saguenay and Saint Lawrence to the gulf. A clever man could make big profits in a magasin général on that part of the St. Lawrence river and Saguenay. "the French colonist enjoys what he has, and often shows off what he has not". Presumably the traders and loggers spent freely after getting paid.

I wonder about the role of high fertility motivated by a desire for "unpaid family labor" in the family business of the general store, tavern or innkeeper though.

Renato Ferreira da Rocha said...

Why anglophone competition was deletery? Was it because of some innate advantage (genetic, cultural)?
Or because of political pressure (prejudice)?

Sean said...

The populations of Charlevoix, Saguenay and Lac-Saint-Jean, have one Tay Sachs mutation and across the river Bas Saint-Laurent has a separate one. These areas together must be only a few percent of the French populations than underwent similar expansion. Based on the Casals explanation there should be lots of other parts of Quebec or non-French Canada with extremely high rates of rare mutations resulting from settlement expansion.

There are areas where Highland Scots were the pioneer settlers. The settlement of Glenngary county by a founder stock of closely related members of clan McDonnell should have had some founder effect for example. So where are the other areas with rates of a mutation that are comparable to Tay Sachs in Rimouski?

Anonymous said...

My grandmother was French Canadian. I have studied her geneaology all the way back to 1620 in Quebec. Girls married at age 12 and had a child every second year. This was the norm. Cousins married cousins. Small villages did not offer a lot of choises in a mate.

Anonymous said...

Ten million Quebecois in Canada alone is simply mendacious. In 2011 la Belle Province accounted for 23.6% of the country's population of ~32 million; ~7.5 million of which 77% are mother tongue French...~5.8 million (elevated by French speaking immigrants from Haiti & N.Africa). There is not another 4.2 million descendants of the founding French population residing outside Quebec in Canada. Thus the Quebecois population, despites its allegedly rapid growth falls well short of the estimated Ashkenazim population of Central Europe (8-10 million) at the end of the 19th century...a miracle by many accounts considering the war, persecution, plague and pogroms (allegedly 500,000 died during the Cossack uprising alone) suffered by that population. Even the Puritans who, according to MacDonald doubled their population every generation, amounting to six million at the end of the 19th century, did not outstrip the stunning reproductive rates of the Ashkenazim of Central Europe.

Anonymous said...

There is also evidence of crypto Jews in New France and the Acadian population of Louisiana.

Peter Fros_ said...


The Coureurs des bois tended to spread their genes among the Amerindians (who suffered population decline because of disease and their inability to adapt to settled society). Farmers were the ones who had lots of children, who in turn had lots of children.

As the land became settled, many young men remained unmarried because they were "sans terre" and unable to support a family. This was why it became important to develop business sidelines of one sort or another.


English Canadians say that French Canadians were less interested in business. French Canadians say that English Canadians excluded them from business.

The truth is somewhere in between. Business was looked down upon in traditional French Canadian culture. The ideal was to have your own piece of land and be a farmer. Even if you did other work on the side, you would still call yourself a 'farmer' (i.e., cultivateur or habitant). If you were educated, you would become a doctor or a lawyer. Either way, business was seen as something rather underhanded or sleazy (and often it was).

On the other hand, English Canadians imposed their language and culture within business to the detriment of French. English was overwelmingly the language of trade and commerce until the 1970s. English Canadians also tended to hire people they knew (who tended to be other English Canadians).


The difference is that English Canadians dominated business elsewhere in Quebec. So population growth was driven by farmers (and later by working-class people who worked at English Canadian factories).

English Canadians tend to move around a lot, so it's harder to trace their genetic/demographic history. Glengarry used to be ovewhelmingly Scottish, but now Scots are in the minority there.


Yes, but those were the French Canadians who left descendants, like yourself. Something like 20% of each generation failed to have any children.


In 2006, Canada had 10,066,290 individuals of French descent. By province, they break down as follows:

Alberta: 667,405
British Columbia: 720,200
Manitoba: 206,360
New Brunswick: 380,915
Newfoundland and Labrador: 241,470
Nova Scotia: 368,940
Northwest Territoires: 6,050
Nunavut: 1,175
Ontario: 2,768,870
Prince Edward Island: 52,350
Québec: 4,474,115
Saskatchewan: 172,365
Yukon: 6,075


The French Canadian Tay-Sachs mutation is structurally different from the Ashkenazi one.

Matt said...

With Tay-Sachs, I wonder if the explanation might be that a general collapse in selective pressures for health (clean air, low population density, plenty of food, etc.) causing greater mutational load combined with maintaining the same pressures for IQ.

Increased mutational load through low health people surviving sets the genetic IQ lower, overclocking balances this by setting it higher - thus relatively high frequencies of cognitive boosters in French Canucks, without much in the way of particularly impressive IQ.

Or it could be something like this happened - - "As human population grows, disease-causing genetic mutations per individual increase, but each mutation is less harmful, when compared with a population that is not growing, says a Cornell study to be published in November in the journal Genetics. .... The most-asked question regarding such population genetics studies is whether people are becoming less healthy or fit based on the findings. The researchers measured fitness and found that the two main results – that there are more mutations, but each mutation is less deleterious – "exactly balance out," Gazave said.". French Canadians grew fast, so high numbers of disease-causing mutations per individual, but they are not any more harmful on net. Whether or not someone like Greg Cochran (who knows about genetic theory, unlike me) would trust that result, is another matter...

Reader said...

Slightly off-topic, but potentially an interesting topic for future discussion:

Here's today's article in the New York Times about new discoveries pertaining to the genetic origins of Jewish Ashkenazi (European Jews):

The latest studies confirm that, as believed, the Ashkenazi are descended from Jewish men but indigenous European women who married them.

Reader said...

Peter, you mentioned something called "relaxation of selection pressures." In the past you've discussed the very challenging (for males) marriage market in the West (particularly in America) due to high sex ratios with a shortage of single women, and contrasted it with the Eastern European/Russian market where men have an easier time getting married.

I'm curious, are there any repercussions of the "relaxed selection pressures" for men in Russia? Is that why, for example, American men tend to be better-looking physically, more successful politically and economically, and leaders of the most prosperous nation on earth, while Eastern Europe is mired in poverty, corruption, and lack of economic might?

Peter Fros_ said...


With French Canadians, we don't see an increase in the mutational load across the board. It seems to be concentrated in the area of neurological diseases. We see a similar phenomenon with Ashkenazi Jews, the mutations tend to be overrepresented in the lysosomal storage pathway.

And French Canadians certainly benefited from lots of space and clean air!

The second hypothesis you refer to is relaxation of selection due to release of constraints on population growth.


I've only read the NYT article, so I'm not sure about the methodology. But it does support a scenario of small numbers of Jewish traders setting up trading posts in eastern Europe between the 8th and 12th centuries and marrying local Slavic women. Dr. Richards sees this happening, however, at least two thousand years ago. I have trouble understanding that kind of time depth because there were no Ashkenazim back then.


The North American marriage market was like the Russian marriage market as late as the 1960s. Back then, single women outnumbered single men at all reproductive ages. Large numbers of never-married women were "left on the shelf." That's too recent for there to be much genetic divergence between North America and Russia.

Eastern Europe has lagged behind Western Europe for a long time. A big reason: the Mongol-Tatar depredations, which hindered growth of large settled populations, pacification of social relations, development of trade, etc.

When I talk with Russians, they say that the marriage market is becoming worse for single men over there, largely because of growing social acceptance of divorce and single motherhood. Single Russian women are becoming, on average, older and tend to have children.

Sean said...

Jews, 827 BC.
There has been some dilution; nonetheless, what is distinctive about Ashkenazic Jews goes back a long way.

Sean said...

OT, what do you think of Bruno Latour?

Sean said...

Not so OT perhaps.

Anonymous said...

Peter:With French Canadians, we don't see an increase in the mutational load across the board. It seems to be concentrated in the area of neurological diseases. We see a similar phenomenon with Ashkenazi Jews, the mutations tend to be overrepresented in the lysosomal storage pathway. discussion of the study which kicked all this off- "What is clear is that for the derived allele (i.e., mutations from the ancestral state) distribution in exonic regions of the genome French Canadians are much more skewed toward the low frequency portion of the spectrum than French proper. This skew is more noticeable for deleterious mutations, such as nonsense and missense mutations (nonsense mutations usually produce nonfunctional protein, while missense mutations may alter the nature of the protein in some specific detail through amino acid substitution)."

That seems pretty much like generalized mutational load across the board.

Peter Fros_ said...


Bruno Latour is correct in his assessment of the current state of science studies. Instead of engaging in a genuine critique of modern science, most academics use this field as a way to make science conform to their own political views.


The French Canadian mutational load does not exist across the board. It's concentrated in the area of neurological diseases. I don't think this reflects some kind of diagnostic bias, since the distribution is much more balanced in genetic diseases reported from the United Kingdom.