On the genes, genealogies, and geographies of Quebec

[Capitalis]

I went looking for papers that dealt with genetic drift in Colonial European-descended North Americans. This was the best paper that I found, but I link the other papers at the end of this post.

Luke Anderson-Trocmé et al., On the genes, genealogies, and geographies of Quebec (2023)

Quoted from p. 1

This study takes advantage of a population-scale spatially labeled pedigree (or spatial pedigree) compiled from more than 4 million Catholic parish records in the province of Quebec.

Using genotype data for 20,451 individuals and newly created pedigree-aware simulation tools, we provide a detailed spatio-temporal model of genetic variation at scales ranging from tens to thousands of kilometers. By including French and British individuals in our analyses, we assess how much ancestral population structure has been preserved from these two populations.

We highlight the relationship between river networks and genetic similarity as the past four centuries of European colonial history has been marked by rapid frontier expansion, beginning along the shores of the St. Lawrence River and eventually expanding up its tributaries. By tracing the genealogical ancestry of millions of individuals across space and time, we describe a constellation of distinct founder events arranged along geographic features that defined transportation and economic activity.

Quoted from p. 1, Results: Regional distribution of genetic variation

Quebec, a province in Canada, has a population of 8.8 million individuals, of which ~6.5 million speak French as a primary language. Most individuals in Quebec derive ancestry from ~8500 settlers who migrated from France in the 17th and 18th centuries—we refer to these individuals as French Canadians. The first 2600 French settlers contributed two-thirds of the French Canadian gene pool (18).

French settlers occupied territory inhabited and used by First Nations for thousands of years (19). Despite folk histories implying large amounts of Indigenous ancestry among French Canadians (20), genetic and genealogical studies show that French Canadians born in Quebec carry on average <1% of ancestry tracing back to Indigenous populations, with the rest being mostly attributed to French ancestry (21).

Catholic marriage records in Quebec have been digitized and combined with civil registry data by the BALSAC Project (22). Because of a strong historical correlation between French ancestry and Catholic religion, the pedigree reconstructed from these records is particularly complete among French Canadians (22).

We therefore focused our genetic and genealogical analyses on 20,451 individuals inferred to be French Canadian from 21,068 individuals among the CARTaGENE [12,064 (23)] and Genizon cohorts [9004, first reported here; see (24) for details on cohorts and ancestry inference]. We used principal components analysis (PCA, fig. S1) and uniform manifold approximation and projection (UMAP, Fig. 1A) (25, 26) to visualize genetic variation in the population.

A schematic summary of the entire French Canadian spatial pedigree and the geographic location of 4882 individuals linked to the spatial pedigree are shown in Fig. 1, B and C. Visual inspection shows strong correlation of genetic and spatial proximity and suggests that gradients in genetic variation coincide with geographical barriers and conduits such as the St. Lawrence, Saguenay, and Chaudière rivers or the Laurentian and Appalachian mountains [...].

Quoted from p. 2, Fig. 1. French Canadian genes and genealogies mirror Quebec’s geographies


Quoted from Supplementary Materials p. 20, Fig. S1: Principal Component Analysis of French Canadians


Quoted from p. 1-2, Results: French ancestry uprooted

Analyses of genealogical records show that most French settlers arrived in Quebec City and came from regions in western (Aunis, Poitou) and northwestern (Normandy, Perche) France, as well as densely populated Ile-de-France [see (27, 28) and table S2]. However, successive waves of migration related to military and colonization objectives had different origins and demographics (28).

To assess whether this history is reflected in the Quebec population structure, we compared the genomes of individuals living in different regions of Quebec and France. In agreement with historical records, French Canadians share more-recent ancestry with individuals from western France as measured by DNA that is identical by descent (IBD) (fig. S2A). We also observe more-modest variation across Quebec regions in their IBD with France, with regions of Saguenay–Lac-Saint-Jean (SLSJ), Beauce, and Bas-Saint-Laurent having slightly elevated values, likely because of founder events (see fig. S2B and discussion in the supplementary materials).

We used F4- statistics to assess whether genetic differentiation between any pair of Quebec regions correlates with genetic differentiation between any pair of European regions (seven French regions and Britain) and found no such correlation (fig. S3). This suggests that most present-day structure among French Canadians is independent of ancestral structure or differential contributions by French and British founders at this broad scale.

Quoted from Supplementary Materials p. 21, Fig. S2: Rates of identity by descent in different regions


Quoted from Supplementary Materials p. 22, Fig. S3: Comparison of Quebec, French and British regions using F4-statistics


Quoted from p. 5, Results: Historical migrations in space and time

To follow the formation of regional substructure in the French Canadian population, we defined three regions on the basis of watersheds with large proportions of individuals driving the top principal components (fig. S13). These empirical regions closely align with the SLSJ, Bas-Saint-Laurent, and Beauce regions defined above, so we reuse these labels to refer to the watershed-based empirical regions.

[...]

SLSJ [realized kinship (l) = 0.005; Fig. 4, A, D, and G] has a dominant early founder event in Baie-Saint-Paul (l = 0.001) and neighboring towns in Charlevoix, where an astrobleme—a geological formation resulting from the erosion of an impact crater—created a small pocket of fertile land within otherwise mountainous terrain (fig. S14) (38). Limited carrying capacity within the astrobleme led to demographic pressure and subsequent rapid expansion up the Saguenay River [(36), p. 91], resulting in a vast majority of kinship predating the colonization of SLSJ.

The Beauce region (l = 0.002; Fig. 4, C, F, and I) has a handful of founder events in Saint-Joseph-De-Beauce (l = 0.0003) and along the Chaudière River, with migrations reminiscent of a hub-and-spoke model.

Finally, the Bas-Saint-Laurent (l =0.002; Fig. 4, B, E, and H) has an assortment of founder events including a dominant founder event in Rivière-Ouelle (l = 0.0004) but also more minor founder events scattered across hundreds of kilometers of shoreline, acting as a one-dimensional regional hub for subsequent inland migrations (Fig. 4E).

Quoted from p. 4, Fig. 4. Historical migrations and founder events define population structure


Quoted from Supplementary Materials p. 32, Fig. S13: Selecting watersheds with individuals driving principal components


Quoted from p. 5, Results: Historical migrations in space and time

As expected in an expanding population, some early settlers (those identified as super-founders) made a large contribution to the present-day population (39, 40). The top 10 super-founders in each region contributed 37, 12, and 14% of the realized kinship in SLSJ, Bas-Saint-Laurent, and Beauce, respectively (figs. S15 to S17).

Quoted from Supplementary Materials p. 36, Fig. S17: Dispersal range of the top five contributors to each region


Quoted from p. 5, Results: Historical migrations in space and time

Not all regions of Quebec exhibit such spatially defined founder events. Even though Abitibi-Témiscamingue was settled by a process of rapid frontier expansion similar to that seen in SLSJ, it did not lead to founder events reflected in leading axes of genetic differentiation.

In contrast to the events in SLSJ, French Canadian settlers to Abitibi-Témiscamingue came from numerous villages scattered throughout the province (fig. S20). While many of the villages in Abitibi-Témiscamingue have measurable founder events, these seldom overlap at a regional level.

To illustrate this, we consider the villages of Rémigny and Rollet separated by 20 km along the Ottawa River (fig. S20). Cross-coalescence rates of these villages have 11% overlap. For comparison, La Baie and Roberval in SLSJ have 70% overlap despite being >100 km apart.

Thus, Abitibi-Témiscamingue shows limited evidence of a shared founder event and almost no isolation by distance (fig. S10), yet the ancestral structure, together with parallel founding events, creates important substructure (fig. S20).

Quoted from Supplementary Materials p. 39, Fig. S20: Realized kinship for Remigny and Rollet


Quoted from p. 5, Discussion

The BALSAC pedigree—a particularly complete population-scale spatial pedigree—has been instrumental in identifying multigenerational demographic effects such as the reproductive advantages of being on a wave-front expansion (39) or the transmissibility of family size (44) and migration propensity (45).

[...]

In this study, we sought to develop a comprehensive genetic model that captured all these effects and more. We used the concept of realized kinship (or coalescence) to break down founder events as a sum of expected contributions from historical individuals and towns. We used this model to highlight how one of the best-studied human founder populations in SLSJ was influenced by the distinctive geography of the region that was shaped by a cosmic event occurring 400 million years ago in Charlevoix.

By providing a first detailed description of other founder events in Quebec, we showed a wide diversity of founder dynamics. We also found concordant genetic and genealogical support for the idea that geographic features such as rivers and mountains played a systematic role in defining major axes of migration and genetic variation.

Finally, the strong correlation between empirical and simulated genetic data provides evidence that the structure within the French Canadian population can largely be attributed to events in North America, while the population kept a genetic signature of the regions in France that contributed more early French settlers.

Quoted from Supplementary Materials p. 42, Fig. S23: Visualization of the PCA and UMAP analyses of the complete genotype dataset


Related
Laurence Gagnon et al., Deciphering the genetic structure of the Quebec founder population using genealogies (2023)

See also
Eunjung Han et al., Clustering of 770,000 genomes reveals post-colonial population structure of North America (2017)
Linda Ongaro et al., The Genomic Impact of European Colonization of the Americas (2019)
Chengzhen L. Dai et al., Population Histories of the United States Revealed through Fine-Scale Migration and Haplotype Analysis (2020)
Melissa L Spear et al., Recent shifts in the genomic ancestry of Mexican Americans may alter the genetic architecture of biomedical traits (2020)

:-)