Let’s face it. A lot of people take DNA tests for the ethnicity admixture estimate. They may not be as interested in solving a family history mystery as they are in knowing the breakdown of their ethnic percentages. How much Irish, Native American, African or Chinese DNA do they carry? So prevalent is this perspective that when I tell people that my job is to identify people’s biological ancestors using DNA tests, they often look at me in wonder and ask “how can you identify a bio-parent with an ethnicity test?” I then proceed to explain genetic cousin match lists, the sometimes ignored but most useful part of DNA test results for the purposes of genealogical research.
To be fair, even if genetic cousin match lists are more helpful for solving family history mysteries, ethnicity estimates can and do provide useful context. Also, consideration of ethnicity results in conjunction with DNA match lists and segment data can lead to interesting and exciting discoveries of their own. When I took a DNA test at 23andMe I was surprised to find I had about 1% Spanish and Portuguese admixture. According to 23andMe’s chromosome paintings, this admixture comes from a single segment of DNA on chromosome 16.[1] Even at higher confidence levels, the Spanish and Portuguese assignment persists. Not only this, but AncestryDNA’s ethnicity admixture estimates also corroborate a small percentage of “Basque” DNA.[2] In my family history research, I have never come across a Spanish or Portuguese ancestor, so I was interested to learn where this DNA may have come from. To do this, I pursued the following process:
- Identify the exact location of my Spanish segment at 23andMe
- Identify genetic cousins sharing DNA in the same region as my Spanish segment.
- Determine which of those genetic cousins are sharing on the same chromosome copy as my Spanish segment.
- Repeat steps 2 and 3 at other DNA testing companies.
- Identify shared matches at AncestryDNA who also have Basque admixture.
- Extend the family trees of all prioritized individuals to see how they are related.
Through this process, I was able to determine that the DNA that 23andMe assigns as “Spanish and Portuguese” and the DNA that AncestryDNA assigns as “Basque” came from my third great- grandfather, Joseph Coumerilh (1850-1929).
23andMe provides an option to download the start and stop positions of your ethnicity admixture. To do this, scroll to the bottom of your ethnicity estimate, click on “view scientific details,” scroll to the bottom of the resulting page, select the confidence level you wish to use and click on “Download Raw Data.” The resulting spreadsheet will give the start and stop positions of each assigned admixture segment. When I pursued this download at the 70% confidence level, I was informed that my Spanish and Portuguese admixture was on a segment of DNA on chromosome 16 between basepairs 105,320 and 13,739,415.
Next, I downloaded the aggregate data from my 23andMe match list to get an idea of which genetic cousins were sharing DNA in the same region. To do this, hover over “Family and Friends,” select “View all DNA Relatives,” scroll to the bottom of the resulting page and click on “Download Aggregate Data.” The spreadsheet output of this download gives you the start and stop positions for the segments shared with genetic cousins with whom you are sharing genomes. Next, I looked for individuals sharing DNA with me in the same region as my Spanish/Portuguese segment.
Remember that each person gets 22 autosomal DNA chromosomes from their dad and 22 autosomal DNA chromosomes from their mom. Therefore, with any given segment you share with a genetic cousin (assuming that it is a true segment due to recent common ancestry and not a false segment) there are two possibilities: you are sharing DNA on your paternally inherited chromosome, or you are sharing DNA on your maternally inherited chromosome. When I reviewed the list of shared segments from 23andMe, I found twelve genetic cousins who shared DNA with me in the same region. Analysis of the matches in common for each of these individuals revealed two groups. The first group of ten matches all shared DNA with each other and I was able to identify several of them as known relatives of my maternal grandfather. However, none of them had Southern European admixture in their ethnicity profiles. The other two matches shared DNA with each other and also had Spanish/Portuguese admixture in their ethnicity profiles suggesting they too carried the same segment of interest.
When I built out the tree of these two paternal genetic cousins, I found that they were my second cousins through my great grandparents: Ira Stoddard and Opal Gay Coumerilh.
In a previous post, I described how I identified all the segments of DNA that I inherited from my paternal grandmother, Fern Stoddard. Indeed, my Southern European segment on chromosome sixteen fits squarely in a region that I inherited from Fern. My maternal DNA in the same region came from my maternal grandfather, Gordon Richard Peterson (1926-2013). Though I did not have the chance to test Fern at 23andMe prior to her death, she did perform testing at Ancestry, Family Tree DNA and MyHeritage. Her ethnicity estimates at those companies corroborate the hypothesis that my Spanish/Portuguese/Basque DNA came from her family. At AncestryDNA, Fern has 4% Basque Admixture and 1% Spanish admixture.[3] When I look at my other close tested relatives, my mother has no Basque or Spanish admixture.[4] Neither does my paternal grandfather’s sister.[5] Lack of Iberian admixture for these individuals supports the hypothesis that my Spanish/Basque admixture came from Fern.
I repeated the approach I took at 23andMe with data downloads from Family Tree DNA and MyHeritage DNA, the other two major testing companies that offer information on shared segments. To download segment data at Family Tree DNA, login to you account, select “Chromosome Browser” from the Autosomal DNA dashboard and click on the “Download All Segments” link in the upper right corner. To download segment data at MyHeritage, login to your account, hover over the “DNA” tab, and select “DNA Matches.” On the resulting screen, there will be a three-dot menu at the top right of your match list. Click on the menu and select “Export DNA shared segment info for all DNA matches,” This will produce a pop-up notification letting you know that the file is being compiled and will be sent to your associated email. Check your email for a download of the data. If you don’t see it soon, check your spam, advertisements and trash folders.
I had tested or transferred my grandmother’s DNA to both of these companies, so it was easier to isolate pertinent genetic cousins sharing in the region of interest. I performed downloads of segment data for both me and my grandmother. Just as I have two chromosome copies to consider on chromosome 16 in the region of interest, my grandmother also has two chromosome copies in the same region. However, in the region of interest, I only inherited one of her copies. Therefore, individuals sharing DNA with me and my grandmother in the region of interest will be related through her Spanish/Basque ancestry.
At MyHeritage, I identified a nephew of Fern who shared the same Iberian segment. I also searched for additional matches at GEDMatch but found none who shared in this same region. At Family Tree DNA, I identified a father and daughter who were distant relatives sharing just one 11 cM segment. I was able to successfully build out the family tree of the father and found that he had ancestry from the towns of Arette, Lourdios-Ichère, Sarrance, Osse-en-Aspe, Bedous, Aramits, and Lanne-en-Barretous in the Vallée d’Aspe in the Pyrénées-Atlantiques Department of France.
AncestryDNA does not provide information on the exact shared segments between tested individuals, but by reviewing the match profiles of individual genetic cousins, I was able to identify some individuals who also had significant Basque admixture. Each of these individuals were descendants or collateral relatives of Joseph Coumerilh.
Fern’s mother was Opal Gay Coumerilh. In a previous post, we explored her childhood and the difficulties of her relationship with Fern. Opal was the daughter of John Henry Coumerilh and Lois Hunt. In turn, John Henry Coumerilh was the son of Joseph Coumerilh who was born in 1850 in Issor, Pyrénées-Atlantiques, France to Jean-Pierre Coumerilh and Anne Ichante. They were natives of Lescun and Aydius respectively in the Vallée d’Aspe. On the opinion of a local priest, the “Ichante surname has a Basque ring to it.” Though we have traced the ancestors of Joseph Coumerilh back to the early 1700s on all lines and to the late 1500s on some lines, to date we have found no ancestors born in the boundaries of the Basque country or modern Spain. Nevertheless, it makes sense that individuals from these communities would be genetically affiliated with these populations. Lescun borders Spain and has a long (and sometime violent) history with the Navarese shepherds across the border. The Vallée d’Aspe is the first valley east of the French Basque country, and according to locals, there is a long history of trade and migration between the two regions.
Issor, Lescun, Aydius and the other towns where Joseph Coumerilh’s ancestors lived are part of the traditional province of Béarn. While Béarn is today a part of modern-day France, it was not always so. In the early 9th century, the viscountcy of Oloron, which included the Vallé d’Aspe was formed under the Dukes of Gascony. In the mid-11th century, the Viscountcy of Oloron was joined to the Viscountcy of Béarn. Over successive centuries, the area came under the control of the kings of Aragon (modern-day Northern Spain) as well as under the kings of England, but throughout this time period had considerable autonomy. Beginning in the 14th Century, under the reign of Gaston de Febus, Béarn gained nearly complete sovereignty over its current borders and subsequently took in large portions of Navarre which includes much of the French and Spanish Basque Countries. This sovereignty lasted until Henri III of Navarre inherited the crown of France and became Henri IV of France. Upon taking the French crown, Henri IV of France entrusted the regency of Béarn to his sister, Catherine de Bourbon. However, after the assassination and death of Henri IV, the reign of Louis XIII saw the incorporation of Béarn and Navarre under French rule. Even so, in the annexation, Béarn was promised continued rights and privileges. Therefore, from annexation in 1620, Béarn still maintained its own parliament, and still considered France a foreign entity until the French Revolution in 1789. Given the long history of jurisdictional changes between several governments on both sides of the current national borders it is not altogether unsurprising that individuals from the Vallée d’Aspe should be genetically similar to the Basques and Spanish surrounding them.
The current boundaries of modern nation states do not align with historic jurisdictions, populations, and ethnic groups. Nor at the time that they existed were those boundaries impermeable to outside influence linguistically, culturally, or genetically. Part of the challenge of modern genetic genealogy is to consider current boundaries and modern understandings of geography and align those perceptions with the genetics of historic population groups. This can sometimes be like pushing a square peg into a round hole. Even though the ancestors of Joseph Coumerilh lived in an area which now is administered under France, they and their descendants had little “French” DNA. Afterall, what is “French” DNA? Is it the DNA of those residing in Brittany which was historically settled by the Britons? What about Normandy which was heavily influenced by gene flow from Viking raiders? What of the regions of Alsace and Lorraine which formerly belonged to kingdoms of the Holy Roman Empire, and which switched between German and French governance several times before their current administration under France? Nice, Marseille, Béarn, Calais, Mulhouse, Corsica, Savoy and countless other cities and regions of France have not always been ruled by France. Likewise, many areas outside of current French jurisdictions were at one time administered by France. What of the genetics of residents of these areas? Is “French” DNA best represented by those with origins in the vicinity of Paris? If so, then as was the case with Joseph Coumerilh and his family, individuals from many of the areas within the current boundaries of France carry very little “French” DNA.
Though I still have not found any “Spanish” or “Basque” ancestors, I have found the origins of the chunks of DNA that various DNA testing companies suggest are of Basque or Spanish origin. They come from my Béarnaise ancestors who lived for centuries on the borders of France, Spain, Béarn, Navarre, Aragon, Gascony and who by nature of their geographic proximity to both Spain and the Basque country are genetically similar to individuals residing today in those areas.
Have you found surprising discoveries in your ethnicity results? Have you been able to trace the origins of particular portions of your ethnic admixture? Feel free to share in the comments below.
[1] “Ancestry Composition,” Paul Woodbury, ancestry composition chromosome painting, 50% confidence level, https://23andme.com, private database, accessed May 2020.
[2] “DNA Story for Paul Woodbury,” Ethnicity Estimate, updated, https://ancestry.com, private database, accessed May 2020.
[3] “DNA Story for Fern Woodbury,” Ethnicity Estimate, updated, https://ancestry.com, private database, accessed May 2020.
[4] “DNA Story for [private],” Ethnicity Estimate, updated, https://ancestry.com, private database, accessed May 2020.
[5] “DNA Story for [private],” Ethnicity Estimate, updated, https://ancestry.com, private database, accessed May 2020.
Genealogenes 