This past Christmas, my four young adult children gave me a really great gift in the form of DNA test kits and spit samples. Although it violates the “oldest first” rule about the order in which we should test family members, I was nonetheless pretty excited to have the kids test. Since a group of three siblings is necessary in order to attempt visual phasing, their data will give me an opportunity to practice that, and since I’ve already tested all four of their grandparents, I’ll also have the “answer key” available.
Additionally, since I have this unique data set at my disposal—all four grandparents, both parents, and four children—I thought it might be fun to create a table comparing all our ethnicity estimates from AncestryDNA to see what insights such analysis might offer about DNA inheritance and also about the limitations inherent to these estimates. This chart is shown in Figure 1, and a larger version can be viewed here.
Figure 1: Comparison of AncestryDNA ethnicity estimates among four siblings, their parents, and grandparents.
Before we dive into the data, let’s discuss the ethnicities in my family based on pedigree. The ancestors of my father-in-law (“Paternal Grandpa” in the chart) were ethnic Poles from the Russian and Prussian partitions as far back as I’ve been able to discover. (A brief discussion of the partitions of Poland and related border changes is found here.) My mother-in-law’s (“Paternal Grandma’s”) ancestors were also ethnic Poles, from the Prussian partition. My mother’s (“Maternal Grandma’s”) family were ethnic Poles from the Russian and Austrian partitions. My father’s (“Maternal Grandpa’s”) ancestry is more mixed. His mother’s family was entirely German, and his father’s family was half German/Alsatian, half English/Irish/Scottish. Since three of the four tested grandparents have all-Polish pedigrees, let’s look at their data first, before moving onto my Dad’s data.
It may be helpful to start by noting that there’s no such thing as “Polish DNA.” Poles are part of the broader ethno-linguistic group known as Slavs, although Ancestry chooses to use the term “Eastern Europe and Russia” which they define as being “primarily located in Poland, Slovakia, Czech Republic, Austria, Russia, Hungary, Slovenia, Romania, Serbia, Ukraine, Belarus, Moldova, Bosnia and Herzegovina, Croatia.” 1 However, examination of the territories that were originally inhabited by the Slavs, shown in Figure 2, suggests that this area corresponds pretty well to the geographic region referenced in Ancestry’s definition of “Eastern Europe and Russia.”
Figure 2: Spread of Slavic tribes from the 7th to 9th centuries AD in Europe by Jirkha.h23, licensed under CC BY-SA 4.0.2
The ethnicity estimates from my mother (“Maternal Grandma”), my father-in-law, and my mother-in-law are pretty typical for individuals with all-Polish pedigrees. All three of them were reported to be as much as 95-100% Slavic, with significant contributions from ancestors whose ancient origins were in the Baltic tribes. Although part of the region inhabited by the Balts is located in Poland today, they were a separate group from the Slavs, as indicated by the map in Figure 2. Each of these testers also shows traces of other ethnic groups (Finnish, Swedish, Norwegian, Germanic, and European Jewish) that make sense in light of Poland’s history and geography.
Notice that when AncestryDNA presents these ethnicity estimates, they indicate their best guess for the percentage of each ethnic group identified, and also the predicted range for this estimate. For example, about 3% of my husband’s (“Dad’s”) DNA was reported to originate in Germanic Europe, but the range was reported to be anywhere from 0-18%, which means he could have no Germanic DNA, or he could have up to almost 1/5 of his genetic ancestry from this region, but their best guess is about 3%. You’ll notice that some of these ranges are tighter than others, so that same 3%, when it refers to the Norwegian component of my oldest son’s (“Child 1’s”) ancestry, falls in a narrower range from 0-7%. The ranges help us to understand how confident we can be in the contribution of a particular ethnic group to our genetic makeup. Moreover, we should remember that there may be more than one explanation for the amount of ethnicity we inherit from one particular ethnic group. My husband’s 3% Germanic component may indicate that he had a single German great-great-great-grandparent, since on average, we inherit about 3.125% of our DNA from each of our 32, 3x-great-grandparents. Alternatively, that 3% could be the combined contribution from multiple, more distant ancestors, and I believe that the broader range suggests that this possibility is more likely.
In some cases, AncestryDNA predicts one or more additional, specific regions from which a tester’s family was likely to have originated. These regions are based on Ancestry’s identification of Genetic Communities,™ a process which incorporates data from the family trees posted by testers and their DNA matches, in addition to the data gained from analyzing the DNA itself. The science behind the Genetic Communities™ is explained in detail here. In the case of my family, all of us with documented Polish roots (i.e. everyone but my Dad) were accurately assigned to the Genetic Community for those with roots in the Czech Republic, Slovakia, Poland, and Lithuania. Moreover, all of us with these Polish roots were assigned to the “Pomerania” region, while my father-in-law and oldest son were additionally assigned to the “Mazovia and Łódź” region.
I found it interesting that my father-in-law and oldest son were the only ones identified as having ancestry in Mazovia and Łódź, for two reasons. First, my mother (and therefore all my children and I) also have documented ancestry from this region, so our pedigrees would predict that all of these tested family members would be noted to have “Mazovia and Łódź” ancestry except for my mother-in-law, whose ancestors were from Prussian Poland, and my non-Polish Dad. The fact that so many of us were not assigned to this category suggests that Ancestry’s algorithm may depend more on on the presence of a particular marker or markers in one’s DNA, rather than on family tree data. However, if that’s the case, then it’s a bit of a red flag that only my father-in-law and oldest son would be assigned to the “Mazovia and Łódź” category, and not my husband. This is the second reason why I found this result interesting: DNA cannot “skip generations.” If my son carries a particular set of markers for Mazovian ancestry, shared by his paternal grandfather, then my husband must carry those markers as well, despite the fact that the test did not identify them.
One should also pay close attention to Ancestry’s definitions of these geographic areas, as they might differ from the usual historical definitions. For example, Ancestry identifies everyone in my family group except for my dad as having roots in Pomerania. Historically, “Pomerania” refers to a region presently divided between Germany and Poland, along the shore of the Baltic Sea, shown in Figure 3.
Figure 3: Approximate border of historic Pomerania region, shown in yellow, by Kelisi, licensed under CC BY-SA 3.0.3
Although the borders of historical regions like this can sometimes be somewhat subjective, the map suggests that at its southernmost point, the Pomerania region extends to about the 53rd parallel, approximately the same latitude as the city of Bydgoszcz. After the final partition of Poland in 1795 and throughout the 19th century, this area was under Prussian control, and my mother has no known Polish ancestry from the Prussian partition. So although it made sense to me that Ancestry identified my mother-in-law, father-in-law, husband, and children as having Pomeranian roots, I was at first puzzled about how my mother and I managed to achieve this designation. While this may reflect an ancient migration of our ancestors from Pomerania to Mazovia that predates the oldest vital records I’ve been able to find, it’s perhaps more likely that the issue lies in Ancestry’s definition of “Pomerania,” shown in Figure 4.
Figure 4: Pomerania geographic region as defined by AncestryDNA.
At first glance, the highlighted region seems to extend much further south and east than the historic Pomerania region shown in Figure 3. We can confirm this suspicion if we zoom in closer on Ancestry’s map (Figure 5).
Figure 5: Detail of Ancestry’s Pomerania geographic region, showing location of Bydgoszcz underlined in red and Konin underlined in blue.
Ancestry’s version of “Pomerania” encompasses Słupca County, which lies between Poznań and Konin. Słupca County was the ancestral homeland of my mother’s Grzesiak and Krawczyński ancestors. So by that definition, it makes sense that my mother and I would have “Pomeranian” roots despite the fact that we have no known ancestors from the historical Pomerania region going back to the early 1700s.
So what other insights can we gain by looking at the data for the four grandparents? In addition to the Slavic (“Eastern European”) and Baltic components that are typical features of DNA from ethnic Poles and which were discussed previously, my mother-in-law’s (“Paternal Grandma’s”) DNA includes possible small contributions from Germanic Europe, Norway, and Sweden. Given the fact that her ancestors were all Poles from the Prussian partition, these ethnicities make sense. We might expect to see some intermingling of German DNA in a region that was under Prussian control. Moreover, the Scandinavian contributions are perhaps more evident in Poles from the Prussian partition than they would be in Poles from the Russian or Austrian partitions because a tiny bit of Pomerania (“Swedish Pomerania“) was under more or less stable control of Sweden from 1630 until 1815.3 In contrast, the Swedish genetic influence in other parts of Poland, which might have resulted from the Swedish Deluge in the mid-1600s, may have been more transient and diluted over time.
Moving on now to my dad’s data, Dad was reported to be 55% Germanic, within a surprisingly tight range of 55-56%, which was a tad less than I might have expected, based on his pedigree. However, Dad’s other major ancestry component, 38-40%, was from “England, Wales and Northwestern Europe,” a region which includes almost all of modern-day Germany by Ancestry’s broadest definition, the outermost circle shown in Figure 6. The balance of Dad’s ethnicity was from Ireland & Scotland (0-5%), and since the whole “United Kingdom” component of his ethnicity based on pedigree should be about 25%, I’d think Ancestry’s estimate is pretty accurate overall.
Figure 6: England, Wales and Northwestern Europe region as defined by AncestryDNA.
Given that Dad’s DNA reflects such a high percentage of ancestry from the “England, Wales and Northwestern Europe” region, I found it somewhat surprising that I did not inherit any of it, and therefore I could not pass it down to my kids. Those genes are entirely lost in my branch of the family. I seem to have inherited virtually all of his Irish & Scottish genes, however, and I managed to pass down about half of these to three of my children. (My oldest son lost out there.) Similarly, I inherited whatever traces of Jewish ancestry may exist in Mom’s DNA, but I did not pass them down to any of my children.
Finally, it’s interesting—if a bit suspicious—that Ancestry did not detect any trace of Norwegian or Swedish DNA in my husband, passed down from his mother, yet two of our children exhibited traces of Swedish DNA (up to 4%) and three exhibited traces of Norwegian DNA (up to 9%). As mentioned earlier, DNA can’t “skip a generation;” it’s not possible for our children to exhibit traits or carry genetic markers which they did not inherit from one parent or the other. To me, this speaks to the imprecision of the test—these are ethnicity estimates, after all. So perhaps my husband does, in fact, carry whatever markers they’re using to assess “Swedish” and “Norwegian” DNA, but for some reason they were missed when Ancestry’s algorithm calculated his ethnicity estimate. Since all four children were independently found to have some trace of Scandinavian DNA, this seems more likely than the alternative hypothesis, which is that the entire Scandinavian contribution is a false positive, although that is certainly possible.
Ancestry DNA continues to refine their ethnicity estimates as the number of testers in their reference panels grows, and as more geographic regions are included, so perhaps in the future these estimates will become more precise and some of these issues with consistency between individuals in the same family will disappear. However, even despite those issues, these ethnicity estimates match reasonably well with expectations based on documented pedigrees for the tested members of my family, especially when one interprets the estimates in light of both the specified ranges for each ethnic contribution, as well as Ancestry’s unique definition of each of the geographic areas.
1 “Eastern Europe and Russia,” DNA Origins for Julie Szczepankiewicz, AncestryDNA (https://www.ancestry.com/dna/ : 10 March 2019).
2 “Spread of Slavic tribes from the 7th to 9th centuries AD in Europe” by Jirkha.h23, is licensed under CC BY-SA 4.0, accessed 10 March 2019.
3 “Pomerania map” by Kelisi, is licensed under CC BY-SA 3.0, accessed 10 March 2019.
© Julie Roberts Szczepankiewicz 2019
3 thoughts on “Comparison of Ancestry’s Ethnicity Estimates for Three Generations of My Family”
Julie, I enjoyed this article very much because it paints such a broad historical sweep and helps us to understand not just the genetic ancestry but also a bit about political borders which for Poland have been so arbitrary in it’s recent history. Thank you for an enjoyable, informative blog.
LikeLiked by 1 person
Thank you, Chris! I’m glad you like the blog in general, and this article in particular.