Using DNA for Clues with Colonial Ancestors: Quality vs Quantity

Have you noticed that some of your family lines have many DNA matches and others very little? My dad is seven-eighths 19th-century German, one-sixteenth Pennsylvania German (18th century) and one-sixteenth colonial American/British. I’ve noticed that my dad has many matches on his Pennsylvania German side. However, there are only a few matches on my dad’s 19th-century German ancestor lines. As much as we want our match list to tell us about our ancestors, we have to remember that our match list also reflects who takes DNA tests. Some populations and some families are well represented and some are not. If we have no matches in a well represented population, that tells us something. If we have no matches in a population that tends not to take DNA tests, it doesn’t tell us as much.

This post arose out of simply noticing that everyone’s match list is different. For example, Nicole’s father-in-law has 6,000 Ancestry close matches (20+ cM) and 60,000 distant matches (less than 20 cM). My dad has 343 and 9290 respectively. Clearly Nicole’s father-in-law has ancestors who have many descendants and many of them have taken AncestryDNA tests! Some Ancestry testers born in another country may have far fewer than my dad who has at least one colonial American ancestor. Simply stated, you have to have matches to use DNA for brick walls. Some ancestors will have lots of matches while others might not have any. 

It pays to think about your tester’s match list. Do their ancestors come from a group who tends to take DNA tests? Does the ancestor’s family have endogamy? Which of your test taker’s ancestors do you expect will have large clusters, and which may have small or no clusters, and why? That will affect how to think about the number of matches you will be able to connect to an ancestor. No matches for a colonial American ancestor (especially of European descent – the most likely group to take a DNA test) in ThruLines may mean you have the wrong family much more than no matches for a recent immigrant line.

What are the advantages of using DNA with colonial American ancestors (ancestors who arrived in the period before or around the Revolutionary War)? Though today’s DNA test-takers’ colonial ancestors are often outside of the 5-6 generations that autosomal DNA can confidently be useful, having many matches on an ancestral line may be able to push this out to 7-8 generations. Since colonial ancestral lines have had several generations in America, and Americans, so far especially those of European descent, take more DNA tests, colonial American ancestors often have many descendants who have taken a DNA test. 

Here are some takeaways for using DNA to help learn more about colonial ancestors:

1. It’s useful to organize your matches into clusters that connect to your great or 2nd-great-grandparent lines.
2. Some ancestral lines have more matches than others.
3. Colonial ancestors have had several generations to leave American descendants.
4. Americans, so far especially of European descent, are more likely to take DNA tests [1].
5. Colonial ancestors often have many descendants who take DNA tests. 
6. Unknown colonial ancestors may often be 4th-5th great-grandparents. 4th-5th great-grandparents are 7-8 generations away, and so outside of the 5-6 generations where autosomal DNA is more consistently effective.
7. 4th and 5th great-grandparent lines with many matches and good trees, may be able to yield DNA hints that narrow down the possible candidates of an unknown ancestor’s family.
8. Establishing the extended family of your colonial ancestors will help you identify their FAN club. The byproduct of using DNA to identify your colonial ancestor is that your matches (descendants of your ancestor’s siblings and their in-laws) lead you to some of who their FAN club was. 
9. If your ancestral line also has a Y-DNA, mtDNA or X-DNA component, be sure to identify DNA testers that would have inherited that type of DNA. Answering a genealogical question may require using every type of clue available [2].

 

Why do most ancestral lines run up against a 5-6 generation limit with autosomal DNA? What situations make it possible to use DNA a little further than that? After we review these questions, we will dive a little deeper into thinking about the idea of quality vs. quantity that I introduced last time.

Why is 5-6 generations the limit of autosomal DNA usefulness?

This is an important topic but one that I won’t fully discuss here. Instead I will give you the main points with other blog posts that you can go to if you want that topic fleshed out more.

• The more distant the ancestor, the smaller the pieces of DNA you inherited from them, and the greater the chance that you inherited no DNA from them. Ancestral DNA Percentages – How Much of Them is in You? | DNAeXplained – Genetic Genealogy plus Q&A: Everyone Has Two Family Trees – A Genealogical Tree and a Genetic Tree.

• Once you get past close relationships (i.e. parental to 1C), the more overlapping the shared DNA amounts will be (e.g. 30 cM can be anything from second cousin once removed, 2C1R, to eighth cousins, 8C). The Limits of Predicting Relationships Using DNA.
• The smaller the amount two test-takers share, the more likely it is that the match is IBS (identical by state, or a segment common in a population) than IBD (identical by descent, or a segment you and a match received from a recent common ancestor). In other words, sharing 20 cM can mean a fairly close relationship, like 3C, or a more distant 5C, or it can even be a population segment that comes from a common ancestor over 500 years ago. Remember endogamy makes longer IBS segments even more likely. Help! My Segments Are So Sticky!

So 5-6 generations is a general limit of autosomal DNA, but what situations make it more possible to push into the 7-8 generation range?

Do large numbers of low cM matches overcome uncertainty?

Since we can expect to have exponentially more 5th-6th cousins than close cousins, it’s possible to find more of these distant matches. The 5C and 6C level matches that a tester will likely have descending from a colonial ancestor can provide hints about which of the possible ancestors our unknown may belong to. This will only be possible if many of a tester’s 5C or 6C have tested their DNA. In my experience, a colonial ancestor’s descendants are often, but not always, in this category.

How many 5C/6C do we have?

The above table comes from the ISOGG wiki page, “Cousin Statistics (Cousin statistics – ISOGG Wiki).” This table shows several important ideas. 

1. We have small numbers of close cousins we share relatively large amounts of DNA with.
2. Conversely we have large numbers of distant cousins we share small amounts of DNA with.
3. Importantly the table also displays the concept that though we have many many distant cousins we can determine by documentary evidence, there will be a smaller percentage of those true cousins we will share DNA with.
4. Nevertheless since there are exponentially more cousins at the 5-6C level (even though we will only share DNA with a fraction of those actual 5-6C and not all will take DNA tests), we still may have many matches for a particular line. Especially an American colonial line with many descendants.
5. The chart doesn’t depict this, but we share on average 25 cM with a 5C and 18 cM with a 6C. The even smaller amounts of DNA we share with 7th (14 cM) and 8th (11 cM) cousins plus the inaccuracy of trees at 6th or 7th great-grandparent levels means that pushing the idea that quantity can make up for quality of matches will break down at some point.

 

Quality vs. Quantity

The greater the amount of DNA shared between two matches, the more certain we can be that a close genetic relationship (i.e. in the last 5-6 generations) exists between them. But the closer the cousin, the fewer possible cousins there will be. The fewer the cousins that are likely to exist, the luckier we have to be to find them. 

This illustration depicts this concept. Taking numbers from the ISOGG Wiki “Cousin Statistics” table above, this illustration has the total expected cousins on one ancestral line on the left, and the number of those expected to share DNA with a tester on the right. For example, of 23,000 cousins 6C we could theoretically have to our 5th-great-grandparents, we will only likely share DNA with 4%, or about 943 of 23,000. If a large number of people with colonial ancestors take DNA tests, we have a good chance of matching enough of these possible 943 people to get a good clue as to who to research first among the many possible candidates who could be our unknown ancestor.

The following diagrams try to illustrate this concept of inverse relationship of quality vs. quantity as our cousins become more distant. The drawing below shows the “heavier” side for a closer 3C is the large (average) amount of DNA shared, showing that quality is the advantage here. Quality means we can be very certain of a close genetic relationship, but we also have considerably fewer 3C matches than distant matches, so we have to be luckier to find them in our match list.

Conversely the more distant a cousin is to the tester, the lower the amount of DNA shared, and therefore the lower the confidence we can have in a recent genetic relationship, as seen in the drawing below. However the more distant the cousin, the greater the number of cousins likely to exist, and may take a DNA test. If we have an ancestor with thousands of descendants, we may match enough to give us a clue we can use to identify our unknown ancestor.

The heavier side here is the number of possible cousins we can expect to have on any one line, showing that quantity is the advantage with distant cousins. If our colonial era ancestor comes from a group that tends to take DNA tests, we may be able to use quantity to help us overcome the lack of quality of any individual distant cousin. How many distant cousins do we need for a clue? I would say that the more distant cousins I find pointing to a certain candidate family, the more I want to research it. 

For example, I wanted to find parents and siblings for my dad’s 4th-great-grandmother, Christine Hittel. Therefore, I was looking for descendants of my dad’s 5th-great-grandparents. My success depended upon starting with a strong foundation in closer matches, as follows:

1. I tested my oldest relatives on my line of interest, connected their family tree to their DNA test, and separated their matches into clusters that connect to their 2nd-great-grandparents when possible.
2. Even though these 5th-great-grandparents are beyond 5-6 generations (i.e. a 5th-great-grandparent couple on average only covers 2 parts out of the 128 of my dad’s DNA), the descendants of these ancestors belong to a group that are many and more likely to take a DNA test.
3. I have 3C and 4C matches on this ancestral line, increasing the accuracy of sorting matches into clusters.
4. Anecdotally, I notice that Pennsylvania Germans post better and more complete trees than any other group I have seen. 
5. 1-4 here meant that ThruLines could identify a group of 17 separate matches that connect to a possible male sibling of Christine Hittel. I identified several more than that by building out unlinked or incomplete trees of shared matches.

 

Autosomal DNA is only useful for giving us clues to who our unknown ancestors might be if we have matches! My dad’s situation illustrates this nicely. While DNA is mostly less useful with his mid-19th century German lines (especially going past 5-6 generations), his Pennsylvania German lines are really in the opposite situation. 

In fact quality vs. quantity in a population that takes DNA tests means that you may be able to use DNA out past the 5-6 generation cut off. Just remember why the 5-6 generation limit is there, and that DNA is indirect evidence, not proof, that needs to be correlated with documentary evidence.

 

Sources

[1] Heather Murphy, “Most White Americans’ DNA Can Be Identified Through Genealogy Databases,” The New York Times, 11 Oct 2018, and Giorgio Sirugo, et. al., “The Missing Diversity in Human Genetic Studies, “ Cell, Commentary Vol 177, Issue 1, P26-31, March 21, 2019 (https://www.cell.com/fulltext/S0092-8674(19)30231-4: accessed 2 Mar 2022).

[2] “Head-spinning DNA,” Ancestor Callings: Georgia and Mississippi Roots (https://ancestorcallings.blogspot.com/2016/05/head-spinning-dna.html : accessed 22 Mar 2022) > Inheritance Chart of Four Type of DNA.