Discrepancies in the Estimated Date of the Human MRCA

Do all humans share a common ancestor from a few thousand years ago or is it more like a few hundred thousand years ago?

Interested in knowing how two people are related on much shorter time scales? Check out the most accurate relationship predictor.

Most people who are interested in the study of ancient human migrations believe that multiple branches of humans split off from an ancestral population that originated in Africa and, until recently, that those branches remained mostly separate from each other to the present day. Is every part of that statement true?

In his fascinating book “A Brief History of Everyone Who Ever Lived,” Adam Rutherford makes the astounding claim that all humans alive today share a most recent common ancestor (MRCA) from somewhere between 3,400 and 3,600 years ago, i.e. at least one person lived during that time who is the ancestor of every person alive today. Rutherford even notes the frowns of disbelief on audience members’ faces when he includes this claim in his speeches. Indeed, it flies in the face of our intuition.

The first thing that needs to be said is that Rutherford’s claim is not completely ludicrous, although it is unconventional. The figure comes from a mathematical model, and you’d be hard-pressed to find an author who claims that their model proves something in the real world; rather, they would most likely say that such and such is the result we see based on the assumptions of the model. It happens that the model was very conservative, but it doesn’t take into account real-world situations in which an extreme exists even beyond the point of being conservative. I also believe that date of the MRCA is very sensitive to the proportion of the world’s population that is being discussed. That is, covering just a few more percentage of the population might, at times, require pushing the date back by several centuries. What follows here is a comparison of the two claims based on the assumption that Rutherford’s claim could be true.

When you start entertaining the idea that the MRCA could have lived 3,500 years ago, it probably won’t take long for you to realize that it would only take one isolated population on an island somewhere for it to be wrong. Then again, it would only take one immigrant on a boat to that same island 3,500 years ago to have easily become an ancestor of everyone on that island today, or one conquistador who left descendants in remote areas of the Amazon. And they may not have left any genes that survive to the present day, which would make it impossible to prove or disprove the claim. However, it seems more likely that some tribe has remained untouched than that every tribe in the world has had at least one recent immigrant. The difficulty with the claim of 3,500 years is that every seemingly isolated population would have had to have had an immigrant sometime in the past 3,500 years, whereas there only needs to be one exception to that supposition to make it untrue.

We now know that our intuition that major branches of humans have remained distinct is verifiably false multiple times over. Much progress has been made in the past decade with respect to the study of ancient human migration. The main theme of Rutherford’s book is that humans have always been enthusiastic about both migration and reproduction — that no population stays in the same place, unmixed, for very long.

In another wonderful book, “Humans: Who We Are and How We Got Here,” David Reich similarly notes that our family “tree” is actually more like a lattice. He and his team have shown, on a few occasions, that there have been ancestral “ghost populations” that no longer exist, but that contributed greatly in varying percentages to populations that exist today. We’ve been fortunate enough in some cases that the remains of a member of such a population have later been found, with DNA that matches the missing population, thusly validating the claims of the scientists who discovered that there was a ghost population.

Reich’s estimation of when our MRCA lived is wildly different than Rutherford’s. Part of the reason for that is that Reich’s estimation is based on genes, and the contributions of many of our ancestors disappear over the generations. Here’s an excerpt from Reich’s book:

Across chromosomes 1–22, the most recent shared ancestor for all present-day people ranges mostly between 5,000,000 and 1,000,000 years ago, and nowhere is it estimated to be more recent than about 320,000 years ago.

So the estimation is ~3,000 from Rutherford and ~300,000 from Reich. One of these figures, given by two experts in human DNA science, is very wrong. How do we reconcile the difference?

In Rutherford’s book, he says that his figure comes from a model by the mathematician Joseph T. Chang in 2003. The mathematical model, first published in 1999, is truly elegant, but it only applies to closed populations with random mating. Chang lists multiple constraints for his model, including that it can’t be used for humans because of geographical and other barriers. The results of this model show us that MRCAs for most populations occur far more recently than anyone would expect, but even when allowing for migration within the model I believe that it doesn’t apply to all populations of humans.

The figures 3,400 to 3,600 are in fact from a paper by Douglas L. T. Rohde published in 2004 with Chang as a co-author. Part of Rohde’s computer simulation takes into account Chang’s probabilistic model from 1999. Rohde’s model was an excellent idea and was well executed, but it has a lot of constraints of its own. For example, the model divides the world into ten nodes that roughly correspond to continents. Migrants are swapped between continents in each generation. The amount of migration in the model is considered a conservative estimate, but in some cases in the real world there may have been no swapping of migrants. This is noted in the paper, and examples are given of populations that remained isolated for long periods of time, such as Tasmanians, but it’s claimed that there are no longer any populations that are known to remain isolated to the present day.

Reich, who gave the figure of 320,000 years ago, describes a few populations that I believe could be used as evidence against Rutherford’s claim. The people of Little Andaman, an island south of India, have been isolated so long that they don’t have any West Eurasian ancestry, whereas the people of mainland India all have somewhere between 20% and 80% of their genomes from that group. The remaining mainland Indian ancestry is Ancestral South Indian (ASI), which was a mixture of south Asian hunter-gatherers and Iranian farmers, but ASI is the sole contributor to genomes of the Little Andamanese. This population as well as presumably some on other Andamanese islands have been isolated from mainland Asia for about 48,500 years. In fact, North Sentinel Island in the same archipelago is even more isolated (at least presently), so much so that we don’t know anything about the genomes of its inhabitants.

Another population that Reich has studied is that of mainland India. DNA analysis has showed that the caste system there has been so strictly adhered to for the past few thousand years that India is actually a country composed of many populations living side-by-side who have more genetic differences than northern Europeans have to southern Europeans, proving that isolation is possible even where no geographical barriers exist. What’s more, the caste systems in some parts of Africa have produced populations that are even more diverse than in India.

Even people from the Italian island of Sardinia show signs of isolation. Sardinians are probably descendants of Anatolian farmers who entered Europe about 8,500 years ago — just one of the waves of farmers who migrated across Europe in the past several thousand years, sometimes genetically replacing most traces of the Europeans who came before them. Unlike in mainland Europe, where steppe pastoralists contributed much of the present-day genomes starting at least 4,500 years ago, it appears that the same group never reached Sardinia. Of course many Sardinians share more recent ancestors with mainland Europeans, but on the whole they are quite a different, isolated, population.

There’s just one problem with this line of reasoning in the few cases above. This problem will remain unresolved and it could work in favor of Rutherford. We’re comparing Reich’s studies of DNA to a mathematical model. And, just as if we were comparing DNA to a family tree, DNA is always missing any traces of some ancestors far enough back in time; the farther back, the more ancestors that are missing. So, although the Little Andamanese don’t carry DNA from West Eurasian populations, that doesn’t mean that there has never been an Indian immigrant to Little Andaman Island in the last 3,500 years. In fact, from what’s known about mixing populations, the DNA of a single immigrant would have likely been selectively pushed out of the gene pool fairly quickly, just as, save for the some very useful genes, Neanderthal DNA was quickly pushed out of the gene pools of what are thought to be traditional Homo sapiens. What we do know is that there hasn’t been much Indian immigration to Little Andaman island relative to the pre-existing population in the past 48,000 years. And that seems like a fairly convincing argument that populations can remain isolated for very long periods of time.

I don’t doubt that a lot of people on Earth share a common ancestor from 3,600 years ago. I simply believe that there are going to be some exceptions. It may be far more insightful to make individual claims, such as that almost all Europeans share a common ancestor from 600 years ago, or that Europeans share a common ancestor with people from India within the last 5,000 to 7,000 years.

What about David Reich’s claim that our MRCA lived much earlier — about 320,000 years ago? DNA studies have suggested that a “Mitochondrial Eve” lived around 160,000 years ago and a “Y-chromosomal Adam” no more than a few hundred thousand years before that. The names “Mitochondrial Eve” and “Y-chromosomal Adam” have been criticized by many scientists for being misleading. They note that a lot of other of women would have been alive at the same time as Mitochondrial Eve. Many of those women may even be our ancestors via other lines.

As far as when our MRCA lived, the date of Mitochondrial Eve (~160,000 years ago) should be treated as earliest possible estimation. After all, it’s very unlikely that the female MRCA of all people alive today is the same woman as our Mitochondrial Eve. And we know, by the property of conjunct probabilities, that it’s more likely that our MRCA is connected to each of us via some combination of male and female ancestors and lived more recently than the woman who connects us all via an unbroken line of women, which would be an additional constraint. For example, if I told you that a common ancestor of everyone alive today who connects to each of us via an unbroken line of tall men who prefer the color green (which is really just an average color) lived exactly one million years ago, you would know that the MRCA of all humans lived no more than a million years ago.

The claim that the MRCA lived less than 160,000 years ago isn’t a criticism of the 320,000 years ago claim because the data sources are different. It just so happens that mitochondrial DNA data produce an earlier date than autosomal DNA, and therefore must be closer to the actual date. Aside from “earlier than the date of Mitochondrial Eve,” it’s difficult to make a more accurate estimation. Directly studying our DNA may have seemed like a promising method, but so much of it is lost over time. There’s a good chance that there’s no DNA left in us from our MRCA. But we may be able to get a better estimation indirectly through DNA. As geneticists such as David Reich discover more about our ancestors’ migration patterns and further constrain the time periods during which populations split off from each other, we could get an estimated date of the isolated population that split off from others at the farthest point in the past. That would be the date of our MRCA. But, as Joseph Chang asks, why are we interested in knowing how long ago this occurred? That date may only be interesting to that one isolated population. For that reason, it may be much more interesting to find the individual dates that different populations split off from each other — a tailored answer for each group.

I believe that the date of our MRCA lies somewhere in between the estimations given by Rutherford and Reich. We can say with reasonable confidence that we all share an ancestor who lived sometime more recently than 160,000 years ago. As ancient genomes become available and are compared to the ones already discovered, we can incrementally move that date farther forward in time.

I think of the people of North Sentinel Island, who are obviously very opposed to accepting outsiders. Maybe they haven’t accepted one in the past four thousand years. I don’t know if their population has been large enough to sustain that kind of endogamy. But maybe they once came from a nearby island with a larger population that had been isolated for thousands of years. I’ve wondered why people feel the need to contact them. Maybe that curiosity is just human nature. Maybe something inside of us wants everyone to be part of one large population. I think we should leave isolated populations alone, but one might argue that that’s just us playing god — wanting to see a different way of life remain intact, knowing that it will never exist again once we’ve undone it. What’s more important is that they don’t want to be contacted. We should respect that. Plus, our diseases would probably wreak havoc on them. But someday human curiosity will win the day. They will accept migrants, abandon their island, or die off for one of many reasons.

I don’t think we will ever find that the people of 2019 all share a common ancestor from 3,600 years ago, though it’s a beautiful idea. Rutherford remarks several times on the beauty of the closing line of Rohde et al., 2004:

No matter the languages we speak or the colour of our skin, we share ancestors who planted rice on the banks of the Yangtze, who first domesticated horses on the steppes of the Ukraine, who hunted giant sloths in the forests of North and South America, and who laboured to build the Great Pyramid of Khufu.

I think it’s a beautiful idea, too. It probably isn’t quite true yet, but maybe it won’t be long.

Feel free to ask me about modeling & simulation, genetic genealogy, or genealogical research. And make sure to check out these ranges of shared DNA percentages or shared centiMorgans, which are the only published values that match peer-reviewed standard deviations. That model was also used to make a very accurate relationship prediction tool. Or, try a calculator that lets you find the amount of an ancestor’s DNA you have when combining multiple kits.