Single Common Ancestor doesn't mean the others died!

A recurrent misconception cropping up over and over again. Some news report comes out to the effect “Every human on the face of the earth had an ancestor in common x, xxx, xxx years ago, says scientist”. Gee, say various and sun-dried people, imagine that, just think of all the other people on the face of the earth back then, and only this dude’s progeny made it!

:smack: :smack: :smack:
a) It’s not so amazing*. I have 2 parents, had 4 grandparents, 8 gread-grandparents, 16 ancestors in the previous generation, 32 in the one before that, 64, 128, 256, 512, 1024, 2048, 4096, etc.; some of those folks had many many kids, so if I trace back 10 generations and then come forward, counting all the descendants, it’s a shitload of people to count. Chances are pretty damn good some of you folks are amongst them. I am neither a statistician nor a biologist, but I’d think if you went back far enough and picked someone, that someone would either have no living progeny or else everyone currently alive could claim that person as an ancestor.

b) Me and my sister are both descended from my great-grandma. Even if we were to rid the earth of anyone not also descended from my great-grandma, that would not mean that none of her contemporaries have surviving progeny. Heck, at least 7 of her contemporaries have me as surviving progeny. Her son my great-uncle had kids with someone other than the person her son my grandpa had kids with, and that someone had parents who were also contemporaries of my great-grandma. So although I am not descended from that person, that person does has living progeny even in a world where only descendants of my great-grandma remain alive.

  • it’s not amazing that a given population (whether everyone or everyone with blue eyes or everyone who has Tay-Sachs disease) all have a common ancestor; it’s interesting to know when. If it’s surprisingly recent, statistically speaking, one may surmise that perhaps, just possibly, everyone inherited a particularly relevant bit of DNA from that person.

Why does there have to be just one most recent common ancestor? To take an extreme case, my four children have two MRCAs (my wife and me). If we are going back several millenia, might there not have been dozens of people who were all MRCAs of a present-day group?

Yes, that’s true, but you will eventually reach a point where an individual is either an ancestor of everyone alive today or of no one at all.

That’s called the Identical Ancestor Point. It’s thought to be no further back than 15k years ago.

Another interesting fact:

You only have to go back about 30 generations before the number of ancestors you have exceeds the number of people on earth. Cue banjo music.

Is it possible to even estimate which percentage of people fell into which camp? In other words, of the people alive on Earth 15,000 years ago, what percentage are ancestors of everybody, and what percentage are ancestors of none?

Of course. That ancestor was called Adam. :smiley:

True, but some of these ancestors are not distinct people. You’re neglecting to account for inbreeding.

The Perfect Master covered this over 20 years ago

It might be possible, although I’m not sure. Here’s the paper (note: PDF file) if you want to see. The results were obtained via a computer model, not actual measurements, so I’m thinking it might have come out of the model. I haven’t read the entire paper myself.

I’m currently reading The Ancestor’s Tale by Richard Dawkins. The whole idea of the book is that each chapter goes one step further back in time finding common ancestors between species (which he dubs “concestors”). Concestor 1 is the MRCA of humans and chimpanzees/bonobos. Concestor 2 is the MRCA of Concestor 2 and Gorillas. And so on. (other Concestors include: 7 is Tarsiers, 14 is marsupials, 19 is coelacanths, 31 is sponges, 39 is eubacteria.

At the beginning of the book a section talks about “Concestor 0” - the most recent common ancestor of all current humans.

There is an example given of a mathematical model of a population on an island and assume a simplified mating model (a random chance any two people will mate). With a population size of 5,000 people on a small island, how far back would you have to go to find the MRCA of that population? The answer from the simple model is the logarithm (base 2) of the population size. So for a population of 5,000 (that you assume was a steady population size during the entire period), the MRCA is 12.3 generations ago (less than four centuries if there are four generations per century).
Then, to address Freddy the Pig’s question:

If you keep continuing back in time after the MRCA (most recent common ancestor), it doesn’t take long until you get to a point where everybody is either a common ancestor or has no surviving descendants. Once you get to that point, about 80% of the people in that generation will theoretically be ancestors of everyone alive in that future population. This point is approximately 1.77 times older than the MRCA point. So in a population of 5,000, the MRCA was 12.3 generations ago, and that “all or nothing” point is 22 generations ago.

Also, the numbers don’t change much is there’s a larger population size. For a population of 60 million people, you reach the MRCA point at 23 generations ago, and the second point where everyone is either a common ancestor or has no surviving descendants is 40 generations ago.

Does this square with Australian aborigines reaching that continent 40k years ago?