Rare genes:
There is something that has nagged me for years. I should have talked about it before. It has to do with effective population size. I read a book on population genetics and got a sort of a drift but didn’t verify numbers I had read in other contexts and was sort of hoping a review article would come out I could cite. That has now happened. (Along Keinan and Andrew G Clark, Recent explosive human population growth has resulted in an excess of rare genetic variants SCIENCE vol. 336 no. 6082 May 11, 2012 page 740)
The bare bones run thus. Every gene has a certain probability of undergoing a mutation in any given generation. I think it’s now thought to be about six per person per generation. So there is a constant inflow of genes into the population, most of them without effect. For the sake of calculation, only neural mutations are considered. While there is a constant inflow of these variants of genes there is a constant outflow. Consider a mutation that prevails in 1% of the genes in a population. If the population numbers 100, then that 1 mutation has about an even chance of vanishing in the first generation. Or maybe it will increase slightly. But over a few generations it is quite likely to vanish. But if the population is 1,000 then the mutation has 10 copies, at least one of which will most probably survive into the next generation. It still faces a chance of being eliminated each generation but a much smaller one than in the smaller population. If you know the rate of appearance of that mutation and you know how frequent the gene is, then with a little algebra you can calculate a population size that would account for it. Do this for a host of genes and you can be pretty sure your “effective population size” reflects something or other.
The hooker is that this only works at equilibrium. If the population size is changing, and has been changing, then you are not quite sure what your effective population size is. But with such calculations most parts of the world have an effective population size in the tens of thousands. I seem to recall the number 20,000 being bandied about. My reaction was, so far as I could get numbers, was, “That does not compute.”
I generally find infertility becoming catastrophic by the time a population size of 1,000 is reached. I can live with 1,000. I could live with 2,000. But twenty thousand boggles my mind. As I have said before (check out http://nobabies.net/Orlando%20meeting.html) that would suggest a time to speciation of 40,000 generations. Mice in the canary island, introduced at or since the time of Columbus about 500 years ago, have undergone speciation to the degree that chromosomes differ in appearance from one valley to the next. That is a late change, of course. Generally you go through a period of “hybrid breakdown” where the offspring of a crossbreed are fertile but the line dies out after a few generations. But assume that the mice have undergone speciation in 500 years. If it takes 40,000 generations for that to happen it means a generation time of 80 per year. A mouse will take about a month to mature. Even proposing a generation time of 1 month, and mice generally can life longer even in the wild, that’s 12 generations per year. That doesn’t seem possible. The same argument can be made for rabbits in the Azores and for camels breaking into Bactrian and dromedary, which started after they had been domesticated and currently is in the hybrid breakdown stage.
Well now the population geneticists have noticed that the human population is not in fact at equilibrium but has been growing, say from 5 million 10,000 years ago to about 7 billion now.
I think even this growth is an understatement. If you look from your train window or a small airplane at the English countryside you will see a sort of quilt pattern in the vegetation. These are called Celtic fields. They represent little plots that were farmed consistently over so long a time that even now long after that pattern of farming has vanished – in the Middle Ages crops were gown in long strips – the variations in the soil can be seen. The word Celtic, of course, has changed meaning. It used to mean Bronze Age conquerors, but since they can’t find any genetic trace of those folks politics has demanded that the definition be extended to mean anybody in Britain before the Saxons. I must use the old definition to make sense.
The people who made the Celtic fields were so pressed for room for their farms that they blanketed the countryside and even terraced hills. That kind of intense farming could probably feed England’s current population of 50 million or come close. And that was probably going on about 4 or 5 thousand years ago. But according to the Celts’ story, when they arrived they found the place empty, “Except for a few giants.” Outbreeding, I have suggested, may make some really big people as well as really low fertility.
So there was a population collapse in Britain. Although that collapse obviously carried away a lot of mutations, it cannot have had much of an effect on people elsewhere.
So I propose that populations have been growing far faster than even the total suggested numbers indicate but there have been an enormous number of local extinctions.
The article points out that an enlarging population is not at equilibrium and will tend to hold onto rare mutations. This results in a spurious increase in calculated population size.
There is also the question of sample size, which the article says is very important. I read in that book on population genetics, “Yes, the number of people we are considering is very small, but the number of genes is very large so our statistics are in fact excellent,” or words to that effect. Ain’t so. Of course you need a large sample size, the genomes of a lot of people. Otherwise you will miss a lot of rare mutations.
The article does not in fact say, at least I did not notice, just what the “effective population size” really is. No matter. I wouldn’t have believed them. I really don’t think their estimate of population growth can have been realistic in terms of the experience of the populations that have actually survived. So for now I see no conflict between my own suggestion of how big a population can be and be sustained over time and their suggestion of what effective population size is. I just don’t think the information is really in yet.
There have been 67,055 visitors so far.