Gene Flow

Someone was wrong on the Internet, saying that there had always  gene flow between human groups, which supposedly inhibits people from becoming different.

If you assume that everything is neutral, it doesn’t take a lot of gene flow to keep populations genetically similar  – something like 1 migrant per generation, regardless of population size.  There’s a relationship between average gene flow and Fst:

{\displaystyle Fst=1/(4N_{e}m+1)}

Fst between CEU ( Mormons) and YRI ( Yoruba, in Nigeria) is 0.165. ( Same as that between timber wolves and coyotes). Going from Fst to number of migrants per generation  :

{\displaystyle Nm=1(1/Fst-1)/4}, where Nm is the number of migrants.

So, the average number of migrants per generation between sub-Saharan Africa and Europe, over the past hundred thousand years or so, was 1.26.  Not a hell of a lot of gene flow.  I blame the Sahara, distance, human cussedness,  and the Ice Age.

Was this enough to keep natural selection from pushing the two populations in different directions?  Not at all. I worked this out in my Ashkenazi paper: for  gene flow to substantially block adaptive divergence of a polygenic trait,  gene flow has to average on the order of  2% or more per generation.  Not two individuals: 2 percent of the population.

Since the actual number for all of sub-Saharan Africa was something like 1 migrant per generation, we don’t have to worry about European admixture damping out local adaptation in Africa, or vice versa.

Now there may well have been significant gene flow between one tribe and a neighboring one in some parts of SSA, which kept them from  becoming too different from each other: but there was nothing keeping them from becoming different from people in the Middle East, or Europe, or East Asia.  In fact, we know some populations in SSA were sufficiently genetically isolated from each other to develop lots of Fst  (Bushmen vs Bantu),  and end up quite different from each other.  Obviously different.

Generally speaking, people seem not to have mixed much in Ice Age times. Fst between the Levant and the Zagros mountains, the west and east ends of the Fertile Crescent, was like that between Germany and China today.

As for the idea that local drift can swamp adaptive evolution — see this.

Lande (1980) argues that, “because of the high mutability of polygenic characters, even small isolated populations of several tens or hundreds…can generate sufficient variation for rapid evolution into a new adaptive zone.”

I knew this, not just by being the sort of nut that reads the literature, but because I knew of many actual examples of rapid adaptive evolution in small populations.

A tip to biologists everywhere: if it happens a lot, it’s probably not theoretically impossible.

{parenthetically, biologists have been known  to say that adaptation ( any adaptation, not some specific adaptation)  is unlikely, because the null in neutral theory is that there isn’t any.  Think about this. }

Eskimos show special adaptations to a high-fat diet, and a body form adapted to cold , consistent with Bergmann’s rule and Allen’s rule. I don’t think there were ever many Eskimos. Tibetans developed their altitude adaptations some time ago, before agriculture: how many hunter-gatherers could there have been in Tibet?

Key deer: need I say more?  An insular populations of possums on an island off Virginia ages more slowly ( lives three years instead of two)  – a consequence of  low predation  risk [ evolutionary theory of senescence].

Drop anybody into a real tropical jungle, wait a few tens of thousands of years (maybe less), and you get Pygmies or Negritos.

Darwin’s Finches on the Galapagos: specialized beaks for different food sources. Specializations that change over a few years of drought [ ENSO cycles].

Every example of island dwarfism or gigantism: tiny Maltese elephant, giant Majorcan dormouse.

Why would someone, not just Dan Graur, say that you need a big population in order for adaptive evolution to occur?  it can help ( generates more favorable mutations) but standing variation is sufficient to allow a lot of selection.

This entry was posted in Uncategorized. Bookmark the permalink.

43 Responses to Gene Flow

  1. magusjanus says:

    I mean, it’s pretty obvious there’s been adaptation in skin color, hair color, avg height, weight, femur length, eyebrow ridge, epicanthic folds, sweat glands, hip width/angle, avg age of menarche, avg age of gestation in many cases, countless diff disease susceptibilities, myopia frequency, blood group frequency, and so on and on and on.

    Forget IQ and cranial size for now…. if there was “gene flow between human groups, which supposedly inhibits people from becoming different”, then why are there ANY differences whatsoever? Why can (well, most of us) tell a Japanese person from an aboriginal from a Dutchman?

    At the end of the day, it doesn’t matter. We’re bringing ‘facts to a feel fight’. There are none so blind as those who will not see. We need (ideally several) M1A2s on Harvard Square, long overdue. That’ll get them opening their eyes real quick.

  2. Coagulopath says:

    Why would someone, not just Dan Graur, say that you need a big population in order for adaptive evolution to occur? it can help ( generates more favorable mutations) but standing variation is sufficient to allow a lot of selection.

    Dan Graur should consider allowing 1.26 migrant thoughts to enter his head per generation. He’s becoming a bottleneck.

  3. Daud Deden says:

    I agree except for this: “Drop anybody into a real tropical jungle, wait a few tens of thousands of years (maybe less), and you get Pygmies or Negritos.” which is based upon a false paradigm, that generalist omnivores (Humans including Pygmies) shrink at about the same rate as dedicated carnivores & vegetarians. But Western science places the average European height as “normal” rather than as “unusually tall” (for a primate, an anthropoid, an ape, a hominin, a human). Pygmies have a stable size (for a non-agricultural diet) that has very gradually increased since “Lucy” & “Selam” etc. Similarly, the early Floresians at Mata Menge were smaller than the later “Hobbits” at Liang Bua.

  4. catte says:

    I’m a bit confused by the logic you’re using in the first part. First you assume everything is neutral and use the equation with known FST values to arrive at the 1.26 migrants per generation figure. Then later on you use this 1.26 figure in a context where neutrality is not assumed:

    Since the actual number for all of sub-Saharan Africa was something like 1 migrant per generation, we don’t have to worry about European admixture damping out local adaptation in Africa, or vice versa.

    Or am I reading it wrong?

    • Eponymous says:

      Point is you need a lot more migrant flow to prevent adaptive differences (cf Greg’s Ashkenazi simulations). FST puts an upper bound on gene flow (which we can measure), but it’s well below what’s needed to prevent adaptive differences.

      So hypothetically you could get groups that differ in meaningful ways due to differential adaptive pressures, but don’t have much FST difference. But not the reverse.

      • gcochran9 says:

        Think of it this way: it takes enormously high levels of gene flow to inhibit local adaptation, levels that would keep Fst near zero. Fst between geographically distant populations is not near zero, and in fact we see many local adaptations.

        • Joe says:

          I’m sorry, I’m dense – in the second paragraph you say that it does not take a lot of gene flow to keep populations similar – one per generation. Then later you work out that there has been 1.26 per generation between Europe and SSA. So that exceeds the one per requirement. What am I missing? Thanks-

  5. dlr says:

    I don’t think you can use Fst to back calculate the number of migrants per generation, over the preceding x thousand years. That would only be valid if all the differences between the two groups were equally adaptive in both locations– if all the differences were neutral. This is obviously untrue, picking one at random, Sickle Cell Anemia. If historically Mormon genotypes were better adapted to Europe, and/or if Subsaharan African genotypes were better adapted to Subsaharan Africa, then you are going to have selection weeding out Mormon genotypes from Subsaharan Africa — and/or vice versa– in every generation. This will make your calculated number of migrants per generation, Nm, smaller, probably much smaller, than the the actual number of migrants that arrived in each generation.

  6. Reziac says:

    Here’s a calculator. I don’t know how accurate it is.
    “Red Lynx is a population genetics simulator that allows students to explore the effects and interactions of several evolutionary forces.”

  7. RCB says:

    So is Graur just not good at polygenic adaptation, or lying?

  8. gcochran9 says:

    You know your theory. And you used to live in the belly of the beast. The high fraction of professionals in genetics that don’t know the theory of their own discipline and publicly say nonsense has always seemed odd to me.Your thoughts on this?

    • RCB says:

      Most of my colleagues were evolutionary anthropologists who didn’t have much background in quantitative evolutionary theory (population/evolutionary/quantitative genetics). Most of them couldn’t tell you about the basics of pop gen: mutation-selection balance, drift, migration, etc. Let alone the infinitesimal model (limits!). Math in general is a weakness for this group (that was my competitive advantage for the time I was in the field). Most of them are field working anthropologists – not theorists (which is fine).

      So most of them just parrot whatever the outspoken PC geneticists or evolutionists of the day are saying – be that Gould, Lewontin, Feldman, Eric Turkheimer, Jonathan Marks. Maybe Dan Graur will break into that list. Sometimes it’s the blind leading the blind (Marks, for example, doesn’t seem very smart. Doubts about Turkheimer…), but sometimes the guy at the top is smart enough that he’s must be lying. Feldman is smart, for example.

      As for professional geneticists, I don’t know them as well. I’ve hung out with Graham Coop and his group a bit. Smart folk. There’s clearly a SJW lean there. Surely that helps them professionally. But what do they really think?

      I’m sure there are a lot of smart folks who are closet crimethinkers. I’ve never personally gotten an admission, though. I bet Razib has: I recall confiding in him when I started having doubts. Perhaps others have.

      • Jacob says:

        I’m in a Facebook group moderated by Marks. I’m convinced that he’s both stupid and dishonest. Nothing he says is interesting for any reason other than creepy levels of vitriol for anyone leaking the wrong information to the general public. He whinges about evolutionary psychology even when it isn’t topically relevant.

        As for confessions, make of this what you will. A professor of mine relayed to me that, in private, Richard Dawkins despises these hacks and is (unsurprisingly) entirely confident in the importance of the gene.

      • magusjanus says:

        I don’t think the “if he has smarts he must be lying” paradigm holds true. Look at Gates or Neil Turok:

        Sometimes they’re very smart, and also capable of believing batsht crazy things, usually aligned with the dominant religion of the time. Newton spends his time on alchemy and bible, Turok and Gates on finding hidden African geniuses.

        • Jim says:

          In Newton’s time it wasn’t clear that alchemy was impossible. A lot of very remarkable transformations can be done chemically. Why not lead into gold? Because they are both elements but that wasn’t obvious in Newton’s day. Substances in nature don’t come with labels indicating whether they are elements or not. Water and air were long thought to be elements but the first is a compound and the second a mixture.

        • Crew says:

          Turok and Gates on finding hidden African geniuses.

          Well, they have been found now, so Turok and Gates will have to find something else weird to do!

      • Dylan says:

        Not a frequent occurrence to hear Cochran praising someone for their grasp of theory. Care to suggest some resources to an interested layman?

        Dr. C, any recommendations from you are of course welcome as well.

        • RCB says:

          I don’t recall a good, readable, intro-level textbook that I can recommend to the lay person. Some of the usual ones mentioned are Hartl & Clark (Principles of Pop Gen) and Gillespie (Pop Gen Concise Guide). You could look for a used copy. Both are kind of old, but the basic theory doesn’t expire.

          Probably better is this free online textbook:
          It’s more content than you’ll want or need. But Joe Felsenstein is the real deal, I believe.

    • RCB says:

      One year I TA’d for an intro evolutionary anthropology course. Basic stuff. One week, instead of learning anything, we played a video that featured folks like Marks and Lewontin talking about race and genes. I think it was PBS’s Race: The Power of an Illusion. You can imagine what it’s like. When we TA’s were told that we were going to do this, no one (including myself) protested, questioned, ec. Most (also including me, at the time) were basically on board with the message, I think. But if anyone wasn’t, they wouldn’t have brought it up then. It’s just not a topic that people like to talk about – and when it comes up, generally people try to change the subject. Some are true believers, sure, but I suspect many aren’t.

    • gabriel alberton says:

      You mean like Craig Venter is apparently doing on Twitter? Perhaps he has some incentives for doing so?

      Anyway, as we all know, although there are ADHD-linked genes (which Venter himself supposedly has) there are no intelligence-linked genes.

      Or there are IQ-linked genes, but IQ is discredited anyway, is not related to intelligence at all, or something. It’s just some test.

      Or okay, there are intelligence-linked genes, but they are equally distributed among all populations. Not randomly, but equally. Somehow Papuans are more intelligent than all of us, though. Whatever.

      You aren’t a racist, are you?

      • Jim says:

        The idea that there is a universal constant that is the average genotypic intelligence of the 6500 or so different ethnicities on this planet is extraordinarily improbable.

  9. mikesmith says:

    “very smart, and also capable of believing batsht crazy things”

    It strikes me there is a big difference between believing something that is “crazy” (sensational, etc.) and believing something that appears to be demonstrably false. I think believing in infinite or multiple universes is crazy, but no one has shown that it is false, However, if you believe the earth is round like a penny rather than round like a ball, that goes beyond merely holding to a “crazy” idea.

  10. Greying Wanderer says:

    “I blame the Sahara, distance, human cussedness, and the Ice Age”

    if close-cousin marriage was the human default (at least since agriculture) that would act as a barrier also – if that was the case then maybe gene flow mostly only came from elite alliance marriages – enough to spread adaptive genes but < 2%?

  11. Greying Wanderer says:

    “Why would someone, not just Dan Graur, say that you need a big population in order for adaptive evolution to occur?”

    not my area of expertise but generally speaking i’d say 99% of SJW inclined people don’t have a head for numbers and so are easily swayed/bullied by people who appear confident in that area.

    if an open debate on the numbers side of things was allowed the genetic side of the argument would win hands down (this has actually been happening online for the last ten years or so and the non-gene side were completely BTFO – hence the mass censorship online).

    so imo it’s 99% people who don’t get numbers and 1% confident liars protected by a media that won’t allow the liars to be directly challenged on the numbers.

    • Dmitry Anisimov says:

      at least non-SJ could have made numbers available in a wiki like “responses to creationists”?

      • Greying Wanderer says:

        wiki is heavily patrolled by the blue-haired thought police but the numbers side of this issue was being explained in scores of other places until recently (harder to find now since social media started trying to cleanse the internet but it’s still ongoing in a more samizdat way – and the main effect of the censorship has been to convince centrists the dissidents are right).

        • I don’t mean wikipedia org… other wikis, like metapedia of infogalactic. You see it’s quite difficult to make references in form of “it was said somewhere in Greg’s blog… or HBDchick’s blog, i’m not sure”…

          I’m pretty skeptical about your latter claim.

          • Greying Wanderer says:

            it’s a mostly under 25 thing that started c. 10 or so years ago when the schools became overtly anti-white leading numerous spergs who didn’t like being painted with original sin to start researching it for themselves – it’s a separate strand from the academic type sites.

            (nb i’m only talking about da youth here, the vast majority of older people are still fully gas-lit on the issue.)

  12. says:

    “Drop anybody into a real tropical jungle, wait a few tens of thousands of years (maybe less), and you get Pygmies or Negritos.”

    Drop 10 million Nigerians in Stockholm, wait a few tens of thousands of years (maybe less), and you get Swedes. Yea or Nay? Of course, the the evolutionary pressures that developed the Swedes have been mitigated by technology…alas!

  13. Rosenmops says:

    According to the BBC children’s series “The Story of Britain” there have always been black people in Britain. And girls wanted to be warriors.

    • J says:

      “…standing variation is sufficient to allow a lot of selection.” Yes, but life is not standing. Start with the standing variation and add mutation and then multiply by 2000 (generations). Yes Shawn etc., you could breed even phenotypical Swedes, although I don’t know why you mention it since already there are plenty of Swedes.

      • says:

        I see from your avatar you are Jewish? Are there “plenty” of Jews? There are only ~8,500,000 ethnic Swedes in Sweden. When the diaspora is added the number is still small. Over 190,000,000 Nigerians in Nigeria, with population explosion to come. 1.386 billion Chinese in China alone.

  14. Antonio Large says:

    Mr. Cochran, this is all very cut and dry, & O.K. for a “special” condition to gene flow (we may make the analogy of special vs. general relativity) where behavior is overly simple & confined in dimension. There is no room in your mathematics for Black Swan events. Your pre-calculus treatment lacks advanced mathematics for discontinuities anyone can observe in real life. Please, think harder.

    • Michel Rouzic says:

      I can’t tell if this comment is meant as a parody of the kind of generic nonsense about fat tails and black swans a Talebite would say or the real thing. Assuming the latter, when we talk about average gene flow between two given populations over the last 100,000+ years that covers everything that actually happened, both the typical and the extreme.

      I discovered this blog through Taleb’s tweeting at the same time as he did and it’s sad that I’ve clearly learnt a lot more from it than he has.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s