The great pruning, and the great synthesis

By Razib Khan | May 9, 2012 12:11 am

Sahul 10,000 years ago

John Hawks has a very long rumination on the story of blonde Melanesians which came out last week. If I can read between the lines I think some of the implications dovetail with John’s thesis in his 2007 paper on adaptive acceleration. But I’ll leave the deep reading of tea leaves to those better versed in such affairs. Rather, I will comment on two issues. The first is specific. I believe that the TYRP1 R93C allele responsible for blonde hair among the Solomon Islanders is going to be found to be the same one responsible for blonde hair around New Guinea and among some Australian Aboriginals.

There are several reasons why I suspect this. First, these Oceanian populations do seem to be a distinctive clade. There are some disagreements among geneticists as to whether they are the descendants of the first settlers of Oceania in totality, or whether they’re a compound lineage. The natural historical details need to be teased apart, but there’s no doubt that they’re a distinct and separate phylogenetic lineage in relation to other human populations (with some admixture from Austronesians in the case of Melanesians). Second, the phenotype of rather light hair, and dark skin, is striking and parallel among all these populations. It is not entirely impossible that random genetic drift could stumble upon an architecture which results in such a suite of characteristics, but I’m generally skeptical of this possibility of having occurred several times in the same human lineage, when it seems relatively rare in other populations (the loci associated with light eyes and light hair in Eurasians seem to have some effect on skin color as well, so it is difficult, though not impossible, to have  very dark skin and light hair). Third, as noted in John’s post, this seems an “old” variant. A new mutation which rose rapidly in frequency should have a lot of associated markers flanking it (ergo, high linkage disequilibrium). Again, we need to take into account the joint information here; there seem several Oceanian populations where the allele is at high, but still minor, frequency (assuming it is the same allele). Oceanian populations, especially those verging toward Melanesia, have low effective population sizes. If this is an old variant I am curious as to why it has not fixed in any of the daughter populations (recall that if it did fix, and admixture recently drove it below ~100 percent, there would be a lot of linkage disequilibrium). Therefore, I think we have to seriously consider the possibility that balancing selection is maintaining this polymorphism in the Oceanian populations. This does not entail that the selection is targeting the hair color phenotype, though it may.

In his post John Hawks seems to suggest that ancient variation in pigmentation, phenotypic and genotypic, may be somewhat different from what was the norm in Eurasia over the past 10,000 years. Rather than focusing on this issue, I will rather put forward a proposition of only moderate boldness, and lay out my own suspicion of what the next few years will tell us about the human past. First, let us set the stage. Between 100,000 and 15,000 years ago modern humans spread out across the world, to every continent where they reside today. There were several “touch & go” moments. The  Toba event may have been one, and the Last Glacial Maximum perhaps another. Human populations waxed & waned. Because of low densities, and resultant barriers to gene flow, many of the deep phylogenies evident in the mtDNA and Y chromosomal record date to this period. It does not seem that modern human expanded rapidly from one or two populations ~15,000 years ago. Mitochondrial Eve tells us that is not so. As I have stated before the “state of the art” in 2005 would have ended with this story. By about ~15,000 years ago, as Amerindians reached the precipice of the New World, pre-Columbian patterns of genetic variation were settling in.

I do not believe that this story holds up. It does not seem to accurately describe how the 20 percent of the world’s population which is South Asian came to be. More precisely, there was always a peculiar discordance in the uniparental lineages for South Asians; the maternal line was closer to East Eurasians, the paternal close to West Eurasians. The autosomal data, which looked at the whole genome, indicated a moderately closer affinity to West Eurasians, though as a function of geography (northwest to southeast) and caste (high to low). These conundrums are solved by a recent admixture of >10,000 years before past. But it is not simply the existence of particular sets of humans. The patterns of genetic variation in Africa today are very new, and owe little to the Last Glacial Maximum. Rather, it seems that the Bantu expansion wiped clean much of Sub-Saharan Africa of pre-existent variation. The Khoisan, Pygmies, and Hadza being the remnants of vast numbers of pre-Bantu populations, which seem to have had marginal effect on the genetics of East African Bantus.

The two cases above illustrate dual dynamics which I believe shaped the nature of modern human genetic variation in the early Holocene: the great pruning and the great synthesis. The great pruning consists of the marginalization and assimilation of vast amounts of cultural diversity, and perhaps genetic diversity as well. It seems likely that the Pygmies and Khoisan have archaic admixture not found in other Africans, while Oceanians have archaic admixture not found in other humans. If it were not for geography, and the practice of horticulture, the populations of  Near Oceania may have suffered the fate of the Negritos of Malaysia and the Philippines. But speaking of Negritos, the Andaman Islanders, and South Asians, illustrate the second aspect: the great synthesis. South Asians are a compound in part descended from the cousins of the Andaman Islanders. In The First Farmers Peter Bellwood lays out an elegant model of demic diffusion from core agriculture hearths, which explains the modern linguistic and genetic patterns we see around us. As an extreme null hypothesis it is useful, but I think we need to be cautious about taking a stylized fact too literally. Bellwood’s thesis came close to positing ancient agricultural apartheid.  But mtDNA remains imply strongly that women of local lineages were absorbed into expanding populations (a result which is eminently plausible given historic patterns).  Racial purity has never been high on the list of human male “to-do’s.” There is a moderate probability that within the next few years other Eurasian populations will be seen to be similar to South Asians in their origin: products of a dialectic between the conqueror and the conquered, between the intruder and the indigene. Finally, in the shadows, and poorly understood from what I can glean, is the third dynamic which comes to the fore in the receding tide of the action of the great pruning and synthesis: the sons of the soil who manage to dodge the fate of being pruned from the human tree by finding a different strategy of survival.

Obviously this is not the whole story. The Columbian Exchange is one chapter, which post-dates the prehistoric prunes and synthesis, though it echoes them in many ways. The expansion of the Han Chinese is another story which post-dates the period to which I allude. But these stories are known and counted.  What I am suggesting here is that in the near future dynamics only faintly recollected will begin to move to out of the shadows, because they have left an imprint in our genes.

MORE ABOUT: First Farmers
  • Karl Zimmerman

    Correct me if I’m wrong, but indications are Australian Aborigines are an ancient admixture themselves, between Melanesians and a second population, right?

    Are there any signs as to who this second population was related to? Or is this one of those cases where even less of the non-Melanesian “zombie” can be reconstructed than ASI?

    Regardless, this sort of complicates matters. Presuming the blondness came from the Melanesian component, admixing with a (Eurasian) population would cause the incidence to decline. There would have to be a secondary selection event, and possibly a population expansion from this new, highly blond core, in order to explain the distribution.

  • Razib Khan

    but indications are Australian Aborigines are an ancient admixture themselves, between Melanesians and a second population, right?

    one group last fall indicates that oceanians absorbed admixture from east eurasians *before* they were oceanians. IOW, in mainland east eurasia. this would probably predate 40,000 years BP. that’s a long enough time that again, i’m wondering why it didn’t fix, or why it didn’t go extinct.

    another group indicated that oceanians exhibited signatures of admixture between an older group, which is wave 1 migrants, and exhibits ancestry which is specific to oceanians (and has the denisovan), and a wave 2 group more similar to andaman islanders and malaysian negritos. this is inclusive of australians & papuans, so the event is likely on the order of < 10,000 years ago.

  • ohwilleke

    “I think we have to seriously consider the possibility that balancing selection is maintaining this polymorphism in the Oceanian populations.”

    One of the arguments for why monarchies survive is that if one person is even slightly notable relative to everyone else that they are the obvious choice to choose for special recognition and to form a symbolic rallying point, even if the reason for that notability, heredity, really confers no actual advantage in ability (or even is disadvantageous in ability).

    Blonde hair probably has no selective advantage per se, and may even be a disadvantage (bad for camo when out hunting). But, maybe having a rare trait that singles someone out and takes that person off the “hunter track” that everyone else is on facilitates some form of rudimentary specialization in an uncontroversial way, and that resulting specialization had adaptive value for the group. If you have too many people who are distinct, this trick doesn’t work. But, an obvious, low frequency trait could do the trick.

  • Sandgroper

    ‘bad for camo when out hunting’ – You can’t be serious. In any case, it’s most common/pronounced in children and, to a much less extent, adult females.

  • ackbark

    My impression is that Oceanians are descended from people who had been spending a lot of time in a thick forest and then found themselves out on the ocean and spending a lot of their time on beaches.

    So, blond hair, reflecting sunlight, would get a lot of attention, and perhaps reflect heat. So that it isn’t necessarily adaptive of anything but attractiveness.

    (does blond hair actually reflect heat to any measurable extent?)

  • T. Kosmatka

    Fascinating post, and comments. If oceanians are the result of two waves of migrants, I’m curious which wave carried the TYRP1 R93C allele. I saw in John Hawks’ post that the Denisovans were excluded as carriers of this variant, so there is the nail in the coffin to that theory. (darn it all) I agree that until evidence proves otherwise, the null hypothesis should be that the pale hair variant of the Solomon Islanders likely has the same genetic cause as pale hair in Australia and New Guinea.

  • Sandgroper

    I think you can also rule out any form of sexual selection, at least in Aboriginal people – there are strong rules about marriage, and people in those arid areas just don’t have that much choice about who to marry, or (not to put too fine a point on it) even that much choice about who to target for rape. If it is balancing selection, I’d say it’s something else being selected for, but it is hard to imagine what.

    The thing to bear in mind is that it occurs in the western desert and inland arid areas – it’s a regional thing, and it’s mostly in children. We’re not encouraged to realise this, but visually there is quite a bit of regional diversity among Aboriginal people, at least superficially.

    There’s also the suggestive evidence of (1) the arrival of the dingo about 3,500 years ago, (2) a major technological change within the past 10,000 years, and (3) two major language groups, one of which is confined to northern parts of the Northern Territory and Western Australia (but a vexed and hotly debated field). Any of those things could suggest a relatively recent second wave, although not necessarily.

    The dating of prehistoric sites suggests that with the first arrival, the coastal and more fertile areas like the Murray-Darling basin were populated quickly, with the more arid areas being populated much more slowly and much less densely. In the event of a second wave, I guess you could assume a similar pattern. Certainly within my lifetime, some Pitjantjatjara were contacted who had never seen white people before, because I remember when it happened – we’re talking *remote* and very inhospitable country, and lots of it. A second wave might not have introgressed that far.

  • Razib Khan

    If oceanians are the result of two waves of migrants, I’m curious which wave carried the TYRP1 R93C allele.

    more investigation needed, but, i had to bet i would say that the older wave (however you skin it). that’s what distinguishes oceanians from other populations. also, philippine negritos don’t exhibit blondism at all, so it may be more recent derived than the origin of the clade before they split off.

    I saw in John Hawks’ post that the Denisovans were excluded as carriers of this variant, so there is the nail in the coffin to that theory.

    john did not say that actually. reread what he wrote, because he entered into the record that it may be problematic to extrapolate from one individual to a whole population. even among solomon islanders if drew a random person you’re likely to miss this allele.

  • T. Kosmatka

    “john did not say that actually. reread what he wrote” “it may be problematic to extrapolate from one individual to a whole population”

    –Good point. I’ll pry that nail back out of the coffin now. I hope we get more examples of Denisovan DNA to test in the future.

  • Jonathan Friedlaender

    I will try to make this short, as it’s getting very involved. My major point is that this region is hyper-variable genetically and phenotypically, from island (and island section) to another. This is almost certainly because of successive founder events and genetic drift. Original founding populations were clearly very small. As noted in many publications (including my own), in many instances you can tell with a fair degree of certitude which island a person is from – Malaitans having a lot of blondism, Bougainville/Western Solomons people being jet-black in skin color (and hair), etc. Using a battery of STRs, we could cluster people very neatly by island and even island section in the Bismarcks and BOugainville (not the Solomons, however). It’s quite a remarkable area for this sort of thing, and it’s inconceivable that any selection is the cause of this structured diversity there.

    As I understand it, the LD questions associated with this variant (hence age and selection estimates) has been difficult to address because of the neighboring highly variable regions. We can hope that more work will sort this out. I would be surprised if it’s much older than the time the Solomons were first settled – roughly 10,000 YBP. But that’s just my conjecture based in its apparently neatly geographically circumscribed distribution. That also would mean no connection to Australian Aborigines, or lighter hair colors in some highland New Guinea populations. Again, just my conjecture.

  • pconroy

    @3 Andrew,
    “Blonde hair probably has no selective advantage per se, and may even be a disadvantage (bad for camo when out hunting).”

    I’ve mentioned many times here over the years that their is a convergence across Northern hunter/carnivore species towards white fur/hair and blue eyes – just look at humans, dogs and polar bears for instance.

    I think it’s exactly because of being great camouflage in snowy landscapes.

    In Oceanians, I can only imagine that it came from a Northern population, like Denisovan, who may have been native to Siberia originally? Alternatively it may have evolved independently as camouflage in desert/savannah conditions, just like Dingos are a yellowish color, as are Lions etc…

  • gcochran

    “it’s inconceivable that any selection is the cause of this structured diversity there.”

    I do not think that word means what you think it means.

  • ohwilleke

    @11 If it provided an advantage, it would have reached fixation in the population long ago. If it was a disadvantage, it would have probably dropped out of the gene pool long ago. The trick then, is to identify a role in which it has an advantage, but only in small amounts. I’m hard pressed to come up with a better story to explain a frequency that is neither 100% nor 0%.

  • Sandgroper

    @13 – Andrew, yes, so am I. Sorry if I sounded unduly derisive. It’s an absolute head-scratcher. But if you saw Aboriginal men hunting in dry grassland, you could never imagine that having sort of ash-blondish hair would be any kind of stand-out. Plus it’s present principally in the kids. (To a mob of kangaroos, it is scent, sound and movement that give the hunter away, far more than some minor issue of hair colour is likely to when you have black bodies moving against an arid landscape.)

    That’s why I’m having a really hard time imagining balancing selection for this kind of trait over a really long time – and I think from the geographical distribution of the trait in Australia it must be ancient, so we’re looking at least around 45,000 years. I can only think it’s pleitropic, but I still can’t imagine what it is that is under selection. Unless it has something to do with avoidance of inbreeding in sparse populations – it’s most highly represented in juvenile females.

    The situation in Australia wrt diversity is very different from the Solomons – so much so that, despite the low probability of this trait evolving multiple times, maybe it is different. It looks different, the blondeness is much less stark.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar