Flickr

  • www.flickr.com
    Abiola_Lapite's photos More of Abiola_Lapite's photos

« Free Speech and Unintended Consequences | Main | Bobby Fischer is Insane »

March 27, 2005

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451afe869e200d834571c3c69e2

Listed below are links to weblogs that reference The Distribution of Human Genetic Diversity:

» Human diversity, race, sexual selection, etc. from Dienekes' Anthropology Blog
Of course, in human terms, we think of "attractive" as someone who elicits a sexual response, i.e., someone we want to have sex with. But, that is not what is important for sexual selection, i.e., for an evolutionary process. [Read More]

Comments

dsquared

This does strike me as terribly mindless; surely to be a pur sang European today (like me, FWIW) you would have to be a descendant of survivors of the Ice Age and a descendant of survivors of the Black Death. That's two bottlenecks that the European gene pool has been pushed through.

Though since the context of the original post was that you'd be picking a small group of people who would be setting about terraforming a planet, I doubt that genetic adaptation would be all that important; after all, white Americans are less genetically well adapted to the environment of Southern California than American Indians, but they live a lot better there.

Abiola Lapite

Yo, brian idioturo,

It's an honor to be called a "racist" by a moron like you, and if I can have you tied up posting from random IP addresses all day and all night, I won't say that my time's entirely been wasted: after all, it only takes a single click to delete everything you've ever written. Troll away, moron.

Abiola Lapite

"white Americans are less genetically well adapted to the environment of Southern California than American Indians, but they live a lot better there."

Sure, but the settler size for white Americans was much, much larger than what I had in mind, which is a population of no more than 100-200 individuals at most. As I assume that the planet is like Earth with the single exception of the absence of humans - think the Americas and much of Eurasia 100,000 years ago - genetic health will matter, as animal husbandry will require coming in contact with all sorts of nasty viruses over time.

Jim

"white Americans are less genetically well adapted to the environment of Southern California than American Indians, but they live a lot better there."

Southern California as an environment is pretty similar to East Africa so presumably any human would be fairly well suited. If people do live better there now than three hundred years ago, and that is very much open to debate considering all the effects of modern population density, the advantage comes form cultural and political adaptations - continental an intercontinental trade, water piped in from north of Mount Shasta and son, and not form any gentic advantage.

Gentic diversity makes a difference in survival and success of a community when it comes to disease resistance and metabolic adaptations to handle dietary challenges. African populatins have to contend with microbes that have co-evolved with the species, and also all the same livestock diseases as the rest of us, so score there. Metabolic adaptations are going to be in the extreme environments that occur outside of Africa, such as lactose tolerance, but even that shows up in some African populations. But that's all - there is no survival advantage at all to have special adaptations across an entire group when it comes to intelligence or whatever you want to call mental aptitude. It is counterproductive to put all your evolutionary eggs in one basket. Even small bands have the full range of all the learning styles.

I didn't realize it was still controversial for some people that Africa is the genetically most diverse area on the planet. Live and learn. The point about the complexity of the population structure is important here - as dsquared points out, Europe has a sizable population, but it has been homogenzied due to bottlenecks and big migrations.

 razib

"Southern California as an environment is pretty similar to East Africa so presumably any human would be fairly well suited."

define "similar." check out a koppen climate map and you'll note they are different (Cs vs.an A of some sort unless you are talking about the highlands of kenya and ethiopia).

FYI, los angeles as about at the same latitude north of the equator as cape town is south of the equator. and, since both are on the western sides of their continents they both have "mediterranean" climates.

 razib

"This does strike me as terribly mindless; surely to be a pur sang European today (like me, FWIW) you would have to be a descendant of survivors of the Ice Age and a descendant of survivors of the Black Death. That's two bottlenecks that the European gene pool has been pushed through."

i would quibble with this assertion based on several issues.

1) from what i recall it seems plausible that europe was settled by modern humans when it was already rather glacial. in other words, it was not settled during a balmy interglacial when populations expanded, after which there was a die off, rather, the initial settlers were venturing into a tundra-grassland. there was a mild withdrawl during the last glacial maximum, but....

2) there were two major "paleolithic" resettlements of europe from the two refugia, one from iberia and one from what is today the ukraine (these show up cavalli-sforza's component maps and other measures of allelic clines).

3) there was a major supplemental demic diffusion from the middle east with the initiation of the neolithic revolution ~10,000 years ago. this diffusion would itself has suffered "bottleneck" effects as the settlers pushed their way up danubian valleys or sallied forth in colonization events toward the western mediterranean.

as for the black death, it was certainly a major die off, but i don't think it was particularly exceptional in the context of post-agricultural civilization. the plague during the reign of the emperor justinian probably wiped out around 50% of the byzantine empire's population in the 6th century, with a skew toward urban areas.

Dienekes

Finding out which population has the highest genetic diversity is interesting, especially if we look at neutral genes, because it helps us track migration patterns, which are assumed (most parsimoniously) to occur from regions of high genetic diversity to regions of low genetic diversity.

On the other hand, if we are interested in the effects of selection on humans, we have to look away from neutral genes, and look at the genetic diversity of each individual gene. In some of those, e.g., the MC1R region which is responsible for human color, Eurasians are more diverse than Africans, while in others the opposite holds true.

Moreover, in some anthropometric attributes it appears that Africans are less diverse than Europeans while in others the opposite is true. So, we have to look at particular traits and trait groups to see which population is more diverse.

Additionally, not all traits are equal in terms of evolution. For example, the shape of the ear lobe does not seem to have a terribly important adaptive value, whereas genes encoding for pulmonary or cognitive function obviously do.

Therefore, we will have to wait to see the distribution of the various human traits and their genetic underpinnings in populations around the world, and determine how diverse human populations are for them, and especially for the most important traits.

Jim

"similar."

Southern California generally includes the Mojave Desert, including Death Valley. Latitude doesn't tell the whole story. California is renowned or notorious for micro-climates. Mainly Southern California is/was savannah-land once you get in from the ocean.

Where are those maps? They sound intersting.

Razib,
Aside from resettlement from the refugia, what do you know about settlement from North Africa? The articles I read seem to conflict, or it may have to do with male versus female lines, but there seems to be a disconnect bewteen Western Europeans and Basques, presumably the resettlers frorm the Iberian refugium.

This has to do with historical linguistics - there are aspects fof Celtic syntax that diverge form Indo-Euroipean and towards Afroasiatic, as it happens. Then there is an even more unusual coincidence, initial consonant mutations in both the Atlantic branch of Niger-Congo and in Celtic. Normally that kind of thing means nothing, but since this particular feature occurs in only these two groups worldwide, in something like close proximity, it begs for an explanation.

 razib

jim,

here is one:
http://www.yale.edu/ceo/Projects/CAP/Koppen-2.jpg

"Aside from resettlement from the refugia, what do you know about settlement from North Africa? The articles I read seem to conflict, or it may have to do with male versus female lines, but there seems to be a disconnect bewteen Western Europeans and Basques, presumably the resettlers frorm the Iberian refugium."

1) there seems to have been minimal genetic exchange over the strait of gibralter. the extent that it occurs it is likely to be the result of the moorish settlement. north africa was mostly settled from the east (noreast africa) via demic diffusion during the neolithic according to the latest data i've seen.

2) the dichotomy between the basques and other european is overplayed. there are some "sharp" differences on some loci, like the rh- factor, but even here it isn't that sharp (there tends to be a drop off of the frequency of the rh- frequency as a function of the distance from northern spain). there is some Y lineage data which indicates that the basques & welsh are phylogenetically closer to each other than either are to the english. similar data comes from sampling the peoples of iberia and the celtic fringe. don't let modern language group distributions fool you, recall that the southern half of iberia was non-indoeuropean speaking until the latinization during the roman period (the northern half was indo-european by virtue of the presence of celtiberian, itself a recent intrusion it seems, so iberia was all non-indoeuropean until very recently, and the same was likely true for much of the celtic fringe).

i don't know enough about linguistics to comment much further on the rest of it. but, remember that cultural diffusion does not always imply genetic exchange. i for one suspect that there was a slight genetic imprint, but far larger cultural imprint, by the prehistoric "megalithic builders" who seem to be found along the shorelines from malta to ireland. i suspect that they were a seafairing peoples who united much of the western mediterranean and and north africa + the atlantnic fringe via a trade & cultural network. commonalities between north africa and western europe might be due to the remnants of this cultural entity (though this is highly speculative, an it is likely unfalsifiable or provable). remember that one of the men buried at stonehenge was born in switzerland (according to concentrations of isotopes in his bones that are indicators of the soil via the foods he consumed)-so prehistoric long distance travel and cultural exchange seems to have been real.

Abiola Lapite

"On the other hand, if we are interested in the effects of selection on humans, we have to look away from neutral genes, and look at the genetic diversity of each individual gene."

1 - How do you know when a gene is "neutral" or not? Just because a certain piece of DNA isn't contained in an open reading frame doesn't mean that variation in it has no effect on phenotypes, and in fact I seem to recall that lactose tolerance is indeed dictated by such a mechanism. A far more likely scenario is that very little of the so-called "neutral" variation is actually anything of the sort, and we just don't have the fine-grained knowledge to pick up on their subtle effects.

2 - Even leaving aside the question of neutrality, the fact remains that on a haplotype basis, Eurasians are essentially a subsample of Africans, and even if individual genes go against this tendency, to expect them to be the common case is to bank on a statistical impossibility; most functional variation at the genetic level *will* be greater for Africans. To quote:

["Sub-Saharan African populations have high levels of haplotype diversity within and between populations, relative to non-Africans, and have highly divergent patterns of LD. Non-African populations have both a subset of the haplotype diversity present in Africa and a distinct pattern of LD."]

"Therefore, we will have to wait to see the distribution of the various human traits and their genetic underpinnings in populations around the world, and determine how diverse human populations are for them, and especially for the most important traits."

The results will be the same, as "the most important traits" will NOT be skin color, nose shape, lip size or any of that kind of superficial variation which has been given attention by racialists far out of any real proportion to its significance. In fact, I'll say right now that the single "most important trait" in human beings is almost certainly HLA variation - what could be more important than the immune system? - and here Africans *definitely* come out ahead in terms of genetic diversity. See e.g.

http://www.pnas.org/cgi/content/abstract/89/6/2277

Dienekes

>> How do you know when a gene is "neutral" or not?

There are statistical tests for neutrality.

>> the fact remains that on a haplotype basis, Eurasians are essentially a subsample of Africans,

Yes, but there is also a 2% (Annual Review of Genomics and Human Genetics, September 2004, Vol. 5: Pages 119-150) which is found in Europeans and not in Africans, and the same is true for other individual races compared to Africans. Let's wait and see what that 2% does.

>> The results will be the same

Perhaps; we will eventually find out. Certainly disparities in disease seem to indicate that the races do not only differ in cosmetic "external" characters, but also in internal health related genes like this:

http://dienekes.blogspot.com/2004/09/positive-selection-on-mmp3-in-europe.html

In any case, the external characters are not unimportant, as they influence the sexual attractiveness of individuals, which is believed by many to be an important force in evolution.

Abiola Lapite

"There are statistical tests for neutrality."

I know what they are and what their limitations are: do you? Tests like Tajima's D or Hudson-Kreitman-Aguade are statistical measures based on modelling assumptions, not laws of nature like F=ma, so you can't assume that an allele must be neutral just because they say so. Even on the assumption that the models underlying these tests are essentially correct, any statistical test will be limited in power by sample size, so these tests would easily fail to pick up on departures from neutrality if the effects are subtle enough, especially given the typical sample sizes in association studies.

"Yes, but there is also a 2% (Annual Review of Genomics and Human Genetics, September 2004, Vol. 5: Pages 119-150) which is found in Europeans and not in Africans, and the same is true for other individual races compared to Africans. Let's wait and see what that 2% does."

A nice way to keep the argument unresolved until no one remembers it exists: another way of stating what you say here is that 98% of all Eurasian genetic variation is already found in Africans, which is completely in accord with what I've been saying from the start. What is more, and more importantly, if we turn the question around to ask how much of African variation is not present in Eurasia, the answer is much, *much* larger than 2%.

"Certainly disparities in disease seem to indicate that the races do not only differ in cosmetic "external" characters, but also in internal health related genes"

Nobody said they didn't - and in fact my prior statements about HLA clearly indicates that they *do* - but this doesn't mean that Eurasian genetic variation isn't mostly a subsample of that in sub-Saharan Africa, only that gene frequencies differ.

"In any case, the external characters are not unimportant, as they influence the sexual attractiveness of individuals, which is believed by many to be an important force in evolution."

So important that Japanese men can develop a taste for blondes in the space of two generations, and Westerners can go from preferring extreme pallor to lusting after tans in 150 years, while interracial marriage rates verge on 50% for blacks and whites in much of Europe ...

Dienekes

>> A nice way to keep the argument unresolved until no one remembers it exists: another way of stating what you say here is that 98% of all Eurasian genetic variation is already found in Africans

2% is the European-specific fraction, there are also Asian-specific, Amerindian-specific etc. fractions, as well as African-specific fractions, and a fraction which is shared by all major racial groups. Therefore, 98% of Eurasian genetic variation is not found in Africans.

Of course, Africans are most variable in neutral genes (as far as can be determined from statistical tests), and they are more variable in some functional genes, while Eurasians are more variable in others. Quantitatively, Africans may turn out to be more variable overall, and in the end we will learn how variable the various races are for the various important genes which remain undiscovered.

>> So important that Japanese men can develop a taste for blondes in the space of two generations, and Westerners can go from preferring extreme pallor to lusting after tans in 150 years, while interracial marriage rates verge on 50% for blacks and whites in much of Europe ...

The fact that Japanese men have developed a taste for blondes does not mean that appearance is unimportant. After all, they did not develop a taste for aboriginal Australians or black Africans, as far as I can tell. In any case, the finding that 47 out of 51 human societies preferred lighter skin than the local spectrum suggests that skin color is a major force in human evolution (Ann Hum Biol. 2002 Nov-Dec;29(6):589-608.).

Frank McGahon

"Let's wait and see what that 2% does."

Lovely. So you get to declare African genetic variation to be "neutral", but we have to "wait and see" what the 2% does?

Abiola Lapite

"2% is the European-specific fraction, there are also Asian-specific, Amerindian-specific etc. fractions, as well as African-specific fractions, and a fraction which is shared by all major racial groups. Therefore, 98% of Eurasian genetic variation is not found in Africans."

Do some elementary arithmetic; you just might find that under reasonable assumptions, 98% *is* indeed a pretty good approximation of the proportion of total Eurasian variation found in Africa.

"After all, they did not develop a taste for aboriginal Australians or black Africans, as far as I can tell."

Which is not very far. Unlike you, I've actually been to Japan, and without going into details, I was *very* popular with the young women there. The Japanese watch basketball and listen to rap like everyone else, and it definitely has had an impact in terms of black sexual appeal.

"In any case, the finding that 47 out of 51 human societies preferred lighter skin than the local spectrum suggests that skin color is a major force in human evolution"

A more plausible suggestion is that the influence of Western culture has been pervasive throughout the globe; after all, the Brazilians didn't get to where they are because of some innate human preference for lighter skin, nor did African-Americans obtain their substantial European ancestry through large numbers of black men chasing white women.

Dienekes

I want to see what all genetic variants do, including the European-specific ones, the Asian-specific ones, the African-specific ones. After all, these private variants, together with the different frequencies in shared variants, are responsible for the differences between the major races.

Dienekes

>> Unlike you, I've actually been to Japan, and without going into details, I was *very* popular with the young women there.

That proves it then.

>> A more plausible suggestion is that the influence of Western culture has been pervasive throughout the globe;

This hypothesis is explicitly tested and rejected by the Japanese researcher who wrote the paper cited above.

delmore macnamara

"Which is not very far. Unlike you, I've actually been to Japan, and without going into details, I was *very* popular with the young women there."

Even this does not _necessarily_ contradict the assertion that skin colour is important in human sexual selection. It might be that, on average, women prefer men with darker skin, & men prefer women with lighter skin. Yes, this argument (or ad hoc hypothesis-saving device, I'm not sure which) is a favourite of Mr Sailer's & the GNXP crowd, but that does not _automatically_ render it false.

"nor did African-Americans obtain their substantial European ancestry through large numbers of black men chasing white women."

Indeed, but opportunistic rape & quasi-rape of those (*wrongly*) regarded as inferiors is surely not at all the same thing as the selection of a wife from among one's "equals". I have seen it asserted (admittedly by explicit racialists) that, now coercion is much less common, black male white female relationships are much more frequent than white male black female. Do you consider this false, or merely irrelevant?

João da Costa

"The fact that Japanese men have developed a taste for blondes does not mean that appearance is unimportant. After all, they did not develop a taste for aboriginal Australians or black Africans, as far as I can tell."

I don't want to derail your main line of thoughut but to ask you to be aware of what happens in other parts of the globe outside the developed countries.
In Brazil you'll find sizeable parts of the population from European, African, and Asian descent plus the few (pure) Amerindians.
Rules for mixing and expression of preferences are not exactly so clear cut. In fact, they love their "mulattas" (some Brazilian tourist operator pretend even that their "mulattas" are an "export product").
The amount of mixing in Brazil shows that, yes, "White (and Asian) men do love black or mulatta women" (at least in some contexts). The relaxed Brazilian atmosphere allows the expression of some feelings which could otherwise get unnoticed.

Abiola Lapite

"That proves it then."

No, we ought to reject empirical evidence gathered by personal observation for your own totally uninformed opinionating. それが破家の主旨です。

"This hypothesis is explicitly tested and rejected by the Japanese researcher who wrote the paper cited above."

Which therefore proves that it *must* be true, as obviously no one's ever written a faulty paper before. At any rate, on the assumption that it is true, what were all those white Southerners doing bedding black slaves when they had white women at their disposal? Why did all those Afrikaners do the same in South Africa? Please do tell.

Abiola Lapite

"Indeed, but opportunistic rape & quasi-rape of those (*wrongly*) regarded as inferiors is surely not at all the same thing as the selection of a wife from among one's "equals""

Natural selection doesn't care about marital status, and the question at issue is what people find sexually attractive, and how innate their preferences are.

Delmore Macnamara

"Natural selection doesn't care about marital status, and the question at issue is what people find sexually attractive, and how innate their preferences are."

Quite. But I suspect natural selection _might_ "care" about paternal investment in offspring, which I would guess (based on data pulled out ouf my arse) tends to be higher where Daddy & Mummy are married or similar. On the other hand this kind of cultural effect might be washed out by genetic ones. I don't know, & I suspect you don't either.

Dienekes

>> Rules for mixing and expression of preferences are not exactly so clear cut. In fact, they love their "mulattas" (some Brazilian tourist operator pretend even that their "mulattas" are an "export product").

Of course, there are different preferences in different parts of the world. That doesn't however invalidate the idea that skin color is an object of selection. In a country where dark skin is valued selection may go one way, and in a country where light skin is valued selection will go another, also accounting for the effects of natural selection which depends on climate. In fact, many scientists believe that racial (and gender) differences in skin color may be the result of sexual selection.

Abiola Lapite

"I suspect natural selection _might_ "care" about paternal investment in offspring, which I would guess (based on data pulled out ouf my arse) tends to be higher where Daddy & Mummy are married or similar. On the other hand this kind of cultural effect might be washed out by genetic ones. I don't know, & I suspect you don't either."

It obviously doesn't matter enough to have kept the quantity of European ancestry from rising in the non-white populations of the USA, Latin America and South Africa.

Dienekes

>> No, we ought to reject empirical evidence gathered by personal observation for your own totally uninformed opinionating. それが破家の主旨です。

The Japanese preference for white skin in women is well-known and see the cited paper for details. Therefore sexual selection for white skin, at least for one of the two genders is a fact, and not an "opinion".
As for the other gender, your experience, may be due to the fact that you are handsome, or it may be that Japanese women dislike black men on average, but you attracted the minority that likes them, or it may be that Japanese women are attracted to black men. Therefore, your experience does not tell us what the preferences of Japanese women are.

>> At any rate, on the assumption that it is true, what were all those white Southerners doing bedding black slaves when they had white women at their disposal? Why did all those Afrikaners do the same in South Africa? Please do tell.

In those relationships there is the confounding factor of social dominance. In the modern USA where there is neither slavery nor apartheid, we see that whites intermarry with Asians at a much higher rate than with Africans, even though there are more Africans than Asians. Hence, it appears that race plays a significant role in the choice of partner.

http://www.census.gov/population/www/cen2000/phc-t19.html

The comments to this entry are closed.

Notes for Readers