Tuesday, December 9, 2008

Looking into the study: Distance from Africa, not climate, explains within-population phenotypic diversity in humans

Lia Betti1, Fran├žois Balloux2, William Amos1, Tsunehiko Hanihara3, Andrea Manica1

December 02, 2008


The relative importance of ancient demography and climate in determining worldwide patterns of human within-population phenotypic diversity is still open to debate. Several morphometric traits have been argued to be under selection by climatic factors, but it is unclear whether climate affects the global decline in morphological diversity with increasing geographical distance from sub-Saharan Africa. Using a large database of male and female skull measurements, we apply an explicit framework to quantify the relative role of climate and distance from Africa. We show that distance from sub-Saharan Africa is the sole determinant of human within-population phenotypic diversity, while climate plays no role. By selecting the most informative set of traits, it was possible to explain over half of the worldwide variation in phenotypic diversity. These results mirror those previously obtained for genetic markers and show that ‘bones and molecules’ are in perfect agreement for humans.

Of course, climate, environment, living conditions, random mutation and genetic drift, and globalization [inter-ethnic miscegenation as a consequence of immigration] chime in in varying forms and in complex ways in influencing cranio-morphometric variation, as I've noted here before, but what these folks seem to be observing, at least from the little mentioned in the abstract above, is more diversity in Africa, especially sub-Saharan Africa, vs. that elsewhere. It goes back to that old age basic lesson: non-African groups derived from a subset of Africans, and hence, loss of diversity or a fraction of diversity. Though variations would occur in OOA, as a result of a number of bottleneck events and elements of the aforementioned factors, the overall diversity within the population is very likely to be impacted by that of the "founding" population, notwithstanding subsequent expansion events. The pre-existing variation in the original OOA subgroups was already a fraction of that in the African homeland, and there is reason to suspect that a series of bottlenecks events, that marked the dispersal of OOA migrants, would have led to further losses in diversity along the way.

A brief summation with regards to the findings of this study:
  • The results of the "within-population" variation strongly correlated with the augmentation of distance away from sub-Saharan Africa, not with climatic differences across the globe.
  • The authors in an earlier study of the same subject were able to study some 19 measurements whose hereditary values were estimated, as part of the overall measurements in accessing the character of within-population variation across the globe; this study builds on that earlier one. Those measurements reinforced the observation of the pattern mentioned above.
  • The results mirror that of their DNA results.
  • Diversity at the time of the founding population has an impact on the development of within-population variation of an inbreeding population, notwithstanding any expansion event(s) subsequent to a bottleneck event. As noted above, the ancestors of OOA derived populations already sported only a faction of African diversity, since OOA migrants constituted only a subset of African populations. As OOA migrants dispersed further outside of African territory, further losses in within-population diversity of the off-shoots would have likely been a strong possibility, which would explain the gradient character of loss of diversity as one moved away from sub-Saharan Africa.
  • A larger database of comtemporary populations is likely to give a bigger picture of the effect of bottlenecks that generally characterize migration events than Upper Paleolithic specimens of anatomically modern humans, of which the current database only has so much [that is, more limited], and even there, more diversity has thus far been seen in the African collection than say, Upper Paleolithic Europe for example. Even if one were to take the effects of natural selection into consideration and its duration thereof, the idea of more diversity in Africa would only go to show that populations there had accumulated more time to accrue such diversity. It has to be remembered that humans [anatomically modern] had been around at least ca. 200,000 years ago, and throughout much of this time until only about ca. 80-50,000 years ago or so, when OOA migrants – from which much of contemporary non-Africans derive – made it sucessfully out of Africa, spent much of its [humanity] bio-evolutionary and social history exclusively in Africa.
  • In addition to the reason provided immediately above, in order to ensure all crania were physically fit for analysis, and alleviate bias from "deterioration", a decision was taken not to include those from over 2000 years old.

    Cranial series of *aboriginal* populations in the major continents were primarily used, judging from the table used for the 2007 study in question; otherwise, relatively conservative inbreeding segments of populations whose ultimate *origins* are clear, were used for analysis.
  • In a related study from 2007, which is obviously *NOT* the one that's the actual subject of this posting, the authors used a statistical approach to assess the effect of the main climatic variables on intra-population craniometric variability, as well as identifying traits that produced the highest hereditary values. Why?

    This allowed for 1) the consideration of both correction for climatic effect on intra-population cranio-metric variability over time and 2) consideration of within-population variation when the said identified climate-influenced component is removed, while 3) assessing the extent of and considering the outgrowth of within-population variation inherited from "ancient" bottlenecks.
  • And taken from said related study [published in 2007], the authors say:

    Realistic geographic distances between locations were computed as the shortest route through landmass that avoid areas with a mean altitude over 2,000m and assuming the following land bridges; a single connection between Africa and Eurasia via a route through the Sinai to the Levant, the Bering Strait between Eurasia and the Americas, and connections between the Malaysian Peninsula to Melanesia and Oceania.

    So why was the above necessary? Simple: for consideration in determining the most parsimonious centroid at the most likely location of geographical origins on the basis of greatest within-population diversity. The probability of a subset or just a fraction of a population having reduced within-population diversity from that of the superset/ancestral population is by normal standards generally quite statistically significant, and hence, serves as a major indicator for identifying the source set.

    In light of the above, it may well be necessary to reiterate, that: Additional bottlenecks have the tendency to decrease within-population diversity yet even further, as the already-reduced "within-variation" bottlenecked population stretches out during dispersals to different locales. This generally manifests itself in a pattern like the one described in the study cited above, with a linear regression-style decrease of within-population diversity as one adds distance away from the centroid/most likely location of origin.
  • Distance from Africa was considered a major variable, because it was only when the centriod at the most *likely location of origin* within Africa was studied in analytical regression models for within-population phenotypic variability, that significant regression or statistics results were attained, displaying a correlation between loss of diversity and accumulation of distance from the putative origin. These regression models were repeated placing the centroid at the most likely origin outside of rather than inside Africa, but these didn't generate the results seen when an African centroid was used; they didn't show the linear style regressed decrease in within-population variation with accumulation of distance away from the putative origin, which would have been necessary in order for the results to be statistically significant.

    In the 2008 study, the authors demonstrated that distance from the putative African origin was the *sole* predictor of intra-population diversity indices, as opposed to climate!

The map [above] of the current study is pretty self-explanatory.

Synopsis: While the trend actually begins within Africa itself, such that intra-population phenotypic diversity decreases as one moves away from the centroid at the most likely location of origin, it is less acute within sub-Saharan Africa from the putative origin than the trend is with respect to groups outside of and further away sub-Saharan Africa. Thus, the trend is not as apparent in the greater part of sub-Saharan Africa for the reasons already noted; sub-Saharan Africa was:

1)The area wherein the ultimate source of modern humanity's ancestors lie.

2)The area wherein population structuring had already existed, was most concentrated and thrived for some 120 or plus thousand years *before* ancestors of contemporary non-Africans left...heck, long before the oldest TMRCA of non-Africans even ever exited!

3)The area wherein a *tiny fraction* of then pre-existing overall African population, as a subset [bearing ancestors of contemporary non-Africans] of just *one of* the populations of the several therein, left the continent for an overseas destination.

Hence, we get plots like this...
Genetic data parallels cranio-metric phenotypic trend; it too is more diverse in Africa, particularly sub-Saharan Africa, than elsewhere. The entire gene pool of contemporary non-African males for example, ultimately traces ancestry to a single most common recent ancestor with the M168 mutation, but this is not so for Africa. In other words, non-African gene pool draws from only a fraction of the overall African gene pool, and hence loss in diversity. This trend continues onto other parts of the human genome, namely autosomes...keeping in mind basics like:

*The entire autochthonous mtDNA gene pool of contemporary non-Africans derives from L3 ancestry, whereas this is *not* the case with the overall African gene pool. In other words, L3 – as one of the autochthonous African lineages – naturally *includes* Africans, but it does not include *all* Africans. However, Hg L3 generally includes all non-Africans. Alternatively, African gene pool contains Hg L3, but Hg L3 does not contain African gene pool!

*The entire autochthonous Y-DNA gene pool of contemporary non-Africans derives from M168 ancestry [as the oldest attestable OOA TMRCA], whereas this is *not* the case with the overall African gene pool.

This is merely the review of the abstract, but as more comes to attention, this post will be progressively expanded on. Watch this space!