Monday, April 2, 2018

Africa's Neanderthals

Skull from Zambia, dated to 110,000 years ago. Modern humans co-existed with archaic groups in Africa, particularly in the south and west.

When and where did modern humans emerge? Anatomical evidence points to somewhere in eastern Africa some 300,000 years ago. The time of origin is different if we look at behavioral and genetic evidence. Sophisticated tool-making, detailed artwork, and other signs of “behavioral modernity” appeared only 70,000 years ago (Brown et al., 2012). Genetic evidence points to a series of demographic expansions between 80,000 and 60,000 years ago in eastern Africa, with the last one spreading throughout Africa and into Eurasia (Watson et al. 1997). At that moment, an innovation in thinking seems to have given these truly modern humans an edge over everyone else.

As these humans spread throughout the world, to what extent did they intermix with the more archaic groups they replaced? We can answer this question for Eurasia by comparing the modern human genome with reconstructed genomes of the now-extinct Neanderthals (Europe, Middle East, and Central Asia) and Denisovans (East Asia, Southeast Asia). Present-day Eurasians have relatively low levels of archaic admixture: about 2% from Neanderthals and up to 5% from Denisovans (Sankararaman et al. 2016).

What about Africa? Unfortunately, we have not yet reconstructed the genome of any archaic population from that continent. We probably never will, given that DNA tends to degrade quickly in tropical climates. In theory, there should be more admixture in Africa than in Eurasia, since many archaic Africans would have been "near-modern," i.e., much more similar in appearance, behavior, and genetic makeup to modern humans than either Neanderthals or Denisovans. Greater genetic similarity would have also made hybrid infertility less likely. Indeed, it looks like male fertility suffered from hybridization with Neanderthals or Denisovans, given that present-day humans have a lower proportion of archaic ancestry on the X chromosome and in genes disproportionately expressed in the testes (Sankararaman et al. 2016). In these parts of the genome, natural selection has stepped in to remove archaic admixture.

The above speculations seem borne out by a recent and still unpublished paper. Its authors, Sriram Sankararaman and Arun Durvasula, came up with a novel way to measure admixture from an unknown archaic group, essentially by using a machine learning algorithm (which they validated with data on Neanderthal introgression in present-day Europeans). When they applied this method to Yoruba from Nigeria, they found a level of archaic admixture higher than in any other human population known to date:

Our results suggest that Yoruban individuals trace about 7.9% of their genomes to an as yet unidentified archaic population. This is in agreement with some results from previous papers in other African populations such as the Biaka and the Baka, suggesting that there was a rich diversity of hominin species within Africa and that introgression was commonplace. (Sankararaman and Durvasula 2018)

This finding is consistent with previous archaeological and genetic evidence, particularly from western and southern Africa. Both regions seem to have had archaic populations until recent times:

- A skull from a Nigerian site (Iwo Eleru) is only about 16,300 years old and yet looks intermediate in shape between modern humans on the one hand and Neanderthals and Homo erectus on the other. It resembles the skull of a near-modern human, like the Skhul-Qafzeh hominins who lived in the Middle East some 80,000 to 100,000 years ago (Harvati et al., 2011; Stojanowski, 2014).

- Genomic analysis of 16 prehistoric Africans suggests that modern humans spread out of eastern Africa and into western Africa, where they mixed with an archaic population as divergent from modern humans as Neanderthals were, the time of separation from modern humans being 200,000 to 300,000 years ago. This archaic admixture is estimated at 9% in Yoruba and 13% in Mende (Skoglund et al. 2017)

- Genomic analysis shows an apparently higher level of Neanderthal ancestry in the Yoruba of Nigeria than in the Luhya of Kenya. This admixture seems to come from a Neanderthal-like population that formerly lived in West Africa (Hawks 2012)

- A skull from Zambia has been dated to 110,000 years ago and yet looks very much like a Homo erectus (Bada et al., 1974; Stringer, 2011). 

-  About 2% of the current African gene pool comes from a population that split from ancestral modern humans some 700,000 years ago. This archaic DNA was then picked up by modern African humans about 35,000 years ago, probably in central Africa because this admixture is highest in pygmy groups from that region (Hammer et al. 2011).

- Genomic analysis of western African pygmies (Biaka and Baka) indicates frequent, low-level interbreeding between archaic and modern humans, including an admixture event within the last 30,000 years (Hsieh et al. 2016). 

- Jawbone fragments from South Africa exhibits significant size and morphological variability, indicating admixture with an archaic population. The fragments fall within the range of 110,000 to 60,000 years ago (Malekfar, 2012)

- Sub-Saharan Africans exhibit dental traits that distinguish them from other modern humans (Sub-Saharan African Dental Complex). These traits are shared with extinct hominids and many extinct and extant nonhuman primates (Irish 1998). When dentitions are compared from western, central, eastern, and southern Africans, these ancestral traits appear to be least present in Kenyans and Tanzanians (Irish 1998). The SSADC thus seems least present in the "homeland" of modern humans (eastern Africa) and more present farther west and south.

Is the estimate of 7.9% archaic admixture a lower bound?

While the new finding of 7.9% archaic admixture is higher than what we see in other modern humans, the actual figure may be higher still. Sankararaman and Durvasula attribute this 7.9% admixture to "a deeply-diverged archaic population," while nonetheless acknowledging the "rich diversity of hominin species within Africa." Dienekes (2018) likewise notes that multiple admixture events had occurred between modern African humans and a range of "Palaeoafrican" groups.

Thus, Sankararaman and Durvasula are measuring admixture only from a highly divergent archaic group, apparently the same one that Skoglund et al. (2017) found in their study of the Yoruba. Indeed, the two studies found almost the same level of archaic admixture in the Yoruba: 7.9% versus 9%. Although Sankararaman and Durvasula validated their methodology with data on Neanderthal admixture in Europe, the two situations are not really comparable. In Europe, modern humans encountered only one archaic group over a relatively short time span, intermixture taking place essentially between 60,000 and 50,000 years ago with a second event more than 37,000 years ago (Yang and Fu 2018).  In Africa, modern humans likely encountered a range of archaic groups over a longer time, including "near-moderns" whose ancestors diverged from those of modern humans less than 200,000 years ago.

If we include introgression from these “near-moderns,” the total for archaic admixture in present-day sub-Saharan Africans should be much higher.  Indeed, 13% of the sub-Saharan gene pool seems to come from a demographic expansion that took place some 111,000 years ago and which probably brought the Skhul-Qafzeh hominins to the Middle East (Watson et al. 1997). Those hominins were anatomically modern, or almost so, but culturally Neanderthal.

Did archaic admixture help or hinder?

Mainstream evolutionary theorists have argued that admixture does more harm than good. As Ernst Mayr (1970, p. 80) wrote:

The claim has been made that species owe much of their genetic variability to introgressive hybridization. However, all the evidence contradicts this conclusion so far as animals is concerned. Not only are F1 hybrids between good species very rare, but where they occur the hybrids (even when not sterile) are demonstrably of inferior viability. The few genes that occasionally introgress into the parental species are not coadapted [...] and are selected against. Introgressive hybridization seems to be a negligible source of genetic variation in animals.

This view has been challenged by Hawks et al. (2007), who argue that gene introgression helped modern humans adapt to new environments. Instead of starting from scratch, they could cherry-pick genes that had already been tried and proven by the populations they were replacing: 

Compared with novel mutations, archaic genetic variants would have had several qualities that, in some cases, may have enhanced their selective value. Because they had long existed within human populations, these alleles had a much lower chance of being strongly deleterious. [...] Alleles with local advantages might never have been selected within the expanding modern population until it reached new climatic regimens. The spread of modern humans may have attained a burst of evolutionary change by drawing on the fruits of the existing adaptations of archaic humans. (Hawks et al. 2007)

The latest findings seem to lie between the above two views. Introgression can in some cases provide useful genes. Usually, however, it’s maladaptive.

We observe a decrease in the frequency of archaic ancestry in the Yoruban populations in more constrained regions of the genome, suggesting that these archaic alleles have been subject to the effects of purifying selection similar to the deleterious consequences of Neanderthal and Denisovan alleles in the modern human genetic background. On the other hand, we find several loci that harbor archaic haplotypes at elevated frequencies (>60%). (Sankararaman and Durvasula 2018)

Similarly, Yang and Fu (2018) note that a "gradual decline in archaic ancestry in Europeans dating from ~37 to 14 ka suggests that purifying selection lowered the amount of Neanderthal ancestry first introduced into ancient modern humans."

This pattern is consistent with findings from nonhuman species. A study of admixture in trout found sharp declines in fitness even with 20% admixture. The decline has two causes:

Hybridization can reduce fitness by either introducing alleles to a population that are not suited to the local environment (extrinsic outbreeding depression) or disrupting co-adapted gene complexes (intrinsic outbreeding depression) (Templeton 1986). These mechanisms are not mutually exclusive, and identifying the contribution of each effect is difficult. However, the high reproductive success of F1 hybrids relative to post-F1 hybrids with similar amounts of admixture suggests that some of the outbreeding depression is intrinsic. (Muhlfeld et al. 2009)

By disrupting co-adapted gene complexes, introgression causes individual genes to lose their adaptive value. Selection will thus eliminate either the introgressed alleles or the previously existing ones. In the second scenario, the complex of co-adapted genes is replaced with a simpler version.


Something “clicked” in eastern Africa 80,000 to 60,000 years ago. A relatively small group of humans acquired a new way of imagining themselves, each other, and the world around them, and this innovation gave them an edge over everyone else. The result: a “big bang” of population growth. They began to spread outward, first within Africa and then into Eurasia.

Their expansion within Africa seems to have proceeded more slowly than in Eurasia. Initially, these modern humans were replacing “near-moderns”—people fairly similar in appearance and genetic makeup. As they pushed farther east and south, however, they encountered populations that were much less similar. West Africa seems to have been home to a people who were as different from modern humans as Neanderthals were, perhaps being related to them. In southern Africa, modern humans encountered people even more divergent: a relic Homo erectus population. Even these highly divergent archaic groups were not rapidly replaced; they may have persisted as late as 15,000 years ago in West Africa and 30,000 years ago in central Africa. Thus, modern and archaic groups seem to have long coexisted in parts of Africa.

In general, archaic admixture reduced fitness: “archaic alleles that introgressed into the Yoruban population were deleterious on average”; neutral alleles were more likely to be retained than those that had functional impacts (Sankararaman and Durvasula 2018). A few, however, seem to have been favored by selection. This is the case with alleles located at a tumor suppressor gene, a gene involved with hormone regulation, and a gene involved with potassium channels. These are individual genes, however, and it is hard to know the impact on co-adapted gene complexes. In theory, archaic admixture should have had a disruptive effect.

Present-day Africans thus have admixture from a range of archaic groups, some being similar to modern humans and others more like Neanderthals or even Homo erectus. This admixture is highest in western and southern Africa and lowest in eastern Africa. In West Africa, admixture from a Neanderthal-like group is estimated at 7.9% by Sankararaman and Durvasula (2018) and at 9 to 13% by Skoglund et al. (2017). Admixture from “near-moderns” is harder to measure. There seems to be a 13% pan-African admixture from a population that had expanded across much of the continent some 111,000 years ago and which perhaps spilled into the Middle East, giving rise to the Skhul-Qafzeh hominins, i.e., early modern humans with Neanderthal culture (Watson et al. 1997, see L1i in Table 2).


Bada, J.L., R.A. Schroeder, R. Protsch, & R. Berger. (1974). Concordance of Collagen-Based Radiocarbon and Aspartic-Acid Racemization Ages, Proceedings of the National Academy of Sciences (USA) 71: 914-917.

Brown, Kyle S.; Marean, Curtis W.; Jacobs, Zenobia; Schoville, Benjamin J.; Oestmo, Simen; Fisher, Erich C.; Bernatchez, Jocelyn; Karkanas, Panagiotis; Matthews, Thalassa (2012). An early and enduring advanced technology originating 71,000 years ago in South Africa. Nature 491 (7425): 590.

Dienekes (2018). Statistical Palaeoafricans, Dienekes' Anthropology Blog, March 25

Durvasula, A., and S. Sankararaman. (2018). Recovering signals of ghost archaic admixture in the genomes of present-day Africans, BioRxiv, March 21  

Hammer, M.F., A.E. Woerner, F.L. Mendez, J.C. Watkins, and J.D. Wall. (2011). Genetic evidence for archaic admixture in Africa, Proceedings of the National Academy of Sciences (USA) 108: 15123-15128.

Harvati, K., C. Stringer, R. Grün, M. Aubert, P. Allsworth-Jones, C.A. Folorunso. (2011). The Later Stone Age Calvaria from Iwo Eleru, Nigeria: Morphology and Chronology. PLoS ONE 6(9): e24024. doi:10.1371/journal.pone.0024024

Hawks, J. (2012). Which population in the 1000 Genomes Project samples has the most Neandertal similarity? John Hawks Weblog, February 8

Hawks, J., G. Cochran, H.C. Harpending, and B.T. Lahn. (2007). A genetic legacy from archaic Homo, Trends in Genetics 24(1): 19-23

Hsieh, P., A.W. Woerner, J.D. Wall, J. Lachance, S.A. Tishkoff, R.N. Gutenkunst, and M.F. Hammer. (2016). Model-based analyses of whole-genome data reveal a complex evolutionary history involving archaic introgression in Central African Pygmies. Genome Research 26(3): 291-300

Irish, J.D. (1998). Ancestral dental traits in recent Sub-Saharan Africans and the origins of modern humans, Journal of Human Evolution 34: 81-98.

Malekfar, L. (2012). An analysis of the Klasies River hominins using a hybrid model, American Journal of Physical Anthropology, Program of the 81st Annual Meeting of the American Association of Physical Anthropologists, p. 201.

Mayr, E. (1970). Populations, Species, and Evolution, Belknap Press: Cambridge (Mass.)

Muhlfeld, C.C.,  S.T Kalinowski, T.E. McMahon, M.L. Taper, S. Painter, R.F. Leary, F.W. Allendorf. (2009). Hybridization rapidly reduces fitness of a native trout in the wild, Biology Letters, March 18

Sankararaman, S., S. Mallick, N. Patterson, and D. Reich. (2016). The combined landscape of Denisovan and Neanderthal ancestry in present-day humans, Current Biology 26(9): 1241-1247.

Skoglund, P., J.C. Thompson, M.E. Prendergast, A. Mittnik, K. Sirak, et al. (2017). Reconstructing Prehistoric African Population Structure, Cell 171(1): 59-71

Stojanowski, C.M. (2014). Iwo Eleru's place among Late Pleistocene and Early Holocene populations of North and East Africa, Journal of Human Evolution 75: 80-89.

Stringer, C. (2011). The chronological and evolutionary position of the Broken Hill cranium. American Journal of Physical Anthropology 144(supp. 52): 287

Watson, E., P. Forster, M. Richards, and H-J. Bandelt. (1997). Mitochondrial footprints of human expansions in Africa, American Journal of Human Genetics 61: 691-704. 0024024

Yang, M.A., and Q. Fu. (2018). Insights into Modern Human Prehistory Using Ancient Genomes, Trends in Genetics 34(3): 184-196


Anonymous said...

The SSADC might be least present in eastern Africa, and more present farther west and south. But E Africa was closest to the source of the backmigration, and a plesiomorphic suite of traits can't be used to diagnose the presence of an unknown, substratal pop/taxon. Frequencies of specific traits like Bushman's canine, might be more informative in this regard.

What do you think of Zubova's claim there were still neanderthal or "archaic" dental traits in the Neolithic populations of Siberia? Ma'lta people definitively cluster with Europeans and other West Eurasians, so when their ancestors reached Siberia, they presumably brought already, the Eurodont suite of mass simplified dental traits. Yet as late as the Neolithic on the Baraba steppes, the local people still had neanderthaloid traits in their teeth. Its hard to interpret this other than the result of admixture, presumably on the stronger scale than happened in Europe or West Asia, unless this was a pleiotropic side effect of some sort. And probably late survival of neanderthal populations, much as mammoths and Irish elk clung on in pockets - demonstrated by Salkhit?

The dental evidence is actually more convincing than the Longlin-Maludong evidence for a supposed new species, of Red Deer Cave people. Isolated populations can show throwbacks, as must be the case with some Neolithic Indians from Kaothe, who show unusual Homo erectus traits in the skull. Though chronologically, erectus grade/clade hominins were long gone by the time sapiens entered India, of course.

Unknown said...

That is quite interesting. Homo Erectus seems to still walk amongst us after all. :)
But I'm a little disappointed in science and its because of things like these. Scientists constantly magically change these numbers of admixture. 15 years ago they told us its proven that Africans and Non-Africans are almost identical genetically and now we learn that somebody obviously made mistakes of some sort! Because now we know there are significant differences in DNA due to admixtures. So can anyone to this day validate a correct number of similar genes between different ethnic groups? I bet not. And to my guess this number might be lower than 90% of similar genes (unlike what we are being told by mainstream science who constantly change their numbers)!

Anonymous said...

Tobias Geiger,
Don't forget how narrative friendly the whole OOA vs MR debate was, and that it existed mostly in the media. Most anthros familiar with the fossil record were quite noncomittal, as were geneticists, because religiously inspired talk about a singular Eve and a localised Eden doesn't reflect speciation in large, mobile mammal species. Nor was there anything scientific about trying to find Eden, by locating the oldest surviving haplogroups, given that haplogroups do die out and relict populations cling on only at the margins. And people forget it was ridiculously biased, so there were people would lie through their teeth that certain Paleolithic fossis, such as Predmosti, were not what we would predict hybrids to look like. Upon analysis of its aDNA, Predmosti turned out to have elevated neanderthal blood after all. Oddly some of the OOA folks committed the same error as their straw man of multiregionalism, faced with such hybrids. Faced with a mosaic neanderthal(?) from Saint Cesaire, they actually invoked the same teleological idea of evolution, independently towards a sapiens grade. Had they not, SC would obviously be a hybrid on the neanderthal end of things, with a mosaic skull and African/sapiens limbs. They would split hairs about whether Skhul 5 was a modern human or a neanderthal, because the narrative was constrained to either-or rubbish. The debate about OOA and multiculturalism wasn't really science, but it managed to tarnish the credibility of physical anthropology and genetics.

Peter Frost said...


Do you mean this study?

There certainly is some Neanderthal or Denisovan admixture, and there seems to have been more of it in older modern human remains in Eurasia. It looks like a lot of the initial admixture was removed over time.

The Tungus of eastern Siberia remember a "monkey people" who may have been an archaic population:


The problem isn't so much science as the way scientific findings are spun in the media.

Anonymous said...

Zubova claims there is 'extreme conservatism' in the molars of people from the Baraba Neolithic, which seems an obvious reference to her earlier description of 'archaicism' in Pleistocene teeth from Western Siberia. The earlier Denisovan teeth also hint at local conservatism on no less than two counts, because Zubova matches them to the dental pattern of Homo heidelbergensis, whilst making comparisons also to prototypical H. erectus from Java, and the seemingly mosaic hominin from Xujiayao (Shanxi/Hebei, 260-370 kya), which has been described as intermediate between AMH and H. erectus, save an oddly neanderthal-like inner ear. Though in fact the complex of plesiomorphic dental traits common to Javan H. erectus and the Denisovans, are otherwise reduced in China: "The dental pattern typical of Denisovans and Xujiayao hominins suggests the presence of one more locus of evolutionary conservation in East Asia." Of most interest is that the unknown species represented at Denisova and Xujiayao themselves had admixture from a Java- or classic Pithecanthropus-type H. erectus, and explicitly not the expected Chinese or Sinanthropus type. This must implicate a late survival of an early Asian H. erectus in Siberia at the time Denisovans arrived, and themselves survived as relict H. heidelbergensis(?) with H. erectus mixture.

The only thing I can think of in Tungusic Siberia, that sounds much like a hominin, are the late accounts of a creature called tchuchunaa or mulen. The source material includes revealing details such as the use of iron tools, so the species encountered must have been Homo sapiens, and not a monkey-like living ancestor. The choice of the word 'monkey' seems to be van Deusen's own attempt to gloss some native word, and likely she confuses the emic with the etic, seeing the absence of monkeys further north than the Japanese forests. How do you gloss a legendary creature, so that casual readers can understand?