Honing in on possible Holocene Contributors to the West African Genepool.
In the last few decades there have been many breakthroughs as it relates to genetic-anthropology. In the motherland of modern humans, such breakthroughs have been lagging. As is, the anthropological catalog of Africa is spotty. Many ancestors of modern day populations have yet to be elucidated irregardless of the discovery of their bones, civilizations or tools in Africa. Ancient DNA analysis have done quite a lot to unveil the mysteries behind many ancient people. But even with such advancements we are left with no definitive explanation behind the development of the peopling responsible for one of the largest expansions of agriculturalists in the world. The Bantu expansion is credited for the distribution of the most prevalent African lineages in Sub-Saharan Africa, but for it’s eventual impact in the genetic and cultural architecture of the continent, explanations of it’s origin leaves a lot to be desired.
If you were enthused by developments in genetic anthropology, you might’ve seen a chart or two with similar tellings as the one here below.
Click here to zoom.
This is an example of an admixture graph where the Yoruba, a West African ethnic group, is shown to have received ancestry from: an unknown group with archaic human admixture vaguely related to East African populations, a second unknown ghost group likely distantly related to Central African hunter gatherers and a third unknown group related to prehistoric North Africans of the Maghreb (made by Revoiye.) Dealing with all of these “unknowns,” makes it easy to over look the practicality in assigning ancestral history. After all, as you can tell by the title, the unknowns are likely known to a degree. In part, individuals or groups overlooked as candidates for West African ancestry could likely be just that. Graphs like the one displayed here were only academically feasible after the sequencing of Ancient Moroccan DNA.1–3 Prior to that, such ancient individuals were seldom compared or considered related to the progenitors of Sub-Saharan Africans, and if they were it was in the context of Eurasian ancestry. Here I want to provide a series of simplified admixture graphs to show relatedness between various African groups with emphasis on West Africans proxied by the Yoruba of Ibadan, Nigeria and the Mada of Cameroon. These graphs further help properly contextualize the genetic make up of West Africans. Those along with some archaeological notes on the morphology and cultures of the West African region including the Sahara should help us uncover the past as it pertains to West Africa. Here you will see some evidence suggesting that West African ancestry coming from more northern regions were likely owed to individuals or populations similar to individuals have discovered in the Malian Sahara. West African ancestors who resided below or in the westernmost extremes the Sudanic belt likely had a complex history relating them to both East Africans and more proximal populations in the western Sahel. And that deep ancestry possibly distantly related to ancient central and east African hunter-gatherers could have had an early and widespread presence in Africa.
Biological Samples from Western Africa and the Sahara.
The most accepted possibility is that the Bantu homeland resided in modern day Cameroon, West Africa, near the border with Nigeria.4–6 With that being the case, the Bantu “marker” is fundamentally West-African. Therefore, in subsequent studies West African biological signals would be the hallmark of Bantu people elsewhere. Ironically though, ancient people of West Africa when discovered in periods rivaling or predating the Bantu expansion, don’t seem to be what we consider “West-African” today. For example, not too far away from the purported Bantu homeland, the archaeological site of Shum Laka had been occupied by people for at least 30k years.7–9 Evidence of occupation has been predominantly marked by the presence of lithics at the site.8 Actual biological samples are fractured, small in number and are representative of the Holocene occupants between 8k and 3k years before present. Preliminary morphological analysis of those samples were only suggestive of potential “Pygmy affinities” due to reduction in stature over time.10 Though the evidence for such a claim was unsubstantial, later aDNA analysis revealed that samples spanning a period of 5 thousand years did show that the individuals there were related to Pygmies. And that they possibly sharing roughly 44% of their Ancestry with modern day Central African hunter-gatherers.3 Needless to say Shum Laka didn’t present biological evidence for being ancestral to, or predominantly related to the ancestors of biological West Africans.
If we look a bit north towards the Sahara, an earlier more complete biological sample was discovered and said to be the earliest West African “Negroid,” dating to roughly between 9.4 and 7kya.11 Subsequent analysis on an undated specimen found in Likasi, Modern day Congo confirmed such associations with Sub-Saharan Africans as the prehistoric Saharan and the individual was described as near identical with the exception of a single cranial measurement; the size of the cranial vault.12 While the Saharan, dubbed Asselar man, did share traits indicative of West African ancestry, he posed traits which lie firmly outside of the range of modern African morphological variation and can be grouped with Biological North Africans (Iberomaurasians and Capsians) under mechtoid traits.13,14 In modern day Nigeria, the Iwo Eleru site contained a skeleton described as a potential donor for West African ancestry dating to 11kya.15,16 However, the sample possessed robust traits which seems to place him out of the range of anatomically modern human variation in a way comparable to non African samples (Upper Paleolithic Eurasians, Skhul, & Qafzeh.)17,18 The sample’s placement seems to be firmly within the spectrum of middle Pleistocene humans possessing archaic traits.17 Though, morphological evidence seems to suggest that they weren’t proto West-African, contribution form this sample into modern West Africans have yet to be ruled out and can be explored with aDNA or subsequent updated analysis.
Regardless, eager to unveil the potential ancestors of West Africans, I revisited some studies and abstracts which highlighted some of the similarities with West Africans and biological samples of the Holocene. Such a distinction can be difficult to pin point due to loose morphological categorization of African samples. Most distinctions of modern African archaeology seem to dichotomize Sub-Saharan and non-African morphology. Where the latter is a broad classification which can include distinct and sometimes unrelated samples and the former reliant on modern Sub-Saharan variation and conjecture as to whether these modern traits were a product of back migration. But nonetheless some inferences can be made and investigated given what we have available.
For instance, a craniometric examination of West Africans has been contextualized with Asselar, a late settlement individual from Shum Laka, a Middle Stone Age (MSA) sample from Egypt; Nazlet Khater, and a contemporaneous sample from the modern day Congo; Ishango.3 Of the ancient samples, Asselar plotted closest to the centroid of Modern West African variation with a position intermediate between them and Ishango. As a whole, modern Sub-Saharan Africans from West and Central Africa seem to show intermediate mandibular traits as it relates to the prehistoric samples, though the Shum Laka individual plotted outside of their variation and is comfortably grouped with Central African Pygmies.3 This analysis by itself isn’t enough to definitively determine a proto-West African but with the autosome of the Shum Laka individuals for comparative analysis with modern West-Central Africans we might be able to piece together the story leading to their genesis. And while Asselar had not been tested for aDNA, proximal samples had been, though not published in full yet.19 Keep this in mind as I’ll revisit this later.
~ Hints from Shum Laka
While the genomes of Shum Laka weren’t fully indicative of being ancestral to modern West Africans there are hints as to what we should expect. As stated before it was discovered that about 44% of their DNA was closely related to Central African Hunter-Gatherers, however the other 56% was loosely related to modern day West Africans.3 How that is as explained by Lipson et al 2020, was that 16% is related to a Ghost North African population and 40 % was related to Basal-West African. It is also worth noting that according to the same model, the “basal West African” lineage was formed by a mixture of a North African related group and a “ghost” African related to the earliest divergent African hunter-gatherer groups. And said ghost population likely inherited minor ancestry related to an unidentified archaic group. A quick linkage disequilibrium test20 of the Shum laka individuals with populations related to each of their ancestral components can reveal the nature or depth in time which their population developed.
More recent than the discoveries at Shum Laka, Ishango, Asselar and Iwo Eleru is the discovery at Gobero, Niger, which encompassed a mass burial containing populations of two occupational phases.14 The Earliest phase referred to as Kiffian is predicted to had lasted between 9.7Kya and 8.2Kya and the later phase, the Tenerean lasted between 7.2Kya and 4.2Kya. Initial reports suggested that climate changes interrupted the settlement at Gobero to which the Earlier population was replaced by an incoming group with a new culture and different subsistence.14,21 The morphological report via craniometrics grouped the early settlement with North Africans as a well as Asselar and other Holocene Saharan Africans, however the later group associated with Tenerean culture clustered a bit farther away with no direct comparison to any population. Culturally, the Kiffians are noticeably African and associated with Nilo-Saharan speakers.22 This is supported by the presence of wavy-line Pottery and barbed bone points which were two cultural elements highlighted on Revoiye prior. The Tenerean culture got it’s label from old archaeological sites discovered in Tenere Niger. They represent the cattle culture of the Sahara for they had appeared after the advent of bos domestication. So far, it has been the skeletons associated with Gobero’s early occupants that had been biologically linked to West Africans, and moreso Bantu speakers via non-metric dental traits.23
So What’s the Hoek
In 2013, a report on another burial site within the Sahara surfaced. At Southwest Libya at Wadi Takarkori.24 Multiple individuals were found in burials segregated mostly by sex across the region. They were likely pasotralists with practices linked to Saharan cattle culture as explained before. Initial analysis of complete skeletons revealed morphological overlap with modern Sub Saharan Africans as well as intermediate clustering with both the early and late settlements at Gobero, whom they are closest to biologically.25,26 Their affinity with those West Saharan groups suggest that both phases at Gobero were related though it seems that based on cranial data, the Tenerean semi-pastoral group was likely influenced by admixture rather than in-situ adaptation.14,27 Post cranial traits of all mentioned individuals indicate that they were likely of African origin, with tropically adapted limbs consistent with Sub-Saharan physiology.25,28 However, it is also true that all individuals at Gobero do show dental affinities with north Africans (including Iberomaurasians), and the Kiffian burials also could have inherited or adopted dental ablation techniques seen in the Paleolithic and Neolithic Maghreb.27,28 Furthermore, two individuals from Takarkori rock shelter have had their genomes sequenced and revealed a novel lineage of macrohaplogroup N (N*).29 Mitochondrial haplogroups M and N were said to diversify outside of Africa though were early descendants from the African haplogroup L3.30 With the discovery of N* in the Sahara, more questions about the diversification process of the mitochondria can be raised considering that Takarkori’s direct or founding lineage hasn’t been found outside of Africa. More comprehensive autosomal data for the individuals at Takarkori will show that they were of a novel ancestry loosely related to early Neolithic and Paleolithic Maghrebi populations whilst having Neanderthal introgression comparable to Sub-Saharan populations. 19
“We find that the majority of the Takarkori individuals’ ancestry stems from a previously unknown lineage that appears to have remained isolated for most of its existence. Both individuals are most closely related to the preceding 15,000-year-old foragers from Morocco associated with the Iberomaurusian techno-complex, whereas both Takarkori and Iberomaurusian individuals are distantly related to sub-Saharan African lineages. The quality of one of the genomes from Takarkori is sufficient to detect prospective Neandertal ancestry and we find evidence for few segments of ancestry that sum to a total comparable to that detected in the genomes of sub-Saharan Africans. Our results therefore support a model of cultural diffusion, rather than human migration, for the emergence of pastoralist subsistence in the Sahara region. “
Abstract from 10th Meeting of the International Society for Biomolecular Archaeology, by Nada Salem & co.
“Genomes from Pastoral Neolithic Sahara reveal ancestral north African lineage”
Things to be aware of
It has been generally accepted that related agriculturalist likely settled in the Sahara prior to the formation of what we consider the modern day West-African genepool. Genetic evidence in the form of uniparental haplogroups shows that the topology of the most prominent Bantu marker (E-M2; E1b1a) highlights distribution from the Sahara.31 The same study shows how that below the Sahara this haplogroup rapidly diversified, and expanded in more recent times.
Simplified Version of E-M2’s Distribution and Phylogeny as seen in D’Antanasio 2019.31 The triangle heights are representative of the age of the most recent ancestors of a sub-branch. Triangle widths correspond to the amount of individuals carrying the mutation. The green band marks the duration of the African Humid Phase. Colored maps highlight the frequency and general distribution in Africa.
Genetic contribution from Stone age samples below the Sahara west to populations such as the Kiffian has not been supported by archeological, morphological or genetic evidence.7,18,27,31 However, Irish 2016 does show that Kiffian non metric traits can easily group them with West Africans from the late Holocene to the modern times, including, Africans found at Shum Laka.23 See the following quote…
“Craniometric data suggest Early Holocene connections to the Maghreb, an inference supported by dental ablation patterns, with no similarity to the Iwo Eleru skull. Dental morphological data, however, suggest a mosaic of north and sub-Saharan African morphology.”
Stojanowski 2017
Further comparisons between the Iberomaurasians and other ancient samples firmly grouped with “Sub-Saharan” populations have been made when considering East African samples. Paleolithic and Neolithic samples from North East Africa, namely, Jebel-Sahaba, Tushka and Wadi Kubbaniya have been grouped with North Africans under mechtiod traits.32 However, their affinity have been widely dismissed due some characteristic differences and overall insignificance of a mechtoid grouping.22 Interestingly enough, though populations related to the West African ancestors of the Sahara can be grouped with North Africans with greater significance, they can also be grouped with the mentioned Northeast Africans to the exclusion of biological North Africans of the Maghreb.14,22 This grouping is further supported by genetic studies of West and East Africans, where it has been reported that Nilo-Saharan groups of East Africa and West Africans could have shared a ancestry dating to the green Sahara.33–36 However, such ancestry likely didn’t facilitate distribution or relatedness of the Bantu marker as haplogroup E-M2 is largely absent from Sudanese Nilotic populations.
Admixture Graphs
I looked to test West African relatedness with some publicly available genomes from prehistoric individuals as well as a handful of modern individuals. The path I wanted to follow was similar to that of Lipson 2020, where they were able to distinctly show the distribution of “Basal West Africans” as well as their ghost North African progenitors.3 Main differences between what was done in that study and here is the methodology for the graph composition and sampling. I used a more recent compilation of tools to best sort formulate relationships between samples using less a prioric assumptions in the process of graph building. However some limitations were implied. And due to the nature of what’s being investigated (the ancestral composition of West Africans), I wanted to make sure that our West African representatives (the Yoruba) were constantly fitted with a minimum of 3 admixture events. Populations constantly held to no admixture were the ancient South African hunter gatherers of Ballito Bay and the Neanderthal of the Altai mountains. This was done to prevent unfeasible and over fitted graphs. 3 modern and ancient reference populations were included in all subsequent analysis. The modern samples included the French as a Eurasian control group, The Biaka pygmy as a Central African control group, and the Mada of Cameroon. The ancient samples were Bayira a boy found in Mota cave Ethiopia ~4.5kya, An ~8kya sample from Shum Laka, Cameroon, and an 7.6kya sample from Ifri Ouberrid cave in Morocco.
Initial 3 Admixture Graphs
(Forcing Ancient Individuals to At Least 1 Admixture event)
Forcing Single Admixture Event for Ifri Ouberrid
Forcing Single Admixture Event for Bayira
Graph3: Forcing Single Admixture Event for Shum Laka
Figure 2. These are Initial graphs allowing 3 Admixture events in total where each is forcing a single ancient individual to have at least one. All three graphs were relatively infeasible, however the third graph holding Shum Laka to a single admixture event (figure 2c) reported the best scores. all of these graphs seems to prefer a case were the Yoruba are a mixture between the Mada and deeper ancestry either similar to East Africans (figure 2b) or Shum Laka / Biaka (figures 2a & 2c).
Forcing 3 Admixture Events for Yoruba
Figure 3. A result of series of graphs processed and compared using Figure 2c as a starting point. Only when allowing for a minimum of 5 events did the graph scores (f4 Z score -outliers) fall below insignificance (Z-score of 3). The main differences between this graph and figure 2c were an addition of basal human or archaic ancestry for the French and another admixture event for Shum Laka which adds additional ancestry from a source related to Bayira/Ancient East Africans.
Making Space for Deeper Ancestry
(2 Graphs)
Maintaining 5 Admixture Events
Allowing 6 Admixture Events
Figure 4. These are a couple of graphs using the Chimpanzee as an outgroup instead of the Neanderthal. Maintaining the limit of 5 events produces infeasible scores but indicates that there is ancestry not accounted for in the diversity of non-West Africans. The same can be deduced when allowing for 6 Admixture events where archaic ancestry is estimated for the Mada who contributes ancestry to Central Africans and the Yoruba.
Adding Mende
Figure 5. In this iteration I added the Mende to investigate differences among West Africans. All graphs generated holding this model to less than 7 events failed. In this case the Mende show’s signs of a similar composition to the Yoruba, both requiring ancestry from a source related to the Mada and a basal human or source more divergent than South African hunter-gatherers. However, the Mende has even more ancestry from the latter source. Also notable, is that including another West African shows that the Aka pygmies (Biaka) received ancestry from West African as opposed to the other way around.
Adding Dinka
Figure 6. Of the graphs printed, adding the Dinka in place of the Mende breaks locality. In this case the Mada and the Yoruba are the only populations held to outlying admixture. The Dinka can be modeled as Mada without archaic ancestry + an ancient East African source. This East African source holds a basal position to Shum Laka’s East African ancestry not found in Pygmies. This graph requires 7 admixture events.
Adding the Toubu
Figure 7. Adding the Toubou yielded a similar graph to adding the Mende (figure 5). The main difference is that the Toubou requires ancestry from a source related to the Epipaleolithic Moroccan and the French possibly hinting at a population with Eurasian admixture in addition to the Mada related ancestry.
Adding Somali
Figure 8. Adding the Somali to the model required 8 admixture events. Interestingly, outside of the addition of the Somali, the major difference between this run and the base model (figure 4) is the dilution of archaic related ancestry in the Mada. The Somali however requires three way admixture from a population similar to the Mada, a Eurasian proxy similar to the Toubou (figure 7) and a population with a similar positioning as the Dinka had in figure 6.
Adding Somali + Dinka
Figure 9. Adding both the Dinka and the Somali required 9 admixture events surprisingly. In this case the Dinka needed additional ancestry related to the central African group contributing ancestry to the Mada. This is possibly due to the Somali having more archaic related ancestry than the Dinka. It’s also worth noting that Shum Laka, the Dinka and the Somali’s additional East African ancestry seems to be undifferentiated.
Adding the Toubu + Mende
Figure 10. The best model printed with both the Toubou and Mende yielded a similar positioning for the West Africans. The Toubou and the Mada in this case now have a similar genesis in which case the Toubou has 2/3rds of their ancestry inherited from a population intermediately positioned between the French and the Epipaleolithic Moroccan. This successful graph only requires 8 admixture events.
Adding Mende, Dinka and the Toubou
Figure 11. The best output using these populations yielded a score slightly above the cutoff for a significant graph. However the output is consistent with being a combination of figures 10 and 6. The extreme Z score is likely due to the discrepancy of archaic related ancestry in the Dinka and Toubou, where as the latter requires it as seen in figures 10 and 7, and the Dinka requires very little to none of this ancestry. See outlying f4 (Chimp Ethiopia_4500BP; Dinka Chad_Toubou). Because of previous outputs, this model was held to 9 admixture events.
Adding Dinka, Toubou and Somali
Figure 12. In aiming for consistency this graph was held to 10 admixture events. See figures 6, 7, and 8. This graph prints as an outlier for similar reasons to the failed graph in figure 11.
A Quick Wrap Up
We can estimate that West African ancestry likely at the very least consists of mixture between three distinct sources as mentioned prior. The most enigmatic source was one which likely had the least impact on Saharan populations overall during the African Humid Phase. Given these details surrounding localization and migratory history, the individuals from the site of Iwo Eleru could have been a part of this complex. How they relate to non-West Africans have not been fully explored here, however it is hinted that this ghost population could have contributed 12-16% of ancestry to West Africans. It also appears that populations located further west have greater proportions of this ancestry relative to their geography. The next two major sources likely merged in Central Africa or the Sahara. One was clearly related to Epipaleolithic and early Neolithic North Africans and the other was related to early hunter-gathers of Central and East Africa and could have had archaic signals possibly due to longstanding population structure41 . It isn’t fully clear yet if Malian Saharans such as Asselar man, or Gobero (late and early) were a product of both these components. However they were highly likely to have been biologically similar to the former.19 It is worth noting that biological dissimilarity between those people and more recent Saharans have been recorded in which the latter can be seen with more archaic and or Sub-Saharan African traits.25,37,38 Nonetheless of the modern populations tested the Mada represents one of the most pristine combinations of these two ancestries. Though they are now represented in modern day Nigeria and Cameroon, their genetic affinity bridges the gap between West and East Africa as they lack additional longstanding local ancestry of either East or West Africa. Populations with more recent and clear non-African ancestry can be modeled as a combination of Mada, Eurasian and Local ancestry(if applicable). This suggests that such an ancestral composition was wide spread in Africa following the humid-phase.
Tools n’ Methods
To formulate the graphs I used the “qpgraph” and “findgraph” functions of the admixtools2 package.39 Positioning in the charts were solely generated using “findgraph” functions and final graphs were calculated using “qpgraph.” Multiple graphs were printed for each iteration to which they were compared and selected for analysis by virtue of “qpgraph_resample_multi” and “compare_graphs” functions. Charts with the lowest inverse likelihood scores were selected at each step. Linkage disequilibrium tests were ran using Malder,20,40 and a table with details of the results can be seen below.
Don’t Be Afraid to Comment, Like and or Share!
References
1. Van De Loosdrecht, M. et al. Pleistocene north african genomes link near eastern and sub-saharan african human populations. Science (80-. ). 360, 548–552 (2018).
2. Lazaridis, I. et al. Paleolithic DNA from the Caucasus reveals core of West Eurasian ancestry. bioRxiv (2018) doi:10.1101/423079.
3. Lipson, M. et al. Ancient West African foragers in the context of African population history. Nature 577, 665–670 (2020).
4. Patin, E. et al. Dispersals and genetic adaptation of Bantu-speaking populations in Africa and North America. Science (80-. ). 356, 543–546 (2017).
5. Fortes-lima, C. A., Burgarella, C., Hammarén, R. & Eriksson, A. The genetic legacy of the expansion of Bantu-speaking peoples in Africa. (2023).
6. Berniell-Lee, G. et al. Genetic and demographic implications of the bantu expansion: Insights from human paternal lineages. Mol. Biol. Evol. 26, 1581–1589 (2009).
7. Brower Stahl, A. African archaeology. Annals of the New York Academy of Sciences vol. 197 (1972).
8. Cornelissen, E. On microlithic quartz industries at the end of the pleistocene in Central Africa: The evidence from Shum Laka (NW Cameroon). African Archaeol. Rev. 20, 1–24 (2003).
9. Lavachery, P. The holocene archaeological sequence of Shum Laka rock shelter (Grassfields, western Cameroon). African Archaeol. Rev. 18, 213–247 (2001).
10. Orban, R., Ribot, I. & Fenaux, S. Les restes humains de Shum Laka. (2000).
11. Boule, M. & Vallois, H. L’homme fossile d’Asselar (Sahara). (1932).
12. Drennan, M. R. REPORT ON THE LIKASI SKELETON. Trans. R. Soc. South Africa 29, 81–89 (1942).
13. Bräuer, G. & Rimbach, K. W. Late archaic and modern Homo sapiens from Europe, Africa, and Southwest Asia: Craniometric comparisons and phylogenetic implications. J. Hum. Evol. 19, 789–807 (1990).
14. Sereno, P. C. et al. Lakeside cemeteries in the Sahara: 5000 years of holocene population and environmental change. PLoS One 3, (2008).
15. Harvati, K. & Stringer, C. The archaeological context of the Iwo Eleru cranium from Nigeriaand Preliminary Results of New Morphometric Studies. 29–42 (1965).
16. Shaw, T. & Daniels, S. G. H. Excavations at iwo eleru, ondo state, Nigeria. West African J. Archaeol. 14, 1–269 (1984).
17. Harvati, K. et al. The later stone age calvaria from Iwo Eleru, Nigeria: Morphology and chronology. PLoS One 6, (2011).
18. Stojanowski, C. M. Iwo Eleru’s place among Late Pleistocene and Early Holocene populations of North and East Africa. J. Hum. Evol. 75, 80–89 (2014).
19. Salem, N., van de Loosdrecht, M., Pelin Sümer, A. & Vai, S. Genomes from Pastoral Neolithic Sahara reveal ancestral north African lineage. Max Planck Inst. Evol. Anthropol. Ger. (2023).
20. Loh, P. R. et al. Inferring admixture histories of human populations using linkage disequilibrium. Genetics 193, 1233–1254 (2013).
21. Riemer, H. Elena A.A. Garcea (ed.). Gobero: the no-return frontier. Archaeology and landscape at the Saharo-Sahelian borderland (Journal of African Archaeology Monographs 9). xviii+293 pages, 196 colour and b&w illustrations, 61 tables. 2013. Frankfurt: Africa M. Antiquity 88, 316–318 (2014).
22. Becker, E. The prehistoric inhabitants of the Wadi Howar : an anthropological study of human skeletal remains from the Sudanese part of the Eastern Sahara. Afrika (2011).
23. Irish, J. D. Tracing the “Bantu Expansion” from its source: Dental nonmetric affinities among West African and neighboring populations. Am. J. Phys. Anthropol. 159, 182 (2016).
24. di Lernia, S. & Tafuri, M. A. Persistent deathplaces and mobile landmarks: The Holocene mortuary and isotopic record from Wadi Takarkori (SW Libya). J. Anthropol. Archaeol. 32, 1–15 (2013).
25. Di Vincenzo, F. et al. Modern Beams for Ancient Mummies Computerized Tomography of the Holocene Mummified Remains From Wadi Takarkori (Acacus, South-Western Libya; Middle Pastoral). Med. Secoli 27, 575–588 (2015).
26. Profico, A. et al. Medical imaging as a taphonomic tool: The naturally-mummified bodies from Takarkori rock shelter (Tadrart Acacus, SW Libya, 6100-5600 uncal BP). J. Cult. Herit. Manag. Sustain. Dev. 10, 144–156 (2020).
27. Stojanowski, C. M. Persistence or Pastoralism: The Challenges of Studying Hunter-Gatherer Resilience in Africa. in Hunter-Gatherer Adaptation and Resilience: A Bioarchaeological Perspective (eds. Temple, D. H. & Stojanowski, C. M.) 193–226 (Cambridge University Press, 2019). doi:DOI: 10.1017/9781316941256.009.
28. Stojanowski, C., Bookman, R. & Carver, C. Population movements throughout northern Africa during the Pleistocene- Holocene transition. Sch. Hum. Evol. Soc. Chang. (2017).
29. Vai, S. et al. Ancestral mitochondrial N lineage from the Neolithic ‘green’ Sahara. Sci. Rep. 9, 1–9 (2019).
30. Metspalu, M. et al. Most of the extant mtDNA boundaries in South and Southwest Asia anatomically modern humans. 25, 1–25.
31. D’Atanasio, E. et al. The peopling of the last Green Sahara revealed by high-coverage resequencing of trans-Saharan patrilineages. Genome Biol. 19, 1–15 (2018).
32. Chamla, M.-C. Les populations anciennes du Sahara et des régions limitrophes. (1968).
33. Mulindwa, J. et al. Genomic evidence for population specific selection in Nilo-Saharan and Niger-Congo linguistic groups in Africa. bioRxiv 186700 (2017).
34. Bird, N. et al. Dense sampling of ethnic groups within African countries reveals fine-scale genetic structure and extensive historical admixture. (2023) doi:10.1126/sciadv.abq2616.
35. Busby, G. B. et al. Admixture into and within sub-Saharan Africa. Elife 5, 1–44 (2016).
36. Shriner, D. & Rotimi, C. N. Genetic history of Chad. Am. J. Phys. Anthropol. 167, 804–812 (2018).
37. Nikita, E. & Crivellaro, F. The Archeaeology of Fazzan: Human Skeletal Remains. in Angewandte Chemie International Edition, 6(11), 951–952. vol. 13 15–38 (2010).
38. Ricci, F., Tafuri, A., Vincenzo, F. Di & Manzi, G. The human skeletal sample from Fewet. 319–362 (2013).
39. Maier, R. et al. On the limits of fitting complex models of population history to f-statistics. Elife 12, 1–62 (2023).
40. Pickrell, J. K. et al. Ancient west Eurasian ancestry in southern and eastern Africa. Proc. Natl. Acad. Sci. U. S. A. 111, 2632–2637 (2014).
41. Ragsdale, A. P. et al. A weakly structured stem for human origins in Africa. 1–12 (2022).
