NetNomad
07-13-2018, 08:41 AM
What is Africa? A human perspective
Luca Pagani1,2, Isabelle Crevecoeur3
1 APE Lab, Department of Biology, University of Padova, Italy
2 Estonian Biocentre, Institute of Genomics, University of Tartu, Estonia
3 UMR 5199 PACEA, CNRS, Université de Bordeaux, Pessac Cedex, France
Africa is, beyond any doubt, the place where the most complete and continuous hominin fossil record is available, ranging from the earliest phases of differentiation from our common ancestors with the chimpanzees to the finding of the most ancient remains of Homo sapiens (White et al. 2003, Clark et al. 2003, Stringer and Galway-Witham 2017, Hublin et al. 2017) ever found to date. Emergence of Homo sapiens in Africa is roughly placed between 300-200kya (thousand years ago), in accordance with genetic data pointing to an early origin before 260kya (Schlebusch et al. 2017). Additionally, fossil and genetic data agree in placing Africa as the starting point of the major expansion that led our species to colonize the whole of Eurasia, Oceania and the Americas from 70kya (Mallick et al. 2016, Malaspinas et al. 2016, Pagani et al. 2016, Reyes-Centeno et al. 2015).
To fully understand the past evolutionary history of our species, as well as the dynamics that led to the Out of Africa (OoA) expansion, it is important to understand what we mean by “Africa”. From a static, geographic point of view such a definition is trivial and, even considering fluctuations of the shore levels, Gibraltar Straight to the North-West, Sinai isthmus to the North-East and Bab el-Mandeb Straight to the East can be regarded as the long-term gateways out of Africa, the latter two traditionally deemed as favoured exit points for the OoA. This said, however, palaeoclimatic understanding of the potential expansion of ecological niches (Breeze et al. 2016) are important to distinguish what should be seen as crossing of a major barrier, from what was simply a temporary stretching of an African environment across the above mentioned gateways. Palaeoanthropological research has indeed pointed to the Levant (Stringer et al. 1989, Grün et al. 2005, Hershkovitz et al. 2018) and more recently even to Arabia (Groucutt et al. 2018) as places inhabited by humans well before the major 70kya OoA expansion. So should these early remains be considered as failed OoA expansions, or should one simply assume that the African human niche stretched all the way to Levant or to the Arabian peninsula during certain time periods and under certain conditions? And, in contemplating this second hypothesis, how far from Africa should a human remain dating between 200 and 70kya be found to be considered “non-African” both geographically and ecologically?
The permeability of the African boundaries highlighted by the palaeoanthropological research can be invoked, in an opposite sense, when looking at linguistic evidence. Afro-Asiatic languages, one of the four major linguistic groups present in Africa today, are spoken throughout most of North Africa, across the Sahara Desert, along the Sahel and in the Horn of Africa. Given the high diversification and deep splits between these languages, one should assume that such a linguistic family may have originated in Africa. A notable exception, however, is represented by Semitic languages (an Afro-Asiatic sub-family), which are thought to have arisen in the Levant and subsequently expanded in North Africa and in the Horn of Africa within the last 3-4ky (Kitchen et al. 2009). Given the dually “deeply-rooted” presence of Afro-Asiatic languages both in Africa and in the Levant, the linguistic debate on the origin of this family is still open (Kitchen et al. 2009, Ehret et al. 2004) and probably settling on an intermediate “across-the-Sinai” solution. This shows that even relatively well studied cultural packages such as languages point to early interactions between Africa and the neighbouring Eurasian cultures or, in other words, to a geographical shrinking of what can currently be defined as “strictly African” in a long term perspective.
The definition of what is “African” from a genetic point of view is not easier. On one hand, the availability of uniparental markers (chromosome Y and mitochondrial DNA) and of whole genome sequences has enabled us to clearly pinpoint a signature of the OoA expansion. The majority of contemporary humans tracing their ancestry to Eurasian, Oceanian or American populations indeed host uniparental markers belonging to a specific subset of the entire human variability. Such a subset is thought to have originated in Africa and, in its derived form, to have expanded out of Africa while the remaining human uniparental lineages did not (Soares et al. 2012, Karmin et al. 2015). Similarly, the genomes of non-Africans harbour the signature of a bottleneck, or reduction in overall diversity, which can be dated at around 70kya (Malaspinas et al. 2016). As said, this provides an excellent toolset to define whether the ancestors of a given individual took part to the OoA expansion. The scenario, however, is not as clear for contemporary African individuals, which in turn complicates the understanding of the whereabouts of the OoA expansion. Since the OoA, several waves of population back-flow, from Eurasia towards Africa, took place. Genetic data from modern African populations unveiled the presence of major Eurasian genomic components in North (Henn et al. 2012, Pagani et al. 2015), East (Pagani et al. 2012, van Dorp et al. 2015) and South Africa (Pickrell et al. 2014), and dated them to ~800, ~3,000 and ~1,500 years ago, respectively. It has already been shown (Pagani et al. 2012, Pagani et al. 2015, Schlebusch et al. 2017) that proper handling of these “non-African” segments may improve the understanding of the autochthonous African human diversity and, in some cases (Schlebusch et al. 2017), drastically revise our understanding of within Africa human population divergence.
Recent availability of African ancient DNA specimens, however, showed that the Eurasian presence in the continent may not be as recent as understood from modern data. On one hand, ancient samples from East Africa (Gallego Llorente et al. 2015) and South Africa (Skoglund et al. 2017, Schlebusch et al. 2017) pre-dating the inferred arrival of the Eurasian component in the area were, as expected, free from non-African genetic traces. On the other hand, Egyptian mummies from more than 2,000 years ago, hence pre-dating by at least 1000 years the inferred Eurasian migration, could be essentially described as “non-Africans” from a genetic perspective (Schuenemann et al. 2017). Similarly, and perhaps more unexpectedly, post-Neolithic remains from Morocco dating back to around 7kya (Fregel et al. 2018) already showed a predominant Eurasian genetic signature. These recent findings point to i) a long term genetic relationship between North Africa and West Eurasia and ii) potential biases towards more recent dates of current available methods for dating admixture events in modern populations. The relative isolation of North Africa from the rest of the continent, often referred to as “Sub-Saharan Africa” due to the strong separation represented by the Sahara Desert, may imply that the consequences of these recent findings have only a limited impact on our understanding of the broader African demographic history. Such a reassurance may not hold in the light of pre-Neolithic findings, once again from Morocco. Human remains found at Taforalt and dating back to as early as 15kya (van de Loosdrecht et al. 2018) pre-date the Eurasian presence in the area by another 8,000 years and point to the Levantine Natufians (or pre-Natufians) as a plausible genetic source population. Such a deeply rooted, pre-Neolithic interaction between North Africa and West Asia may lead to envisage at least two (entirely speculative) scenarios.
The first one concerns the potential implications for a more, widespread presence of West Eurasian traces in contemporary West and Central Sub-Saharan populations. Currently available data points to the presence of Natufian-like or, in general, West Asian components in Morocco 15kya (van de Loosdrecht et al. 2018) and 7kya (Fregel et al. 2018). The period in between these two dates is known as the “African Humid Period”and was characterized by higher carrying capacity within the broader North African region and, among other features, by the so called “Green Sahara” (D’Atanasio et al. 2018). Such an ecological scenario may have facilitated some level of gene flow of the Natufian-like component attested in Morocco, towards the ancestors of contemporary West and Central Africans, which may have, in turn, spread over most Sub-Saharan populations, hitchhiking on the post-Neolithic expansion of Bantu speaking groups (de Filippo et al. 2012). If this was the case, the wide-spread nature of this component in a gradient-like manner all over Sub-Saharan African populations may make its detection particularly difficult. Overlooking such a presence may reflect in over- or under-estimating splits between Sub-Saharan populations, as already shown by Schlebusch and colleagues (Schlebusch et al. 2017) for the more recent Eurasian components in southern Africans hunter-gatherer populations. Therefore, one may wonder whether the closer genetic relationship between West and Central Africans to Eurasians, compared to South African hunter-gatherer groups, should be seen as a simple South-to-North African differentiation towards East Africans and Eurasians, or whether the above mentioned Natufian-like leaking across the Green Sahara may partly account for such a genetic relationship. Furthermore, abrupt changes in the Natufian-like proportion within neighbouring African populations may create a genetic signature similar to the presence of deeply-rooted genetic components within the population with the smallest Eurasian fraction, similarly to what reported by Skoglund and colleagues (Skoglund et al. 2017). A potential way to rule out this speculation would be to check whether genomic regions that appear as deeply divergent in population A are also particularly attracted by Eurasian-like genomes in population B used as “African reference” in this particular example. Additionally, one may apply local ancestry or chromopainter (Lawson et al. 2012) approaches to mask out genetic regions that are showing a Eurasian-like signature within contemporary African genomes. If these regions are simply reflecting shared ancestry between any African and any Eurasian population, rather than the presence of a Natufian-like genetic leaking across the Green Sahara, then they should be equally present in all African populations and the resulting population split estimates should not be affected by the masking procedure.
The second scenario that stems from the presence of a non-African component in Morocco at least 15kya is a more radical one, and it reflects back on what we may call, genetically, Africa. Putting together genetic evidence for a Northern exit OoA (Pagani et al. 2015) with the archaeological and palaeoclimatic evidence for a drastic reduction of human presence along the lower Nile Valley from MSI4 (60-70kya) until 25kya (Vermeersch et al., 1990; Van Peer, 2004; Vermeersch& Van Neer, 2015), one may postulate that the progressive drying out of the North-East African region from 70kya triggered a population fragmentation in the area. The majority of these fragmented human groups may have, eventually, died out or merged back with the broader Sub-Saharan population. A small subset of them may instead have reached the Mediterranean shores and, subsequently, expanded West, along the North African coast and East, towards Eurasia and eventually admixed with and replaced pre-existing Neanderthal groups in West Asia (Green et al. 2010). This hypothesis is consistent with the fossil record in lower Egypt and upper Nubia during this Late Pleistocene period and until the beginning of the Holocene. Indeed, the oldest and only complete skeleton in this area, the Nazlet Khater 2 individual, dated around 40 Kya, exhibits a combination of plesiomorphic and unique morphometric features that lie outside of extant phenotipyc diversity and could be the consequence of fractioning population in the middle part of the Nile valley at a time of extreme climatic conditions (Crevecoeur, 2008). Such robust phenotype can be detected in subsequent samples in the area dated at around 14-12ky (Wendorf, 1968; Anderson, 1968; Irish, 2005; Crevecoeur, 2008; Antoine et al., 2013) and up to 9kya (Honegger, 2006; Crevecoeur, 2012; Honegger & Williams, 2015), while the fossil record in the northern part of the Nile Valley between 70 and 10.5kya has so far yielded little or no evidence of human presence (Kuper and Kröpelin 2006 Science). Human remains from the Nubian region dating to after 8kya show instead more gracile morphologies (Honegger & Williams, 2015, Crevecoeur 2012), consistent with a reopening of the Nile Valley corridor. The admixture of allegedly African-like middle/lower Nile Valley “more robust” groups with Levantine or North-West African “more gracile” groups after 8kya is compatible with contemporary Egyptians harbouring within their African genetic components (hence the component contributed by the more robust groups) the haplotypes that are most represented outside of Africa today (Pagani et al. 2015). Such a scenario would imply the following: i) a potential cause for the genetic bottleneck that characterizes all non-African group was the progressive increase of aridity of the Nile corridor, rather than the expansion out of Africa of a few wanderers during environmentally permissive conditions; ii) such a bottleneck did not take place at the gateways of Africa but, rather, within Africa (along the Nile basin), and iii) as a consequence of the MIS4 arid period, all subsequent coastal North African populations should be considered “non-Africans” from a genetic viewpoint or, in other words, they should be expected to share the same 70kya genetic bottleneck signature that characterizes all contemporary non-Africans. An interesting implication of this highly speculative scenario would be the presence, in North-Africa, of genetically non-African populations since 60-70kya and, if our understanding of the whereabouts of the human-Neanderthal introgression are correct, then these North African populations should be predominantly free from any archaic genetic signature. Recent ancient DNA studies on West Asian and European populations inferred the presence of a “Basal Eurasian” population (Lazaridis et al. 2014), which is thought to have originated from the same OoA expansion, but to have diverged from other human groups prior to the separation between extant West and East Eurasians and, in all likelihood, even prior to the Neanderthal introgression event (~55kya, Fu et al. 2014). Notably, among modern and ancient populations for which DNA is available to date, Levantine Natufians are among the ones with the highest presence of this so-called “Basal Eurasian” component (Lazaridis et al. 2016). The hypothetical scenario outlined here may therefore point to an identification of Basal Eurasian groups with the populations that inhabited North Africa after the bottleneck event that took place 60-70kya along the Nile Valley and that acted as a Neanderthal-free, non-African reservoir. The separation between Basal Eurasian groups in North Africa and other Eurasians out of Africa may have been facilitated by the reported reduction or absence of human presence along the northern Nile valley between 70kya and 10kya (Vermeersch & Van Neer, 2015, Kuper & Kröpelin, 2006). These two groups may have come again into contact after 25kya, consistently with a potential coastal connection pointed out by the Levantine (Belfer-Cohen and Goring-Morris 2014) and North African (Douka et al. 2014, Jacobs et al., 2017. Barton et al. 2013) material culture. The re-opening of the North Africa – West Asia connection may then have facilitated the entrance of the basal Eurasian component into Eurasia (Lazaridis et al. 2014, Lazaridis et al. 2016) as well as the arrival of Levantine components in North-West Africa. In this light, the Natufian-like component in 15kya Taforalt (van de Loosdrecht et al. 2018) may be seen as an autochthonous North African component enriched by Levantine contacts post 25kya, while the Sub-Saharan signature within the same genomes as the consequence of a relatively recent gene flow from Sub-Saharan human groups. Morphological or genetic analyses on North African human remains dating between 70 and 25kya are needed to conclude whether the Sinai isthmus or the Saharan belt are to be considered as the ultimate gateways of Africa.
Luca Pagani1,2, Isabelle Crevecoeur3
1 APE Lab, Department of Biology, University of Padova, Italy
2 Estonian Biocentre, Institute of Genomics, University of Tartu, Estonia
3 UMR 5199 PACEA, CNRS, Université de Bordeaux, Pessac Cedex, France
Africa is, beyond any doubt, the place where the most complete and continuous hominin fossil record is available, ranging from the earliest phases of differentiation from our common ancestors with the chimpanzees to the finding of the most ancient remains of Homo sapiens (White et al. 2003, Clark et al. 2003, Stringer and Galway-Witham 2017, Hublin et al. 2017) ever found to date. Emergence of Homo sapiens in Africa is roughly placed between 300-200kya (thousand years ago), in accordance with genetic data pointing to an early origin before 260kya (Schlebusch et al. 2017). Additionally, fossil and genetic data agree in placing Africa as the starting point of the major expansion that led our species to colonize the whole of Eurasia, Oceania and the Americas from 70kya (Mallick et al. 2016, Malaspinas et al. 2016, Pagani et al. 2016, Reyes-Centeno et al. 2015).
To fully understand the past evolutionary history of our species, as well as the dynamics that led to the Out of Africa (OoA) expansion, it is important to understand what we mean by “Africa”. From a static, geographic point of view such a definition is trivial and, even considering fluctuations of the shore levels, Gibraltar Straight to the North-West, Sinai isthmus to the North-East and Bab el-Mandeb Straight to the East can be regarded as the long-term gateways out of Africa, the latter two traditionally deemed as favoured exit points for the OoA. This said, however, palaeoclimatic understanding of the potential expansion of ecological niches (Breeze et al. 2016) are important to distinguish what should be seen as crossing of a major barrier, from what was simply a temporary stretching of an African environment across the above mentioned gateways. Palaeoanthropological research has indeed pointed to the Levant (Stringer et al. 1989, Grün et al. 2005, Hershkovitz et al. 2018) and more recently even to Arabia (Groucutt et al. 2018) as places inhabited by humans well before the major 70kya OoA expansion. So should these early remains be considered as failed OoA expansions, or should one simply assume that the African human niche stretched all the way to Levant or to the Arabian peninsula during certain time periods and under certain conditions? And, in contemplating this second hypothesis, how far from Africa should a human remain dating between 200 and 70kya be found to be considered “non-African” both geographically and ecologically?
The permeability of the African boundaries highlighted by the palaeoanthropological research can be invoked, in an opposite sense, when looking at linguistic evidence. Afro-Asiatic languages, one of the four major linguistic groups present in Africa today, are spoken throughout most of North Africa, across the Sahara Desert, along the Sahel and in the Horn of Africa. Given the high diversification and deep splits between these languages, one should assume that such a linguistic family may have originated in Africa. A notable exception, however, is represented by Semitic languages (an Afro-Asiatic sub-family), which are thought to have arisen in the Levant and subsequently expanded in North Africa and in the Horn of Africa within the last 3-4ky (Kitchen et al. 2009). Given the dually “deeply-rooted” presence of Afro-Asiatic languages both in Africa and in the Levant, the linguistic debate on the origin of this family is still open (Kitchen et al. 2009, Ehret et al. 2004) and probably settling on an intermediate “across-the-Sinai” solution. This shows that even relatively well studied cultural packages such as languages point to early interactions between Africa and the neighbouring Eurasian cultures or, in other words, to a geographical shrinking of what can currently be defined as “strictly African” in a long term perspective.
The definition of what is “African” from a genetic point of view is not easier. On one hand, the availability of uniparental markers (chromosome Y and mitochondrial DNA) and of whole genome sequences has enabled us to clearly pinpoint a signature of the OoA expansion. The majority of contemporary humans tracing their ancestry to Eurasian, Oceanian or American populations indeed host uniparental markers belonging to a specific subset of the entire human variability. Such a subset is thought to have originated in Africa and, in its derived form, to have expanded out of Africa while the remaining human uniparental lineages did not (Soares et al. 2012, Karmin et al. 2015). Similarly, the genomes of non-Africans harbour the signature of a bottleneck, or reduction in overall diversity, which can be dated at around 70kya (Malaspinas et al. 2016). As said, this provides an excellent toolset to define whether the ancestors of a given individual took part to the OoA expansion. The scenario, however, is not as clear for contemporary African individuals, which in turn complicates the understanding of the whereabouts of the OoA expansion. Since the OoA, several waves of population back-flow, from Eurasia towards Africa, took place. Genetic data from modern African populations unveiled the presence of major Eurasian genomic components in North (Henn et al. 2012, Pagani et al. 2015), East (Pagani et al. 2012, van Dorp et al. 2015) and South Africa (Pickrell et al. 2014), and dated them to ~800, ~3,000 and ~1,500 years ago, respectively. It has already been shown (Pagani et al. 2012, Pagani et al. 2015, Schlebusch et al. 2017) that proper handling of these “non-African” segments may improve the understanding of the autochthonous African human diversity and, in some cases (Schlebusch et al. 2017), drastically revise our understanding of within Africa human population divergence.
Recent availability of African ancient DNA specimens, however, showed that the Eurasian presence in the continent may not be as recent as understood from modern data. On one hand, ancient samples from East Africa (Gallego Llorente et al. 2015) and South Africa (Skoglund et al. 2017, Schlebusch et al. 2017) pre-dating the inferred arrival of the Eurasian component in the area were, as expected, free from non-African genetic traces. On the other hand, Egyptian mummies from more than 2,000 years ago, hence pre-dating by at least 1000 years the inferred Eurasian migration, could be essentially described as “non-Africans” from a genetic perspective (Schuenemann et al. 2017). Similarly, and perhaps more unexpectedly, post-Neolithic remains from Morocco dating back to around 7kya (Fregel et al. 2018) already showed a predominant Eurasian genetic signature. These recent findings point to i) a long term genetic relationship between North Africa and West Eurasia and ii) potential biases towards more recent dates of current available methods for dating admixture events in modern populations. The relative isolation of North Africa from the rest of the continent, often referred to as “Sub-Saharan Africa” due to the strong separation represented by the Sahara Desert, may imply that the consequences of these recent findings have only a limited impact on our understanding of the broader African demographic history. Such a reassurance may not hold in the light of pre-Neolithic findings, once again from Morocco. Human remains found at Taforalt and dating back to as early as 15kya (van de Loosdrecht et al. 2018) pre-date the Eurasian presence in the area by another 8,000 years and point to the Levantine Natufians (or pre-Natufians) as a plausible genetic source population. Such a deeply rooted, pre-Neolithic interaction between North Africa and West Asia may lead to envisage at least two (entirely speculative) scenarios.
The first one concerns the potential implications for a more, widespread presence of West Eurasian traces in contemporary West and Central Sub-Saharan populations. Currently available data points to the presence of Natufian-like or, in general, West Asian components in Morocco 15kya (van de Loosdrecht et al. 2018) and 7kya (Fregel et al. 2018). The period in between these two dates is known as the “African Humid Period”and was characterized by higher carrying capacity within the broader North African region and, among other features, by the so called “Green Sahara” (D’Atanasio et al. 2018). Such an ecological scenario may have facilitated some level of gene flow of the Natufian-like component attested in Morocco, towards the ancestors of contemporary West and Central Africans, which may have, in turn, spread over most Sub-Saharan populations, hitchhiking on the post-Neolithic expansion of Bantu speaking groups (de Filippo et al. 2012). If this was the case, the wide-spread nature of this component in a gradient-like manner all over Sub-Saharan African populations may make its detection particularly difficult. Overlooking such a presence may reflect in over- or under-estimating splits between Sub-Saharan populations, as already shown by Schlebusch and colleagues (Schlebusch et al. 2017) for the more recent Eurasian components in southern Africans hunter-gatherer populations. Therefore, one may wonder whether the closer genetic relationship between West and Central Africans to Eurasians, compared to South African hunter-gatherer groups, should be seen as a simple South-to-North African differentiation towards East Africans and Eurasians, or whether the above mentioned Natufian-like leaking across the Green Sahara may partly account for such a genetic relationship. Furthermore, abrupt changes in the Natufian-like proportion within neighbouring African populations may create a genetic signature similar to the presence of deeply-rooted genetic components within the population with the smallest Eurasian fraction, similarly to what reported by Skoglund and colleagues (Skoglund et al. 2017). A potential way to rule out this speculation would be to check whether genomic regions that appear as deeply divergent in population A are also particularly attracted by Eurasian-like genomes in population B used as “African reference” in this particular example. Additionally, one may apply local ancestry or chromopainter (Lawson et al. 2012) approaches to mask out genetic regions that are showing a Eurasian-like signature within contemporary African genomes. If these regions are simply reflecting shared ancestry between any African and any Eurasian population, rather than the presence of a Natufian-like genetic leaking across the Green Sahara, then they should be equally present in all African populations and the resulting population split estimates should not be affected by the masking procedure.
The second scenario that stems from the presence of a non-African component in Morocco at least 15kya is a more radical one, and it reflects back on what we may call, genetically, Africa. Putting together genetic evidence for a Northern exit OoA (Pagani et al. 2015) with the archaeological and palaeoclimatic evidence for a drastic reduction of human presence along the lower Nile Valley from MSI4 (60-70kya) until 25kya (Vermeersch et al., 1990; Van Peer, 2004; Vermeersch& Van Neer, 2015), one may postulate that the progressive drying out of the North-East African region from 70kya triggered a population fragmentation in the area. The majority of these fragmented human groups may have, eventually, died out or merged back with the broader Sub-Saharan population. A small subset of them may instead have reached the Mediterranean shores and, subsequently, expanded West, along the North African coast and East, towards Eurasia and eventually admixed with and replaced pre-existing Neanderthal groups in West Asia (Green et al. 2010). This hypothesis is consistent with the fossil record in lower Egypt and upper Nubia during this Late Pleistocene period and until the beginning of the Holocene. Indeed, the oldest and only complete skeleton in this area, the Nazlet Khater 2 individual, dated around 40 Kya, exhibits a combination of plesiomorphic and unique morphometric features that lie outside of extant phenotipyc diversity and could be the consequence of fractioning population in the middle part of the Nile valley at a time of extreme climatic conditions (Crevecoeur, 2008). Such robust phenotype can be detected in subsequent samples in the area dated at around 14-12ky (Wendorf, 1968; Anderson, 1968; Irish, 2005; Crevecoeur, 2008; Antoine et al., 2013) and up to 9kya (Honegger, 2006; Crevecoeur, 2012; Honegger & Williams, 2015), while the fossil record in the northern part of the Nile Valley between 70 and 10.5kya has so far yielded little or no evidence of human presence (Kuper and Kröpelin 2006 Science). Human remains from the Nubian region dating to after 8kya show instead more gracile morphologies (Honegger & Williams, 2015, Crevecoeur 2012), consistent with a reopening of the Nile Valley corridor. The admixture of allegedly African-like middle/lower Nile Valley “more robust” groups with Levantine or North-West African “more gracile” groups after 8kya is compatible with contemporary Egyptians harbouring within their African genetic components (hence the component contributed by the more robust groups) the haplotypes that are most represented outside of Africa today (Pagani et al. 2015). Such a scenario would imply the following: i) a potential cause for the genetic bottleneck that characterizes all non-African group was the progressive increase of aridity of the Nile corridor, rather than the expansion out of Africa of a few wanderers during environmentally permissive conditions; ii) such a bottleneck did not take place at the gateways of Africa but, rather, within Africa (along the Nile basin), and iii) as a consequence of the MIS4 arid period, all subsequent coastal North African populations should be considered “non-Africans” from a genetic viewpoint or, in other words, they should be expected to share the same 70kya genetic bottleneck signature that characterizes all contemporary non-Africans. An interesting implication of this highly speculative scenario would be the presence, in North-Africa, of genetically non-African populations since 60-70kya and, if our understanding of the whereabouts of the human-Neanderthal introgression are correct, then these North African populations should be predominantly free from any archaic genetic signature. Recent ancient DNA studies on West Asian and European populations inferred the presence of a “Basal Eurasian” population (Lazaridis et al. 2014), which is thought to have originated from the same OoA expansion, but to have diverged from other human groups prior to the separation between extant West and East Eurasians and, in all likelihood, even prior to the Neanderthal introgression event (~55kya, Fu et al. 2014). Notably, among modern and ancient populations for which DNA is available to date, Levantine Natufians are among the ones with the highest presence of this so-called “Basal Eurasian” component (Lazaridis et al. 2016). The hypothetical scenario outlined here may therefore point to an identification of Basal Eurasian groups with the populations that inhabited North Africa after the bottleneck event that took place 60-70kya along the Nile Valley and that acted as a Neanderthal-free, non-African reservoir. The separation between Basal Eurasian groups in North Africa and other Eurasians out of Africa may have been facilitated by the reported reduction or absence of human presence along the northern Nile valley between 70kya and 10kya (Vermeersch & Van Neer, 2015, Kuper & Kröpelin, 2006). These two groups may have come again into contact after 25kya, consistently with a potential coastal connection pointed out by the Levantine (Belfer-Cohen and Goring-Morris 2014) and North African (Douka et al. 2014, Jacobs et al., 2017. Barton et al. 2013) material culture. The re-opening of the North Africa – West Asia connection may then have facilitated the entrance of the basal Eurasian component into Eurasia (Lazaridis et al. 2014, Lazaridis et al. 2016) as well as the arrival of Levantine components in North-West Africa. In this light, the Natufian-like component in 15kya Taforalt (van de Loosdrecht et al. 2018) may be seen as an autochthonous North African component enriched by Levantine contacts post 25kya, while the Sub-Saharan signature within the same genomes as the consequence of a relatively recent gene flow from Sub-Saharan human groups. Morphological or genetic analyses on North African human remains dating between 70 and 25kya are needed to conclude whether the Sinai isthmus or the Saharan belt are to be considered as the ultimate gateways of Africa.