John, I’m puzzled by the apparent 20,000-year gap between 70,000 years ago when early modern humans crossed into southwest Asia and 50,000-to-43,000 years ago when that founder population started interbreeding with Neanderthals in the same region.
Did modern people stay in one relatively confined area without expanding or wandering for 20,000 years? Did they simply not encounter Neanderthals, or did they encounter them without interbreeding?
Or did something change in the modern human population or culture in that 20,000-year period which suddenly caused them to surge out of their once-confined area, start expanding their geographic range and continue to do so on an ongoing basis, and start interbreeding with the Neanderthals (and Denisovans?) they encountered?
This is a great question! We're all puzzled a bit by this timeline. Let's start with what we know.
1. There was a bottleneck of this founder population long enough to cut its genetic variation by a substantial degree. This requires a length of time roughly on par with the effective number of individuals in the population. 1000 people would take roughly 1000 generations, or 20,000–30,000 years. This is a ballpark estimate only. A smaller effective size could be a shorter bottleneck and we're talking order of magnitude, not precise estimates.
2. All bottlenecks have a tighter effect on mtDNA and Y chromosome systems than on the autosomal DNA. By our current way of looking at these systems, they seem to indicate a separation from African groups in that time range of roughly 75,000 to 65,000 years ago. The autosomal estimates of this separation time are harder to pin down, largely due to the need to somehow correct for later gene flow after the founders started expanding. Ancient DNA would help a great deal.
3. The release from the bottleneck is better timed, although there is an ongoing mismatch between mtDNA and autosomal estimates for this. That mismatch is giving rise to the disagreement between long and short chronologies for island SE Asia, for example. Short chronology is expansion after 50,000 years, long chronology expansion after 70,000 years. (Long chronology consequently pushes back the bottleneck onset to 90,000+) At the moment the short chronology is supported by genome-wide LD including comparisons that work in ancient genomes; the long chronology supported mainly by mtDNA also including some ancient genomes and certain mutation rate assumptions.
Your question about what the founders were doing while they hung around bottlenecked in North Africa or Southwest Asia is the right one. We have no idea. We can't identify this population archaeologically, and from the genetic point of view it's a ghost.
Southern route proponents would probably say that the founders were taking refuge in Green Arabia at that time, pushed south by the Neanderthal resurgence of Amud-Kebara times. An alternative is that this was a lower Nile valley or coastal Egypt-Libya population that was just waiting for the right release point.
If the short chronology is correct, then it's a good hypothesis that the initial Upper Paleolithic in the Levant represents the early descendants of the expanding founder population. We are talking about 3000 years to get them from the Levant across Anatolia, the Caucasus, and well into the Zagros, and that can be made to fit the chronology of IUP. I'm wary of shoehorning, but the timeline doesn't give me vapors there.
I believe uniparental and autosomal genomic data are actually fairly consistent with each other. Autosomal results in many recent papers show a strong bottleneck peaking around 70-60 kY ago. Some are actually more specific, e.g. TMRCA for all Eurasians show an abrupt cutoff at 70-75 kY in this picture:
from the 2020 paper "“Insights into human genetic variation and population history from 929 diverse genomes”.
It coincides with the emergence of CT->{CF,DE}->{C,F,D,E} Y-chromosome lines and M and N mtDNA macrohaplogroups. I believe the quick emergence of these major haplogroups ancestral to all Eurasians can be best explained by a split of early humans into small isolated groups during the bottleneck and separate genetic fixation in each of these surviving groups.
Then around 55 kY ago there is an explosion of many new haplogroups and beginning of exit from the bottleneck according to autosomal data. This likely indicate the start of a vigorous demographic and geographic expansion at that time.
To follow up on my earlier question, is there any evidence in the genome or in physical archeology of some kind of physical, cultural or technological advancement during this 20,000-year bottleneck, such as an improved toolkit or signs of enhanced vocal articulation or symbolic behavior, which could have given this founder population a competitive advantage over other modern human populations as well as Neanderthals and Denisovans, to enable their ensuing dispersal and growth, analogous to Neolithic agriculture or (perhaps) the Zamnaya’s spoked wheel, or is their dramatic expansion primarily attributed to opportunities created by climatic change?
This is an astute question that gets to the point of why there was such a difference in population intrinsic growth after the bottleneck. It deserves its own post and hopefully I'll be able to return to the topic soon. My intuition is that the mixture itself is part of the explanation.
Not only the archaic admixture itself, but I also feel like there had to have been some sort of selection on the OOA populations sex chromosomes themselves that led to some sort of fertility boosting effect. I hope some lab really does try and pursue this angle more in the not too distant future.
One more conclusion one can make from the Neanderthal admixture geographic distribution - refutation of an old myth of the "rapid coastal travel out of Africa". It was never more than a purely mental construct without any archaeological or genetic evidence, but was largely brought to explain early human presence in SEA and Australia.
This myth (and corresponding maps) still fills nearly all anthropological textbooks and museums.
This just never happened.
If early humans settled the Indian ocean coast first and then expanded inland, the geography of Neanderthal introgression would be very different, of course. But data shows that N. admixture in any ancient indigenous people in SEA (including such extremely old isolates as Andamans) is very similar to other Asians. It is clear that modern humans came to SEA from the north (all the way from Siberia), not along the coast. This is also confirmed by a shockingly small genetic distance between Tianyuan man and modern Andamanese, as well as prevalence of old Y-DNA F->K and C lineages in indigenous SEA and Sahul region.
As for myself I agree with this. I can't entirely render the point of view of the "southern route" proponents, because I don't see the evidence for it. The Andaman Islands DNA evidence is very important. On this point I think they have to lean on the idea of later migrations covering up the first wave. (Or vague hand-waving about glacial sea levels covering up the evidence).
I think the best proxy to the geographic distribution of the Neanderthal admixture is neither latitudinal nor longitudinal, but the "Movius line" - an old archaeological concept, that divided Eurasia roughly in the Southwest and Northeast parts. Basically, migration patterns of early humans in Eurasia were either south of Himalayas and west of the mountain range around the Bangladesh-Burma border, or north and east of that line.
It roughly coincides with the Y-chromosome geographic pattern, with the northeast containing haplogroups F->K as well as C and D (the latter surviving today only in SEA refugia, such as Japan and Andaman islands), while South Europe, Middle East and South Asia containing E and F->GHIJ lineages and these lineages reasonably coincide with "Basal Eurasians". It seems the separation of these two big groups occurred early, around 55-50 kY ago, and they had very different propensities to travel. The first one (F-K, C, D) relatively quickly settled huge areas of Europe and N and E Asia all the way to Sahul, while the second (E, F->GHIJ) stayed in the larger Middle East and much later developed first Neolithic cultures.
The split between northern and southern dispersal routes around the Himalaya is well substantiated by several lines of genetic evidence. Regarding the Neanderthal mixture what we know from extant genomes is consistent but the comparison shown here is hampered by the very low availability of ancient DNA from SE Asia. The high contribution of mainland China agricultural transition to SE Asia may make more of a difference in the east than today's pattern than the early European farmers did in the west. Great comment!
"Only in Africa were people clearly much lower in Neanderthal heritage, small enough to be undetectable with the first statistical approaches to the question. Today we know that across Africa south of the Sahara Desert, living people have only a trace of Neanderthal genetic ancestry. This pattern is a legacy of thousands of years of migration and mixing between Africa and Eurasia. The level is not very high: a fraction of a percent. Still, the sheer distances involved—some 8,000 km across the continent north-to-south—are a vivid reminder of how gene flow has diffused across groups."
Do we actually know this for certain, that Africans (aside from post-Neolithic, obviously Eurasian admixed East Africans) really do have even trace Neanderthal ancestry? There was a study a year or two ago that suggested reported trace-Neanderthal ancestry in African populations was essentially a mislabeling of a previously undetected signal of modern human ancestry into the ancestral Neanderthal population 100-200k years ago; once that signal was accounted for, the Neanderthal ancestry in modern Africans went away. Not sure if there's been any follow-up on that front, though.
It would seem like by now, that the conventional Neanderthal introgression into proto-Eurasians is well accounted for enough that it shouldn't be so difficult to confirm if Africans really do have some of the Neanderthal ancestry that Eurasians have.
There's quite a lot of research on this now, with multiple approaches giving similar outcomes. I wrote about this in 2020 and the fundamental story has not changed, but the levels have been refined in various ways. The quote you pull from this post is basically a summary of the southern African ancient genomes paper from Mattias Jakobsson's group that was published in December.
The fact that researchers were able to finesse this cline out of these larger samples is great! This seems to imply that Neanderthals preferred higher latitudes and were "cold weather" species. But I'm reminded of Hublin & Roebroeks (2009) who concluded, "their [Neanderthal's] ability to cope with extreme glacial environments seems to have been very limited." They also argued that during cooler times Neanderthals went locally extinct in higher latitudes.
So did Neanderthal genes, in the sapiens gene pool, increase proportionally with latitude because Neanderthals were already further north or did arriving sapiens push Neanderthals to higher latitudes and this increase in proportion is a result of continuous contact along the leading edge of sapiens migration?
I suppose with the differences discussed being less than one percent, it's likely both. I find it hard to believe that Neanderthals had abandoned southern Eurasia.
You ask a good question. The cline tells us about mixture, but not necessarily about the number or density of either population. Finding more Neanderthal ancestry further north may only indicate that modern people had to pass through more Neanderthals to get further north. That would be the genetic drift scenario. But it may be possible that climate made some difference directly through selection. To test that possibility will require looking more closely at the frequencies of particular genes to see if any of them have a higher relationship with environmental variables such as latitude.
John, I’m puzzled by the apparent 20,000-year gap between 70,000 years ago when early modern humans crossed into southwest Asia and 50,000-to-43,000 years ago when that founder population started interbreeding with Neanderthals in the same region.
Did modern people stay in one relatively confined area without expanding or wandering for 20,000 years? Did they simply not encounter Neanderthals, or did they encounter them without interbreeding?
Or did something change in the modern human population or culture in that 20,000-year period which suddenly caused them to surge out of their once-confined area, start expanding their geographic range and continue to do so on an ongoing basis, and start interbreeding with the Neanderthals (and Denisovans?) they encountered?
Thanks for your provocative series.
This is a great question! We're all puzzled a bit by this timeline. Let's start with what we know.
1. There was a bottleneck of this founder population long enough to cut its genetic variation by a substantial degree. This requires a length of time roughly on par with the effective number of individuals in the population. 1000 people would take roughly 1000 generations, or 20,000–30,000 years. This is a ballpark estimate only. A smaller effective size could be a shorter bottleneck and we're talking order of magnitude, not precise estimates.
2. All bottlenecks have a tighter effect on mtDNA and Y chromosome systems than on the autosomal DNA. By our current way of looking at these systems, they seem to indicate a separation from African groups in that time range of roughly 75,000 to 65,000 years ago. The autosomal estimates of this separation time are harder to pin down, largely due to the need to somehow correct for later gene flow after the founders started expanding. Ancient DNA would help a great deal.
3. The release from the bottleneck is better timed, although there is an ongoing mismatch between mtDNA and autosomal estimates for this. That mismatch is giving rise to the disagreement between long and short chronologies for island SE Asia, for example. Short chronology is expansion after 50,000 years, long chronology expansion after 70,000 years. (Long chronology consequently pushes back the bottleneck onset to 90,000+) At the moment the short chronology is supported by genome-wide LD including comparisons that work in ancient genomes; the long chronology supported mainly by mtDNA also including some ancient genomes and certain mutation rate assumptions.
Your question about what the founders were doing while they hung around bottlenecked in North Africa or Southwest Asia is the right one. We have no idea. We can't identify this population archaeologically, and from the genetic point of view it's a ghost.
Southern route proponents would probably say that the founders were taking refuge in Green Arabia at that time, pushed south by the Neanderthal resurgence of Amud-Kebara times. An alternative is that this was a lower Nile valley or coastal Egypt-Libya population that was just waiting for the right release point.
If the short chronology is correct, then it's a good hypothesis that the initial Upper Paleolithic in the Levant represents the early descendants of the expanding founder population. We are talking about 3000 years to get them from the Levant across Anatolia, the Caucasus, and well into the Zagros, and that can be made to fit the chronology of IUP. I'm wary of shoehorning, but the timeline doesn't give me vapors there.
Thank you for this detailed comment!
I believe uniparental and autosomal genomic data are actually fairly consistent with each other. Autosomal results in many recent papers show a strong bottleneck peaking around 70-60 kY ago. Some are actually more specific, e.g. TMRCA for all Eurasians show an abrupt cutoff at 70-75 kY in this picture:
https://substackcdn.com/image/fetch/$s_!28ew!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa2aa6467-2c86-44d2-bedb-2fb55b7d2299_624x601.png
from the 2020 paper "“Insights into human genetic variation and population history from 929 diverse genomes”.
It coincides with the emergence of CT->{CF,DE}->{C,F,D,E} Y-chromosome lines and M and N mtDNA macrohaplogroups. I believe the quick emergence of these major haplogroups ancestral to all Eurasians can be best explained by a split of early humans into small isolated groups during the bottleneck and separate genetic fixation in each of these surviving groups.
Then around 55 kY ago there is an explosion of many new haplogroups and beginning of exit from the bottleneck according to autosomal data. This likely indicate the start of a vigorous demographic and geographic expansion at that time.
To follow up on my earlier question, is there any evidence in the genome or in physical archeology of some kind of physical, cultural or technological advancement during this 20,000-year bottleneck, such as an improved toolkit or signs of enhanced vocal articulation or symbolic behavior, which could have given this founder population a competitive advantage over other modern human populations as well as Neanderthals and Denisovans, to enable their ensuing dispersal and growth, analogous to Neolithic agriculture or (perhaps) the Zamnaya’s spoked wheel, or is their dramatic expansion primarily attributed to opportunities created by climatic change?
This is an astute question that gets to the point of why there was such a difference in population intrinsic growth after the bottleneck. It deserves its own post and hopefully I'll be able to return to the topic soon. My intuition is that the mixture itself is part of the explanation.
Not only the archaic admixture itself, but I also feel like there had to have been some sort of selection on the OOA populations sex chromosomes themselves that led to some sort of fertility boosting effect. I hope some lab really does try and pursue this angle more in the not too distant future.
https://www.johnhawks.net/p/a-shorter-sharper-out-of-africa-story/comment/158071495?utm_source=activity_item
One more conclusion one can make from the Neanderthal admixture geographic distribution - refutation of an old myth of the "rapid coastal travel out of Africa". It was never more than a purely mental construct without any archaeological or genetic evidence, but was largely brought to explain early human presence in SEA and Australia.
This myth (and corresponding maps) still fills nearly all anthropological textbooks and museums.
This just never happened.
If early humans settled the Indian ocean coast first and then expanded inland, the geography of Neanderthal introgression would be very different, of course. But data shows that N. admixture in any ancient indigenous people in SEA (including such extremely old isolates as Andamans) is very similar to other Asians. It is clear that modern humans came to SEA from the north (all the way from Siberia), not along the coast. This is also confirmed by a shockingly small genetic distance between Tianyuan man and modern Andamanese, as well as prevalence of old Y-DNA F->K and C lineages in indigenous SEA and Sahul region.
As for myself I agree with this. I can't entirely render the point of view of the "southern route" proponents, because I don't see the evidence for it. The Andaman Islands DNA evidence is very important. On this point I think they have to lean on the idea of later migrations covering up the first wave. (Or vague hand-waving about glacial sea levels covering up the evidence).
I think the best proxy to the geographic distribution of the Neanderthal admixture is neither latitudinal nor longitudinal, but the "Movius line" - an old archaeological concept, that divided Eurasia roughly in the Southwest and Northeast parts. Basically, migration patterns of early humans in Eurasia were either south of Himalayas and west of the mountain range around the Bangladesh-Burma border, or north and east of that line.
It roughly coincides with the Y-chromosome geographic pattern, with the northeast containing haplogroups F->K as well as C and D (the latter surviving today only in SEA refugia, such as Japan and Andaman islands), while South Europe, Middle East and South Asia containing E and F->GHIJ lineages and these lineages reasonably coincide with "Basal Eurasians". It seems the separation of these two big groups occurred early, around 55-50 kY ago, and they had very different propensities to travel. The first one (F-K, C, D) relatively quickly settled huge areas of Europe and N and E Asia all the way to Sahul, while the second (E, F->GHIJ) stayed in the larger Middle East and much later developed first Neolithic cultures.
The split between northern and southern dispersal routes around the Himalaya is well substantiated by several lines of genetic evidence. Regarding the Neanderthal mixture what we know from extant genomes is consistent but the comparison shown here is hampered by the very low availability of ancient DNA from SE Asia. The high contribution of mainland China agricultural transition to SE Asia may make more of a difference in the east than today's pattern than the early European farmers did in the west. Great comment!
"Only in Africa were people clearly much lower in Neanderthal heritage, small enough to be undetectable with the first statistical approaches to the question. Today we know that across Africa south of the Sahara Desert, living people have only a trace of Neanderthal genetic ancestry. This pattern is a legacy of thousands of years of migration and mixing between Africa and Eurasia. The level is not very high: a fraction of a percent. Still, the sheer distances involved—some 8,000 km across the continent north-to-south—are a vivid reminder of how gene flow has diffused across groups."
Do we actually know this for certain, that Africans (aside from post-Neolithic, obviously Eurasian admixed East Africans) really do have even trace Neanderthal ancestry? There was a study a year or two ago that suggested reported trace-Neanderthal ancestry in African populations was essentially a mislabeling of a previously undetected signal of modern human ancestry into the ancestral Neanderthal population 100-200k years ago; once that signal was accounted for, the Neanderthal ancestry in modern Africans went away. Not sure if there's been any follow-up on that front, though.
It would seem like by now, that the conventional Neanderthal introgression into proto-Eurasians is well accounted for enough that it shouldn't be so difficult to confirm if Africans really do have some of the Neanderthal ancestry that Eurasians have.
There's quite a lot of research on this now, with multiple approaches giving similar outcomes. I wrote about this in 2020 and the fundamental story has not changed, but the levels have been refined in various ways. The quote you pull from this post is basically a summary of the southern African ancient genomes paper from Mattias Jakobsson's group that was published in December.
For more context, my earlier post is here:
https://www.johnhawks.net/p/how-much-neandertal-dna-do-todays-african-peoples-have
The fact that researchers were able to finesse this cline out of these larger samples is great! This seems to imply that Neanderthals preferred higher latitudes and were "cold weather" species. But I'm reminded of Hublin & Roebroeks (2009) who concluded, "their [Neanderthal's] ability to cope with extreme glacial environments seems to have been very limited." They also argued that during cooler times Neanderthals went locally extinct in higher latitudes.
So did Neanderthal genes, in the sapiens gene pool, increase proportionally with latitude because Neanderthals were already further north or did arriving sapiens push Neanderthals to higher latitudes and this increase in proportion is a result of continuous contact along the leading edge of sapiens migration?
I suppose with the differences discussed being less than one percent, it's likely both. I find it hard to believe that Neanderthals had abandoned southern Eurasia.
You ask a good question. The cline tells us about mixture, but not necessarily about the number or density of either population. Finding more Neanderthal ancestry further north may only indicate that modern people had to pass through more Neanderthals to get further north. That would be the genetic drift scenario. But it may be possible that climate made some difference directly through selection. To test that possibility will require looking more closely at the frequencies of particular genes to see if any of them have a higher relationship with environmental variables such as latitude.