August 3, 2011

Did Homo sapiens outnumber Neanderthals 9 to 1?

That is what a recent paper suggests:

I got a copy and lots of complementary information and interesting debate (in Spanish) at Mundo Neandertal.


Mellars and French deal specifically with the transition between Chatelperronian and Aurignacian cultures in the context of the Franco-Cantabrian region.  It is generally accepted that Chatelperronian was a Neanderthal culture and Aurignacian instead was mad by Homo sapiens, our species.

The study estimates population sizes based on an archaeological record that is assumed to be quite complete (something only available for Europe probably). 

All this brings us to an obligated earlier reference:

Which was previously discussed by me here.
To get us fully in situation, I must mention that Bocquet-Appel estimates the following population for Aurignacian Europe:

average: 4424 
(min.: 1738 - max.: 28,359)

And suggests the following geographical distribution:

Aurignacian population est. (Bocquet-Appel 2005)

Even though the demographic preeminence of the Franco-Cantabrian region in Upper Paleolithic Europe becomes more clear later,  with the Last Glacial Maximum, already in the Aurignacian period (c. 41-28 Ka ago) it was the most populated area of Europe without doubt, holding surely 1/3 of all Europeans (later it becomes at least 2/3). 

This makes all of the estimates by Mellars & French 2011 more than quite relevant: what happened in the Franco-Cantabrian Region affected to most Europeans back then, Neanderthal, Sapiens or whatever else.

Nine Sapiens for each Neanderthal (9:1)

Bocquet-Appel found only minor differences in population density between Aurignacian and Solutrean, i.e. for most of the Upper Paleolithic generally accepted to have been made by Homo sapiens, only noticing a clear (and quite brutal) increase in population with the Late Upper Paleolithic, notably in Magdalenian contexts.

Based on genetic estimates, I thought that there would be roughly the same number of Neanderthals as early Homo sapiens in Europe, that some 5000 people (1800-28,000) was the number that Europe could support in such technological conditions and that it did not matter if these were Neanderthals or Sapiens. 

Fig. 1. Site densities. Left: Chatelperronian, right: Aurignacian

But Mellars and French challenge this idea. They compared the densities of Chatelperronian and Aurignacian sites in a rectangle centered in Perigord (the Paleolithic metropolis of Europe) and found that Aurignacian sites are more common than Chatelperronian, after adjusting for time, some 2.5 times.

Fig. 2A

That means 2.5 more Sapiens settlements than Neanderthal ones. But more: they find that the density of use of the sites is also much greater for the Aurignacian ones, reaching a ratio of 9 to 1.
Fig. 4

Is this for real?

Assuming that the average figure proposed by Bocquet-Appel for the Aurignacian population of Europe (4424) would be correct, then there would be only 491 Neanderthals in all the continent. I do not think this is a credible figure. 

Let's assume then that the correct figure of BA is the maximum (he may well have been too conservative in his estimates, I guess): 28,359 people. This implies 3151 Neanderthals. It is a more credible figure but still a bit too low. 

Remember that the Neanderthal Genome Project estimated 1500-3500 fertile Neanderthal women at any given time (in Europe only), what means at least three times actual people (i.e. 4500-10,500 min.) 

Of course there may be an error here and it is another different way of estimating populations, not directly comparable and much more uncertain than estimates based on factual archaeological data. Still, I have the strong impression that the archaeology-based estimates are too low, that the actual populations are being underestimated. 

Magdalenian harpoons
For example, if the lowest genetic estimates would be correct, then European Neanderthals would be c. 5000. Following Mellars & French, the earliest (Aurignacian) Sapiens population could be then 45,000 (almost double than the maximum figure estimated by Bocquet-Appel), what in turn would make the Late Upper Paleolithic (Magdalenian and Epigravettian) population of Europe (6.5 times that of Aurignacian) roughly of 292,500 people (200,000 or so in the Franco-Cantabrian region). 

It could well be double too. Half a million hunter-gatherers roaming Europe upon the arrival of the first farmers? Does this make any sense? I'm not sure but it's not totally impossible, specially if we consider that sea resources (seals, fish, mollusks and crustaceans, whales maybe) had become very important by then.

Of course, fine tuning of the figures available may also be recommended. Anyone?

Update: comparing with the Aleuts

A very interesting observation is made by Joy in the comments section: the Aleut people, whose technology and way of life can well be compared with that of Late UP or Epipaleolithic  Europeans, numbered 25,000 people upon contact with modern Europeans (now they are only 15,000 however).

This is a quite sizable population for a relatively small and cold territory. I have calculated that the traditional Aleut territory has some 30,000 km², what is just 5.3% of the territory of modern France (552,000 km², excluding overseas dependencies) and just 0.3% of all Europe. Extrapolating only to "France", we'd get some 500,000 people. Sure that "France" is not all coast, as the Aleut territory was but it's neither so far North and cold (the Aleut lands' climate and geography is more comparable to Norway in fact).

In any case, what I find with this comparison is that the figures mentioned above of 300,000 or more people in late UP Europe are perfectly plausible, even maybe a bit shy and low.


  1. Lots to cover here.

    First of all, population sizes in the area could easily have been higher than the genetic estimates. The Aleut people of the Aleutian Islands of Alaska (completely hunter-gatherer) numbered circa 25,000 at the time of European contact. These folks were essentially still living an ice age lifestyle in the 18th century.

    The ratio of H.s/H.n in this paper rings true to me. H.n lived for circa 300,000 years in Europe with no major changes in megafauna. Yet when H.s shows up, H.n and the megafauna melt away.

    My latest hypothesis is that H.n was much more dependent on big game than H.s. When big game dropped below a certain density, H.n would die back to a sustainable level. Thus they stayed in balance with the megafauna for repeated glacial cycles.

    If H.s was better at exploiting a variety of food sources, for example by better fishing or more efficient taking of small game such as hares, or even pursuit of deer to exhaustion (something H.n sprinters were unlikely to be suited for) -- then H.s could keep going even when the big game was rare. Yet H.s would still take big game whenever it was encountered. In this scenario, H.n fades away not so much from warfare, as from loss of their main food source. H.n would have suffered the same way the Plains nations of North America suffered from European hunting of bison.

    Just a thought.

  2. The Aleut people numbering 25,000 is a great reference, because their technology should have been similar (+/-) to that of Magdalenian and Epipaleolithic peoples of Europe, also with emphasis in sea resources. Their area was not too large either, and was quite to the North (comparable with at least North Sea Europe, rather than the FC region). Thanks.

    "Yet when H.s shows up, H.n and the megafauna melt away".

    But not so fast, right? Some megafauna (U. speleoarctos notably) did vanish with the arrival of H. sapiens but others did not. We have evidence of almost all megafauna (mammoth, bison, rhino, horse, auroch, brown bear, wolf...) up to the Magdalenian period. It is only with the end of the Ice Age when some of these animals suffer a lot, even up to extinction in some cases (mammoth notably).

    "... they stayed in balance with the megafauna for repeated glacial cycles".

    One and a half glacial cycle. Neanderthals as such are not much older than our own species, so they lived through the Riss and most of the Würm glacial periods.

    "H.n would have suffered the same way the Plains nations of North America suffered from European hunting of bison".

    I seriously doubt that any amount of Paleolithic humans could have any impact in the numbers of bison. I'm imagining huge packs such as the ones found in North America and of course I do not imagine any of the brutally destructive practices of industrial era white hunters in railroad payroll or anything like that.

    I think that the demise of Neanderthals, even if probably aided by climatic accidents (HE4 cold spell notably) was more of a matter of "who controls and exploits this nice territory: you or me?" And it was "me", i.e. our ancestors by maybe even direct armed conflict at times.

  3. The Aleuts' territory has some 30,000 km² of land, which is 5.3% of the territory of the French Republic (excluding overseas dependences).

    A linear ratio would yield 500,000 people for "France", making the figures I considered at the end of this entry (for all Europe) perfectly plausible and even shy and conservative in fact.

  4. Erratum: U. spelaeus, not "U. speloarctos" as I said above. I'm uncertain about the dates of extinction of the hyena and the saber-toothed tiger as well, but I think that they did not exist by the time of arrival of H. sapiens, what may make H. neanderthalensis suspect of their extinction, in any case. All were carnivores, no one was a typical hunter's prey.

  5. I think you are underestimating the area of Aleut territory.

    Also, while the precise ratio suggested by the study may be in tension with genetic estimates of the size of the Neanderthal population (which, it is worth recalling, are from a very, very small sample of individuals and is likely to have meaningful hard to quantify systemic error given the dififculties involved in the ancient DNA recovery process), the bottom line conclusion - that early UP anatomically modern humans were much more common than Neanderthals given the archaeological evidence, whatever the precise ratio may have been, is well supported.

    Another point to keep in mind when reconciling the two is that the Neanderthal DNA indicates two or three subpopulations, and that archaeological evidence is not evenly distributed geographically. It is not implausible that the populations were closer to parity in some regions and widely unequal in others. (Probably closer to parity in the North and more unbalanced in the F-C area.)

    The hardest issue with the archaeological evidence is the extent to which preservation and recognition is a factor. Distinguishing MTA artifacts from background randomly broken rocks is non-trivial and requires a fair amount of skill relative to the skill needed to identify later technologies as something hominin made, the degregation of older artifact deposits is expected to be greater with time, and it is also entirely plausible that Neanderthals were less heavy tool users than archaic modern humans.

    Parsing out both (1) differential rates of artifact discovery due to identification and degredation over time factors, and (2) different rates of tool production per capita, makes it difficult to make inferences from the archaeology alone.

    However, to the extent that both roughly confirm each other, even if the error margins don't quite overlap, the findings that the populations were somewhere on that order of magnitude are more robust.

  6. "I think you are underestimating the area of Aleut territory."

    The territory is almost exactly the Aleutians East Borough and the Aleutians East Census Area (plus lesser islands in Russian territory, which I ignored for simplicity).

    I simply added up the non-water area of these two Alaskan administrative divisions. With water are it'd be double, so I'm almost sure that they are counting parts of the sea, what is not really valid.

    "It is not implausible that the populations were closer to parity in some regions and widely unequal in others".

    Not implausible but not likely either. You'd have to demonstrate it by means of an archaeo-statistical study like the one done by Mellars and French.

    "Distinguishing MTA artifacts from background randomly broken rocks is non-trivial"...

    It doesn't matter because the paper only makes passing mention of MTA and what they really compare is Chatelperronian with Aurignacian, two early Upper Paleolithic techno-cultures with, seemingly, different authorship.

    "... it is also entirely plausible that Neanderthals were less heavy tool users than archaic modern humans".

    This is an interesting idea but it would only divide by two (or was it 1.8?) the stats.

    We may indeed want to fine tune some data and inferences but it's a task that I cannot perform: it would require another academic study by experts on the matter.


Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).