July 25, 2012

The long, drawn-out road to us

Timothy Weaver has an intellectually challenging recent article in the Journal of Human Evolution. He proposes that there was no punctuational "rise of modern humans", but rather a long, drawn-out continuous process since our split with our closest cousins, the Neandertals.

The argument for "no bottleneck" leading up to modern humans in Africa was recently made by Sjödin et al. Weaver examines critically three of the arguments of the "recent Out of Africa" model:

  • the shallow coalescence of mtDNA Eve
  • the emergence of anatomically modern traits in Africa c. 200ky ago
  • the idea that these modern traits are quite different from the ones that preceded them -the idea that a new species was born at the time.

With respect to mtDNA, Weaver makes the argument that the fact that mtDNA Eve coalesces to about 200ky ago (actually 177ky according to the latest estimate) does not mean that there was some bottleneck at the time associated with the rise of modern humans; the same coalescence age could occur under quite different scenaria: perhaps there were bottlenecks all the time, and nothing special happened at that time; or, there were no bottlenecks at all: if the effective population size was constant, then mtDNA would still coalesce at some time, depending on what that effective size was.

This of course raises the question: if we know what the human effective population size is, could we estimate when mtDNA ought to coalesce. First, coalescence theory does not provide a hard time for the coalescence, but rather an expected value; coalescence at 177ky is compatible with lots of different effective sizes, and different effective sizes are compatible with a coalescence at 177ky.

(More importantly, and contrary to popular belief and recent commentary: we have absolutely no idea what the human ancestral effective population size is. Figures like 10,000 people are sometimes quoted around, but we must remember where they come from: there is a triangle of doom between the human-chimp divergence date, the effective population size, and the mutation rate, and you need to know two of these to infer the other. Actually, we are beginning to get a hold of the mutation rate -thanks to the ability to sequence full genomes- but we have absolutely no clue when human-chimp divergence actually happened, at least not within a few million years.)


There is a different blow that can be directed to the idea of using mtDNA to infer a "rise of modern humans": if we look at Denisova and Neandertal hominins, they are autosomally about equidistant to us, but Denisova carries an mtDNA lineage that is about half a million years more ancient. If that doesn't stop us from repeating the "recent mtDNA Eve = recent African origin" meme, I don't know what is.

Getting back to the Weaver article, the author argues that the appearance of cranial modernity is expected if we only make an assumption about the narrow-sense heritability of human traits; that is, working backwards from the present, and taking into account drift and mutation, we expect that "modern traits" will start appearing in the anthropological record at the time of the supposed "rise of modern humans". This is simply a consequence of the fact that anthropologists label traits as modern or archaic with respect to extant human variation; so, there is nothing special about the fact that such "modern" traits appear on the record, since they are expected to do so by the mere fact that the people who lived 100-200ky ago are ever-more related to us.

The final aspect of the Weaver article has to do with the supposed punctuation in the appearance of modern humans in Africa. He makes a good point here, that the African record is so fragmentary that we hardly know what people were like before the supposed rise of modern humans. It's tough to argue about the emergence of a new species when you have no good comparative base. I would also add that even after the supposed emergence, "modern" and "ancient" traits co-exist, with no clear overall pattern discernible in the data. If modern humans suddenly arose in Africa and replaced pre-existing African hominins, the evidence for this sudden emergence and replacement is lacking.

Overall, I would say that Weaver makes a good argument against the idea of us being something special in the grand scheme of things. Perhaps we're not mutant world conquerors after all, but rather the latest phase in a long and drawn-out evolution of Homo. It's a less dramatic and more mellow theory about our origins, but one that may very well be true.

Journal of Human Evolution
Volume 63, Issue 1, July 2012, Pages 121–126


Did a discrete event 200,000–100,000 years ago produce modern humans?

Timothy D. Weaver

Abstract

Scenarios for modern human origins are often predicated on the assumption that modern humans arose 200,000–100,000 years ago in Africa. This assumption implies that something ‘special’ happened at this point in time in Africa, such as the speciation that produced Homo sapiens, a severe bottleneck in human population size, or a combination of the two. The common thread is that after the divergence of the modern human and Neandertal evolutionary lineages ∼400,000 years ago, there was another discrete event near in time to the Middle–Late Pleistocene boundary that produced modern humans. Alternatively, modern human origins could have been a lengthy process that lasted from the divergence of the modern human and Neandertal evolutionary lineages to the expansion of modern humans out of Africa, and nothing out of the ordinary happened 200,000–100,000 years ago in Africa.

Three pieces of biological (fossil morphology and DNA sequences) evidence are typically cited in support of discrete event models. First, living human mitochondrial DNA haplotypes coalesce ∼200,000 years ago. Second, fossil specimens that are usually classified as ‘anatomically modern’ seem to appear shortly afterward in the African fossil record. Third, it is argued that these anatomically modern fossils are morphologically quite different from the fossils that preceded them.

Here I use theory from population and quantitative genetics to show that lengthy process models are also consistent with current biological evidence. That this class of models is a viable option has implications for how modern human origins is conceptualized.

Link

6 comments:

terryt said...

"Overall, I would say that Weaver makes a good argument against the idea of us being something special in the grand scheme of things. Perhaps we're not mutant world conquerors after all, but rather the latest phase in a long and drawn-out evolution of Homo. It's a less dramatic and more mellow theory about our origins, but one that may very well be true".

Almost certainly is true.

"modern human origins could have been a lengthy process that lasted from the divergence of the modern human and Neandertal evolutionary lineages to the expansion of modern humans out of Africa, and nothing out of the ordinary happened 200,000–100,000 years ago in Africa".

I have long been very suspicious of any alternative theory.

"He proposes that there was no punctuational "rise of modern humans", but rather a long, drawn-out continuous process since our split with our closest cousins, the Neandertals".

I'd take it even further back, from the time of Homo erectus. And most likely long before that time.

eurologist said...

The H. Li, R. Durbin paper cited by Weaver is fascinating - I don't think I paid attention to it, previously:

http://www.nature.com/nature/journal/v475/n7357/full/nature10231.html

* their calculation results in a population maximum coinciding with the climatic optimum ~150 - 100 kyr ago;

* differentiation of European and Asians from Youruba started 120–100 kyr ago (!)

Rob said...

Great argument

terryt said...

"The H. Li, R. Durbin paper cited by Weaver is fascinating - I don't think I paid attention to it, previously"

You're right.

"All three populations have an elevated effective population size between 60 and 250 kyr ago, possibly due to population substructure. We also infer that the differentiation of genetically modern humans may have started as early as 100–120 kyr ago12, but considerable genetic exchanges may still have occurred until 20–40 kyr ago".

That again argues against the simple, 60 kya. OoA.

Aaron said...

Unknowns:

1) Divergence Time
2) Mutation Rate
3) Population Size
4) Population Growth Rate

and many other unknowns.

I learned early on that to solve for 4 unknowns you need 4 equations.

One of the major problems with genetic time estimates currently is that we estimate three of the unknowns to solve the fourth.

We don't have a crystal ball that can look back into the past to know for certain the values of the other unknowns, and often we are very wrong.

Perhaps the most destructive assumption in genetics is a stable population size, many algorithms depend on that, and if you look at recent human history, that is anything but true.

terryt said...

"One of the major problems with genetic time estimates currently is that we estimate three of the unknowns to solve the fourth".

But it is very interesting to play with the figures even though we can be certain that they are wrong.

"Perhaps the most destructive assumption in genetics is a stable population size, many algorithms depend on that, and if you look at recent human history, that is anything but true".

Yes, but just as destructive is an assumption of steady increase in population. Like most species our numbers have almost certainly fluctuated wildly.