Content created: 2014-11-25
Content revised: 2018-08-26
File last modified:
Startlingly, it seems to be the case that every human alive today is descended from the same woman, who lived somewhere in or near Ethiopia, probably not more than 200,000 years ago. It also seems likely that those of her descendents who left Africa, perhaps as recently as 90,000 years ago, displaced all of the earlier human populations of Europe and Asia. This shared female ancestor is popularly named "Eve," and because her existence is postulated based on the evidence of mitochondrial DNA, she is usually called “Mitochondrial Eve.”
It is easy to misunderstand the “Mitochondrial Eve” model based on DNA analysis, and the topic is made more complicated by its relevance to the (earlier) "Out of Africa" (OA) and "Multiregional Evolution" (ME) hypotheses. This essay aims to provide the greatest possible clarity.
When hominid fossils were very few, the fact that older ones generally looked less like us and newer ones looked more like us led to the view that the various human forms distributed in small pockets around the world evolved more or less independently over hundreds of thousands of years into the same animal: us. This was the view, broadly, of Franz Weidenreich, the original interpreter of Peking Man.
Unfortunately, if this were true, it would be the only case of such a perfectly parallel evolution known for any plant or animal, and no serious biologist entertains this view at this time, to my knowledge.
Because the view postulates parallel evolution, producing diagrams with parallel lines, like the candles in a candelabra , it is sometimes mockingly called the "Candelabra Hypothesis" or "Candelabra Model."
What is the alternative? The following two fossil-based models were commonly proposed:
Return to top.
The "Out of Africa" or "OA" Hypothesis pre-dated the work with mitochondrial DNA. Based on skeletal evidence alone, it argued that modern human populations in Eurasia had to be the products of a comparatively recent (post-Homo erectus) migration out of Africa, and could not have evolved from earlier Eurasian fossil populations. The work was done primarily with European and Near Eastern Neanderthal specimens, and began by noticing that more recent European Neanderthals looked less like modern Europeans than older ones did. If Neanderthals had been ancestral to modern Eurasians, they should have resembled them more and more over time, not less and less.
The "OA" hypothesis also argued that there had not been sufficient time or sufficient changes in selective pressure since such a migration for subsequent speciation to occur, which would explain why all modern humans were the same species.
This solved the problem of postulating implausible “parallel evolution” —there wasn’t any— and it also explained why we did not find convincing “missing links” between earlier Eurasian populations, like the Homo erectus “Java Man,” and later ones. The older populations had died out and the later populations were descended from a later wave of migration from Africa.
Return to top.
Some scholars —a minority— argued that speciation was a slow enough process that Afroeurasia (Africa, Europe, and Asia) had been home to a single great proto-human hominid population from the time of the first expansion of hominids from Africa to Eurasia, probably more than 1.5 million years ago. Gene flow might not have been enough to sustain total homogeneity, but it also was not so restricted as to lead to completely independent local species. The "ME" model therefore held that hominid forms evolved, not in parallel (as the "Candelabra Model" proposed), but as a single, very diverse species, probably with some isolated sub-populations becoming extinct here and there along the way. (I tend to think of this as the "whoopee model": when populations are in contact with each other, there is … um … gene flow.) This was in fact probably closer to Weidenreich's view than the strict parallel evolution he is sometimes accused of.
This model explained why there were similar forms across the world at about the same time. It also explained why some distinctive physical types were localized, although they were not different enough to represent different species.
The few remaining scholars who subscribe to this seemingly very reasonable view today are usually referred to as “multiregionalists.” The primary reason why the view has been abandoned by most specialists is that the very convincing evidence now available from DNA research simply does not support it.
Return to top.
Continuing research on mitochondrial DNA (and later on the Y chromosome) has brought to light evidence that is relevant to these hypotheses and to transcending them. The research, briefly, goes like this:
Return to top.
Caution: This does not make Mitochondrial Eve the first woman, or the first human, or the first member of a new species.
Further Caution: This does not mean that other women alive when Eve was do not have descendants today; they simply do not have living descendants who are descended only through female links.
Yet Further Caution: If a person were to be discovered whose mtDNA showed a pattern of mutations of greater time depth, then the status of Mitochondrial Eve would be reassigned to the most recent female ancestor shared by both that person and the person we now call Mitochondrial Eve. In other words, although M.E. was necessarily a real person, she is also a logical necessity (just as your mother's mother’s mother is).
Study after study, both of mtDNA and of the Y chromosome (which is transmitted only through and to males), suggests that the basic logic of the Eve hypothesis is probably correct, even down to its surprisingly recent date.
Obviously, Mitochondrial Eve presents significant support for the old, Neanderthal-fossil-based "OA" hypothesis, and reduces the credibility of the old "ME" hypothesis.
In theory, one might accept the mtDNA analysis underlying Eve and still subscribe to the ME hypothesis if one could convincingly demonstrate that there had been a huge error in calculating the rate or the constancy of mutation. If that were the case, then Eve could in fact have lived much earlier and been a predecessor to Homo erectus, say. Current evidence seems to be against this, however.
Return to top.
Mitochondrial DNA study presents powerful evidence and convincing reasoning favoring a single human female ancestor (traced entirely through female links) for all living humans and also providing an approximate date for her. In the absence of modern humans who are not descended from Mitochondrial Eve, we must hypothesize a complete extinction of all other Homo forms in both Africa and Eurasia. Given the recent date of Eve, most of the fossils formerly thought to belong to populations ancestral to modern humans simply belonged to lines of descent without modern issue, at least as traced only through female links.
The date, based on assumptions about the statistical regularity of genetic mutation in DNA, is therefore where the current action is.
Remember that all African populations were also replaced by the "children of Eve." Despite the lingering application of the older name "Out of Africa" and the relevance of Eve in explaining the population of Eurasia, the claims made for Eve include ALL modern humans, not merely those outside Africa.
Return to top.
It is important to remember that we all carry the DNA of other women from Eve's era too. What is special about Eve is only mitochondrial DNA, which is transmitted only through women. Eve, like other women, had descent lines that included men, but her mtDNA came to an abrupt halt at any male link. A woman’s nuclear DNA (as opposed to her mitochondrial DNA) is transmitted through both her daughters and her sons. But in the simplified research world of mtDNA, having a male child, like having no child at all, represents the end of a line. Thus a woman can have millions of descendants after a few hundred generations, and still pass on no mtDNA at all.
Recent statistical models have made much of the fact that we have two (2^1) parents, four (2^2) grandparents, eight (2^3) great grandparents, sixteen (2^4) great, great grandparents, and so forth. Eventually this exponential increase as one moves up one’s genealogy produces a set of same-generation ancestors several times the total population of the world at that era. This can be true because some ancestors occur at more than one place in our genealogies, that is, they are related to us in more than one way, contributing their genes to us more than once. The further back one goes, the more times each individual must be "used" in the genealogy. If one adds the fact that some people have no offspring (and hence are not ancestors and sometimes remove blocks of ancestors above them from the family tree as well), the numbers become more extreme yet.
Pursuing this logic, it is possible to estimate the probability that everyone in the world today is related, say, to Charlemagne, or for that matter to Cleopatra, or to Confucius. The probability is overwhelming that everyone reading this is related to all three of them as long as we consider both male and female links.
The logical simplicity of mtDNA is obviously a great boon to researchers. But the fact that we are all descended through all-female links from a single woman (or through all male links from a single man — the "Y-Guy" of Y-chromosome research) hardly means each of us is not also descended through mixed links from virtually everybody else in the distant past with modern living descendants. We are. But that analysis will probably take several more centuries!
Return to top.
The Y chromosome descends down a male line the same way mtDNA does down a female line, and it is reasonable to hypothesize a single man who would logically have to be the most recent male ancestor shared by all living humans the same way Mitochondrial Eve is on the female side. (He has tentatively been named “Y-Guy.”) Unfortunately, there is only one Y chromosome per cell, but a lot of mtDNA, so the Y-Guy research is more difficult. However some Y-chromosome researchers have produced provocative findings.
For example, geneticist Chris Tyler-Smith from the Welcome Trust Sanger Institute in Britain has concluded that about 15 million people now living in Mongolia and north China share a common ancestor, who seems to have been a named person: Giocangga, the father's father of Nurhaci (1559-1626), the founder of what was to become the Qing or Manchu Chinese dynasty (1644-1911). (Simply Google "Giocangga" for some interesting details of this controversial assertion.)
In a separate study, Tyler-Smith argued that about 16 million men in Central Asia were descended from none other than Chinggis Khan (1167?-1227), founder of the vast Mongol empire (and in China the Yuan dynasty, 1280-1367).
On the mtDNA front, Oxford researchers recently concluded that about 97% of modern Europeans are descendents of seven women who lived sometime between ten thousand and forty-five thousand years back.
These are heady findings, and by no means uncontroversial. However as DNA work continues, driven especially by its utility in everything from medicine to forensics to paternity suits, it is clear that we will see new hypotheses appear about human evolution as well. The study of human paleontology will never again be limited to the study of "old bones."
Return to top.
It has been a standard rule of thumb —the kind you learn early on in high school biology— that species cannot usually interbreed, and, if they do, they do not produce fertile offspring. (A horse and a donkey can interbreed, and the result is a mule. But mules are sterile.)
The concept of evolution has always raised an awkward question: When one biological form evolves until it gives rise to another, where does the change occur? Can it really be the case that one generation is not inter-fertile with the next? Can two populations become unable to mate because they have lived apart for some number of generations? How many generations? Under what circumstances? How much variation is there is this process? Shouldn’t we imagine a kind of gradualism, in which the success of mating decreases over a long period of time? If so, shouldn’t we expect intermediate individuals to turn up? We usually don’t.
For example, if Neanderthals and Homo sapiens are different species overlapping in time —the majority view— then, since we are Homo sapiens, we should not have any Neanderthal genes. On the other hand, if we and Neanderthals are a single species (Homo sapiens neanderthalensis and Homo sapiens sapiens), then fertile mating should have worked fine. But given the range of variation within a species, and the small number of fossils available for classification, researchers working only from the anatomy of fossil forms could not agree whether we and our Neanderthal cousins belonged to one species or two.
Recent DNA research has shown that in fact species boundaries are more fluid than our rule of thumb suggested. The Neanderthal component of the modern human genome is extremely small, but it seems undeniable. Although modern individuals have little Neanderthal DNA, about 20% of the known Neanderthal genome turns up somewhere in modern populations.
Denisovans. Most interestingly, recent research has turned up very human-like materials from a cave called “Denisova.” in the Altai mountains, a range running through south-central Siberia and northern Mongolia. (To find it on a map, find the point where Kazakhstan and Mongolia touch each other, and move northwest slightly.)
Most specimens have been very tentatively dated between 150,000 years ago and 30,000 years ago. (Although the first date will no doubt eventually be pushed back as more evidence emerges, extremes are probably wrong.) The specimens are too fragmentary to provide much anatomical information, but have been successfully mined for DNA evidence. The population is called “Denisovans,” after their cave.
The remarkable thing about Denisovans is that some earlier DNA mutations turn up in some later Neanderthal specimens. One researcher described Denisovans and Neanderthals as “close cousins.” And a few Denisovan mutations turn up in modern populations as well, especially in eastern Eurasia and, to the southeast, in the islands of Melanesia, including Papua New Guinea. (One estimate is that 3% of the genes of people living today in Papua New Guinea come from Denisovans). But no distinctively Denisovan DNA traces turn up in modern African populations.
Because of the small number of specimens, it is hard to develop stable dates for Denisovans. Furthermore, other relatively recent finds have extended our date-range for European Neanderthals. Indeed it may be more accurate to think of Neanderthals as a Denisovan population that got stranded in ice-age Europe and developed a somewhat separate set of adaptations to that harsh environment.
On the other hand, Neanderthal DNA has been shown to occur in Native American populations, believed to descend from migrants from northeastern Eurasia, suggesting that Neanderthals were more mobile than formerly supposed. And one bone from a site in Spain, formerly thought to be Neanderthal, has DNA that is more Denisovan than Neanderthal. But what would a Denisovan be doing in Spain, some 4000 miles from home? (Home? Denisova Cave was the site of their discovery; there is no special reason to assume it was the center of their distribution.)
A find announced in the prestigious journal Nature on August 22, 2018, was a tiny limb fragment from a cave in southern Siberia, not far from the Mongolian border. Based on its DNA, the fragment, too small to be sure whether it came from an arm or a leg, was tentatively interpreted as from a girl of about 13 or a little over, with a Denisovan father and Neanderthal mother, probably living about 90,000 years ago. Not all researchers agree in detail; for example, an alternative interpretation might be that the girl came from a population with a roughly equal mix of Denisovan and Neanderthal genes. But it has become clear that Neanderthals and Denisovans were inter-fertile, at least some of the time, even though the two species retained some genetically (and probably anatomically) distinctive characteristics.
Bottom line (for now). As of September, 2018, the best guess is that Denisovans can, without too much distortion, be thought of as a kind of central- and eastern-Eurasian Neanderthaloid population. If so, further DNA research may even reveal a connection between them and coeval Chinese fossils such as Maba whose relationship to the western Eurasian fossil record has remained obscure based on anatomical studies alone.
Return to top.