Out of Africa: mtDNA doesn't tell the whole story

arlin at ac.dal.ca arlin at ac.dal.ca
Fri Nov 22 01:23:29 EST 1991

It seems that correspondents have been assuming since this discussion
started that the question of the geographic origin of _homo sapiens_
is equivalent to the question of the geographic origin of human mtDNA.
This assumption is false.  The mtDNA origin question is distinct from
the human gene pool origin question, which is itself distinct from the
question of _homo sapiens_ origins.  Here's why:

1. The "Eve" concept:

Since some females have no daughters, some mtDNA lineages fail to be
represented in the next generation.  Starting from an arbitrarily
chosen generation T, with N females representing N lineages, the
number of lineages represented in generation T + 1 will be less than
N, and will continue to decrease inexorably, until N = 1.  Conversely,
at any point in time we can look back many generations to the single
female ancestor of all extant mtDNAs.

2.  The succession of "Eves" in any population:

The mitochondrial "Eve", from whose lineage all present-day mtDNA
lineages arose, is postulated to have lived ca. 200,000 years ago.
This female does not represent a unique phenomenon in hominid history:
there were other "Eves" before her, and there will be yet others in
the future. For instance, 200,000 years ago, the mtDNA in the human
population could have been traced back to an "Eve" that existed much
earlier (perhaps this earlier "Eve" was not _homo sapiens_, but _homo
erectus_).  The lineage of human "Eves" will also continue into the
future. Suppose that one million years into the future another survey
of human mtDNA is carried out.  The "Eve" for this future population
may be found to have existed 300,000 years from now, and to have lived
on Mars.

3.  The limits of inferences from mtDNA:

The above paragraph should make perfectly clear that it takes
important assumptions to go from statements about mtDNA "Eves" to
statements about the origin of the human species.  In the example
given, humans of the future could be correct in stating that the
origin of human mtDNA was a woman who lived on Mars in 300,000 A.D.
They would be quite wrong in concluding from this that a) all humans
contributing to the gene pool lived on Mars; and b) the human species
originated on Mars.

The point is very simple, really.  Whenever one looks at mtDNA of a
population, one will be able to make a tree relating all of the types.
Whenever there is reason to believe that the mtDNA types are
non-recombining and maternally inherited (these assumptions are made
for the case of humans), then one can suppose that the root of this
tree represents a single historically real female.  It is inevitable
under these conditions, regardless of the history of the species, that
a single female will be thus identified.  It is likewise inevitable,
no matter how the population evolves, that this single female will
have occupied a geographic location.  This will be true even if the
population is geographically widespread.  That is, even if the
population is spread out over seven continents and is panmictic (all
individuals interbreed, and there is no real geographic "origin" of
the gene pool), each mtDNA "Eve" for a time period will still have
belonged to only a single region.

4.  The necessity of other genetic data:

It is more likely than not that the human *mtDNA* gene pool can be
traced to a single female who lived in Africa (as opposed to Eurasia)
ca. 200,000 years ago, though there are considerable uncertainties
associated with this conclusion.  Even if the conclusion about *mtDNA*
were rock-solid (which it isn't!), we would be on shakey ground in
drawing inferences about the origin of the entire gene pool from the
mtDNA data ALONE.  That is, it is conceivable that the Y-chromosomal
"Adam" (who can be discovered from non-recombining
paternally-inherited sequences) will be found to have lived 300,000
years ago in Asia.

However, if evidence for African origins is found for Y-chromosomal
and other nuclear loci, we will be justified in making comments about
the gene pool as a whole.  If not, extrapolations from mtDNA to the
entire gene pool are out of order.  Finally, we will be justified in
making statements about the geographic origin of the *species* (not
just mtDNA or gene pool), if a combination of genetic and
paleontological evidence suggests that the origin of the gene-pool
predates the origin of the chronospecies _h. sapiens_.

Arlin Stoltzfus

Arlin at ac.dal.ca

More information about the Mol-evol mailing list