> On July 11, 1997, the announcement was made in the journal Cell that
> Neanderthal mitochondrial DNA (mtDNA) had been successfully recovered
> and sequenced by Svante Pääbo and his team.25  Of course there were
> statistical differences between Neanderthal DNA and the DNA of modern
> humans.  These differences were used to calculate the evolutionary
> divergence of Neanderthals from a common ancestor to around 550,000 to
> 690,000 years ago.  It is thought that Neanderthals then became
> extinct without contributing mtDNA to the modern human genome.  In
> other words, Neanderthals were just one of many offshoots or splinter
> groups that became extinct but who were not direct links to modern
> humans in our evolutionary branch.

> However, there are just a few problems with this theory.  One problem
> has come to the forefront with movies such as Jurassic Park and with
> the publicity of the O.J. Simpson murder trial where DNA technology
> played a prominent role.  Some problems that were brought more clearly
> to light by these media events is that DNA does not last very long.
> It breaks down fairly rapidly depending on environmental factors.
> Even under the most favorable conditions, many scientists believe that
> DNA cannot remain identifiably intact beyond a few tens of thousands
> of years. 27,28  In light of this fact, how could Neanderthal DNA
> avoid decay over the "120,000 to 150,000" years that is the
> hypothesized age of these bones?  This seems to be quite a significant
> problem.  Of course, the reports of recovering DNA from amber that is
> millions of years old are now being discounted because of this very
> problem.29

> Repeatability seems to be a real issue as well.  Not only is it
> difficult to confirm claims of ancient DNA recovery from amber
> specimens, but mtDNA recovery from Neanderthals has only been done
> three times.  Others have found it very difficult to repeat Pääbo's
> results.  Since the scientific method is based on repeatability, the
> whole issue of Neanderthal mtDNA and its implications comes into
> serious question.  Consider the following comment from Pääbo himself:

> "Preserved Neandertal DNA is likely to be rare, and the DNA in the
> type specimen [the 1856 Neander Valley Neandertal fossil] may result
> from its unique preservation conditions. … Most Neandertal
> specimens are therefore unlikely to contain amplifiable DNA. …"
> 30

> Despite these problems, Pääbo et al. seem to have overcome them in
> this particular case and presented their conclusions in the following
> statement:

> "The Neandertal sequence was compared to 994 contemporary human
> mitochondrial lineages, i.e., distinct sequences occurring in one or
> more individuals, found in 478 Africans, 510 Europeans, 494 Asians,
> 167 Native Americans and 20 individuals from Australia and Oceania.
> Whereas these modern human sequences differ among themselves by an
> average of 8.0 ± 3.l (range 1–24) substitutions, the difference
> between the humans and the Neandertal sequence is 27.2 ± 2.2 (range
> 22–36) substitutions. Thus, the largest difference observed
> between any two human sequences was two substitutions larger than the
> smallest difference between a human and the Neandertal." 25

> The conclusions drawn were as follows:

> "When the comparison was extended to 16 common chimpanzee lineages,
> the number of positions in common among the human and chimpanzee
> sequences was reduced to 333. This reduced the number of human
> lineages to 986. The average number of differences among humans is 8.0
> ± 3.0 (range 1–24), that between humans and the Neandertal, 25.6
> ± 2.2 (range 20–34), and that between humans and chimpanzees,
> 55.0 ± 3.0 (range 46–67). Thus, the average number of mtDNA
> sequence differences between modern humans and the Neandertal is about
> three times that among humans, but about half of that between modern
> humans and modern chimpanzees.
>     To estimate the time when the most recent ancestral sequence
> common to the Neandertal and modern human mtDNA sequences existed, we
> used an estimated divergence date between humans and chimpanzees of
> 4–5 million years ago and corrected the observed sequence
> differences for multiple substitutions at the same nucleotide site.
> This yielded a date of 550,000 to 690,000 years before present for the
> divergence of the Neandertal mtDNA and contemporary human mtDNAs. When
> the age of the modern human mtDNA ancestor is estimated using the same
> procedure, a date of 120,000 to 150,000 years is obtained, in
> agreement with previous estimates. Although these dates rely on the
> calibration point of the chimpanzee-human divergence and have errors
> of unknown magnitude associated with them, they indicate that the age
> of the common ancestor of the Neandertal sequence and modern human
> sequences is about four times greater than that of the common ancestor
> of modern human mtDNAs.
>      The Neandertal mtDNA sequence thus supports a scenario in which
> modern humans arose recently in Africa as a distinct species and
> replaced Neandertals with little or no interbreeding." 25

> If mtDNA was in fact isolated from the Neanderthal bones, these
> conclusions might seem reasonable until one considers a few more
> facts.  The Cell article itself noted that the range of sequence
> differences for modern human mtDNA goes from 1 to 24 with an average
> of 8 substitutions.  The mtDNA sequence differences between modern
> humans and the single Neanderthal fossil range from 22 to 36
> substitutions, with the average being 27.  In other words, the two
> most different humans analyzed in this study, as far as mtDNA
> substitutions are concerned, are different by 24 substitutions.  The
> closest that any human in this study was to the single specimen of
> Neanderthal mtDNA was 22 substitutions.  This means that there are
> some people living today that are closer to Neanderthals in their
> mtDNA sequencing than they are to some other modern human beings.
> Someone might be found to be only 22 substitutions away from our
> Neanderthal, but 24 substitutions away from his own next-door
> neighbor.  Interesting isn't it?  If Neanderthals are classed as
> separate species because of these differences, which one of our modern
> human volunteers should be classify as a separate species?  Perhaps
> the one who had only 22 substitutions different from the Neanderthal?
> Or, maybe his neighbor who had 24 substitutions away from him?

> There have been attempts by popular scientists to describe exactly how
> this mtDNA evidence turns Neanderthals into separate species.  Some
> describe it as a group of early Homo sapiens huddled around a fire.
> Some are shoulder to shoulder while others, on the other side of the
> fire might be several feet away... maybe even 24 feet away.  However,
> the average distance that any one person is from any other person is
> just 8 feet.  Now we notice a dark Neanderthal figure  in the shadows
> far from the fire.  He averages 27 feet away from any given Homo
> sapien huddled around the fire.  Obviously therefore, he is an
> outsider, a different species all together.

> However closer inspection  reveals that some of those huddled around
> the fire are closer to the Neanderthal than they are to certain others
> that are also huddled around the fire.  Does that make them more
> closely related to Neanderthals, who belong to a completely different
> species, than to certain members of their own species?  This sounds
> rather silly does it not?  And yet, this is what must be the obvious
> conclusion.  For example, what if we started with the Neanderthal
> specimen and then picked a person at random out of a crowd.  We might
> get someone who is different by 24 substitutions from our Neanderthal
> specimen.  Now, we pick someone else out of the crowd who just so
> happens to also be different by 24 substitutions from our Neanderthal
> specimen and by 24 substitutions from our first human volunteer.
> Which one is the new species?  They are all equidistant from each
> other.  In order to visualize the problem, draw three dots on a piece
> of paper, one for each of our two volunteers and the third dot for our
> Neanderthal "volunteer."  Make sure to draw each one on the paper
> separated by 24 units of measure from each of the other two dots.
> Now, pick the one that is the new species and the two that belong to
> the same species.  Maybe there are three separate species represented
> here?  However, all one would have to do to disprove this notion is
> get two of the volunteers to "produce offspring" so to speak.  If that
> happened, the entire notion that a separation of 20 or so
> substitutions makes for a new species, would have to be... well...
> revised somewhat.

> Why this problem has not been more publicly recognized seems rather
> strange.  I am sure that I am not the first one to wonder about this.
> And yet, popular scientist seem not even to be aware that there is
> this problem.  It seems that the statistical averages of 8 and 27 are
> so different that this squelches any suggestion that there might be a
> problem.  Using this average difference as a basis for their
> conclusions,  Kahn and Gibbons wrote in the journal Science that these
> averages put Neanderthal out of the statistical range of modern human
> variation.31

> This statement is clearly misguided because not only are there humans
> living today with wider separations between them than our Neanderthal
> friend, but it is a statistical error or pitfall to compare many
> different entities with just one entity.  In other words, we do not
> know what the Neanderthal mtDNA average is if there is just one
> specimen.  How then can we know if this one Neanderthal was not a
> statistical outlier?  How do we know that if we but had more
> Neanderthal samples that the average would not be closer to that of
> modern humans?

> As it turns out, since the first sequence was obtained by Pääbo and
> his team, there have been two more Neanderthals found who's mtDNA was
> intact enough to sequence.  The second sequence was done in 1999 on a
> baby discovered in Mesmaiskaya Cave in south-western Russia.  This
> Neanderthal baby is thought to have died 29,000 years ago.  The
> sequence of this baby differed from the first sequence by 12
> substitutions.  The average number of substitutions between the second
> Neanderthal mtDNA ("Baby M" for short) and a given human is 22 as
> compared to 27 from the first Neanderthal sequence.46,47  In other
> words, Baby M was "closer to the Homo sapien fire" than the first
> shadowy Neanderthal.  This means that some living humans might be even
> closer to this second Neanderthal than they are to other living humans
> by quite a fair margin.  Unfortunately however, no figures for the
> minimum, average, and maximum distances between the second Neanderthal
> and modern humans was provided.

> The third Neanderthal who's mtDNA was successfully sequenced was found
> in a cave at Vindija, Croatia.  In 2000, scientists announced the
> mtDNA sequencing of this third Neanderthal specimen.  This new
> sequence fell within a 3.75% cluster of the first two sequences.48
> Modern humans cluster at around 3.5%.  This is a rather narrow level
> of diversity when one compares these clusters to chimps (15%) and
> gorillas (19%).  Various human ethnic groups also have rather narrow
> ranges of diversity in their mtDNA sequencing.  Of course, the problem
> still remains that some humans from certain of these ethic groups are
> more closely "related" to Neanderthals than they are to certain other
> living humans from other groups.  The question remains as to who
> should be classed as a separate species?

> Maryellen Ruvolo (Harvard University) points out that the genetic
> variation between the modern human and Neanderthal sequences is within
> the range of other single species of primates. She goes on to say:
> "… there isn’t a yardstick for genetic difference upon
> which you can define a species." 31

> Further confusion comes from the comments in the Cell article that
> seem to indicate that Neanderthals are more closely related to the
> ancestral "chimpanzee" than modern humans are.  This might not have
> been the actual intention of the authors, but one could easily get
> confused by the wording of the article.  The fact of the matter is
> that the single specimen of Neanderthal mtDNA was actually farther
> away from chimp mtDNA than humans are from chimp mtDNA substitutions.
> Clearly then, Neanderthal DNA is no closer "related" to chimp DNA than
> human DNA is. 32

> Also, the idea that mtDNA mutations can be used as a molecular clock
> have been recently called into question by the journal Science.  As it
> turns out, former ideas about the timing of this clock might be in
> error by as much as "20-fold."  The famous "Mitochondrial Eve" once
> thought to be around 100,000 to 200,000 years old, might now have to
> be revised to as young as "6,000" years old. 33,34

> D. Melnick and G. Hoelzer (Columbia University) tested the assumptions
> of mtDNA based phylogenic relationships and concluded:

> "Our results suggest serious problems with use of mtDNA to estimate
> 'true' population genetic structure, to date cladogenic events, and in
> some cases, to construct phylogenies." 35

> Jonathan Marks (Yale University) declared mtDNA determined
> relationships to be highly biased:

> "Most analysis of mitochondrial DNA are so equivocal as to render a
> clear solution impossible, the preferred phylogeny relying critically
> on the choice of outgroup and clustering technique." 36

> Given all of these findings, what seems most reasonable?  Are
> Neanderthals anything but human?  It seems like they fall well within
> human ethnic variation.  How then can we say that, based on such
> variations in mtDNA sequences that Neanderthals belong to a different
> group or species than Homo sapiens?