Re: What determines if two species can produce a viable fertile offspring?

Hi Michael,

  As a scientist, I loved the early episodes of the X-files because many of
the storylines either used mildly plausible science or simply bypassed the
science entirely (which is entirely acceptable for ghost stories and the
like).  As the series matured it became harder and harder for me to suspend
disbelief, as the writers tried to make their storylines more plausible to
a naive audience by incorporating ever greater amounts of pseudoscience. 
Nonetheless, one of the things I really enjoyed about the show was fielding
questions from kids about the accuracy of the science.

  Which brings us to your question.  What prevents species from
interbreeding?  In order to answer this, I need to digress briefly into the
definition of a species.  Technically, a species is a group of genetically
similar individuals capable of interbreeding and producing fertile
offspring.  The key here is "fertile": donkeys and horses can breed to
produce viable offspring (mules) but mules are not fertile.  

  As an aside, there is some wiggle room in defining a species: birders
tend to define species in terms of populations of individuals which
naturally interbreed to produce fertile offspring.  Thus, species can be
defined as genetically isolated groups that might produce a fertile
offspring but do not due to different mating behaviors.  The subtleties of
this debate tend to be lost on normal people, but for hard core birders
working on their life list, there is something of a religious war between
"lumpers" who want to merge similar species and "splitters" who want to
elevate subspecies to full species status.  The lumpers are currently
winning, but the pendulum will probably start to swing back in the future.

  Returning to the question at hand, first let's examine factors that can
affect fertile offspring.  Chromosome count is critical for fertility:
humans and other mammals are diploid, so when our germ line goes through
meiosis we produce gametes with half the normal chromosome complement. 
That is, the normal chromosome count in humans is 23 pairs of chromosomes,
but our gametes carry one and only one copy of each.  When the sperm
fertilizes the egg, the normal complement of 23 pairs is reconstituted. 
However, if an error occurs in meiosis, and one of the gametes gets an
extra copy of almost any chromosome, the fertilized zygote will not develop
normally, and will spontaneously abort.  The only exceptions to this rule
are chromosome 21 and the sex chromosomes (X and Y).  Zygotes carrying
three copies of 21 will develop nearly normally, into kids with Down
Syndrome.  The sex chromosomes are special cases, and result in even less
dramatic syndromes.  My point is, it is critical that each gamete carry
exactly one copy of the genome for normal development. 

  Returning to our Mule, the Horse has 32 pairs of chromosomes, and the
donkey has 31 pairs.  Thus, the mule carries 63 chromosomes. 
Developmentally, horses and donkeys are similar enough that the mule
survives.  However, the mule has problems when the time comes to segregate
the chromosomes into gametes: the uneven chromosome count means that there
will always be one excess chromosome at meiosis.  The unbalanced chromosome
count is just the beginning of the problem: when a pair of species diverge,
their chromosomes tend to rearrange over time.  That is, large chunks of
DNA containing many genes are swapped between chromosomes over evolutionary
time scales.  For example, most of the genes on Human chromosome 21 can be
found on Mouse chromosome 16, but several large chunks of human 21 are
homologous to Mouse 10 and mouse 17.  Thus, when the time comes for a mule
to segregate the 63 chromosomes into gametes, it not only needs to generate
a gamete with 31 or 32 chromosomes, but it needs to get all 32 horse
chromosomes or all 31 donkey chromosomes into a single gamete.  This would
be akin to my passing on to my son only the chromosome set I received from
my father: possible, but very very unlikely.

  So chromosome count (and structure) are major determinants of whether
interspecies offspring are fertile.  Now, as I mentioned earlier, the
barriers to producing viable offspring are not as clear cut.  The major
factor I'm aware of is basic body plan.  For example, just as horses and
donkeys can prodice viable, sterile offspring, so can horses and Zebras. 
For that matter, lions and tigers can produce viable, sterile offspring in
captivity (sometimes known as "ligers).  I'm really not sure how extreme
the difference in body plan can be to produce viable offspring, but I'd
guess that many if not most within genus hybrids are viable.

  Now, to return to more familiar territory, what about humans?  Where
humans have 23 pairs of chromosomes, the rest of the great apes (chimp,
gorilla and Orangutan) carry 24 pairs.  I want to make this absolutely
clear: I do not condone or suggest any experiment to determine whether a
human/great ape hybrid is viable, because of the tremendous ethical and
legal issues at hand, but I'd have to say it's not obviously biologically
impossible.  However, if a human/great ape hybrid were viable, I can almost
guarantee that it would be sterile.

  As for human/alien hybrids, the aliens would probably have to be as
closely related to us as we are to the other great apes.  Thus, they'd need
to be less than 5 million years diverged from us, or thereabouts.  On the
other hand, if the body plan similarity between human and alien was a
result of convergent evolution to a similar body plan, I would say the
chances of a viable hybrid would be roughly the same as the chance of a
human/oak tree hybrid. 

  Chris

Horse and Donkey details: http://www.geocities.com/moredonkeys/equinebasics.html

Human chromosome 21 vs. the mouse genome: http://www.ncbi.nlm.nih.gov/Homology/view.cgi?map=ncbi_mgd&chr=21&tax_id=9606

Ligers: http://www.sierrasafarizoo.com/animals/liger.htm