Q: "I thought inbreeding always caused problems.  Couldn't it have caused
the problems in ferrets we see like adrenals?"
 
A: Ever notice in the "Back to the Future" trilogy that when Marty goes
forward in time, his son and daughter look just like him, and when he
goes back to the old west and sees his great grandfather, the two look
exactly alike, and his mother looks just like his great grandmother?
I've always suspected THAT was the result of the yucky kind of
inbreeding!  Zip your McFly, will ya?
 
I don't think adrenal disease in ferrets is due to inbreeding; as I've
said before,I think it is a species trait.  Nonetheless, there is no
doubt inbreeding can cause problems, but the same kind of problems can
occur regardless of the genealogy of the two breeding lines.  Breeding
domesticated animals is a lot like baseball; you've got your bats, your
balls, and you have statistics that rule the show.  Like a spitball
increasing the chance a batter won't get a hit, inbreeding increases
the chances (the risks) that recessive traits will become problems, just
like they increase the chance a preferred trait will show up.
 
For this discussion, there are three basic kinds of inheritance in
genetics: dominant, incomplete, and recessive (it is somewhat more
complex than this, but lets keep the discussion simple).  If a problem
gene is dominant, you only need one copy of the gene for it to be
expressed, but you know it is there because you can see it, so if you
are ethical or intelligent you don't breed the line and the problem ends
there.  If the trait is incomplete, it means different forms of the gene
are not dominant over the other ones, and the two genes act together in
governing the trait.  In this case, while you might see a difference, it
may not be bad for the ferret, which can fool you into thinking both of
the genes are benign.  The quandary is that one gene may be faulty, but
the other gene fills in, hiding the trouble.  The crisis occurs when you
have either two copies of a bad gene that is incompletely dominant over
other genes, or when you have two bad genes that are recessive.  In
those cases, the genes can express bad things--not good, not good at all.
 
If you recall your high school biology, you might recall a guy named
Gregor Mendel, an Austrian monk who could hold his peas for years.
Mendel was the first person to work out the statistical ratios for basic
genetic traits, first publishing his results in 1866.  What Mendel
discovered was that genetic traits are displayed in predictable
mathematical ratios, which means they can be statistically predicted.
For example, suppose you had 4 unrelated ferrets: two pure sables
(dominant trait, SS) and two albinos (recessive trait, aa).  If you bred
the sables to the albinos, all the offspring would be sables, but each
one would carry the hidden recessive albino trait (Sa).  If you bred
one of the offspring from the first group to one of the offspring from
the second group, you could statistically predict the sables would
out-number the albinos in a 3:1 ratio (3 sables to 1 albino).  You could
additionally predict that out of every four kits, one would be pure sable
(SS), two would be half sable and half albino (Sa), and one would be pure
albino (aa).
 
Those are statistical probabilities, which means for the ratios to really
be seen, you need a large population, because random chance can throw
off the ratios in individual litters; you might have half albinos, or
all sables.  Just like in baseball, you know someone with a .250 batting
average will get a hit once out of four times at bat--the problem is,
that is only true in the long run.  In the short run, you have streaks of
hits or outs that superficially appear to be proving the ratios wrong.
You can prove this yourself, just by tossing a coin.  Each time you flip
the coin, the chance is 50% it will land on the heads, yet you might flip
4 or 5 heads in a row.  Later, you might have streaks of tails.  In the
long run, such chance streaks balance each other out, preserving the
ratio.  Likewise, individual litters might not seem to have the proper
ratios, but if you breed for a number of years, or run a huge ferret
farm, the ratios are surprisingly accurate.
 
What inbreeding does is to change the probabilities of getting the
desired trait, making it more likely to get albinos, if that is what you
are breeding for.  As such, inbreeding is a very valuable tool to insure
offspring will have the best chance to have the trait you want.  The
reason this works is because inbreeding DECREASES genetic variation.
This is how it works: Jill and Hob are unrelated, and--for the sake of
this example--they have completely different sets of genes.  In other
words, Hob might be albino, Jill is sable, Hob is longhaired, Jill is
shorthaired--that sort of thing.  In this hypothetical situation, there
is 100% variability and 0% relatedness.  After mating, all the offspring
have half of their genes from Jill and half from Hob, so they have 50%
variability and 50% relatedness (by relatedness, I mean they share 50%
of their genes with their siblings).  If a couple of kits show a trait
you are looking for and you decide to inbreed them to conserve it, the
resulting offspring will have a range of variability and relatedness
depending on the genetic shuffle.  It can be as low as 50%-50%
(unlikely), or as high as 100%-0% (also unlikely); they will likely
average to 25% variability and 75% relatedness, meaning you roughly have
a 75% chance of any one ferret showing the favored trait.  If you inbreed
long enough, the genetic variability can drop close to zero, with nearly
100% relatedness.  If you just compared females to females, and males to
males, they can theoretically be identical twins of each other.  That is
how scientists bred certain strains of animals to be used for medical
research; if you want to study the effect of a single gene mutation, it
is really handy to have a population where the rest of the genes are
nearly identical; one of the many reasons why learning how to clone
is so attractive to genetic researchers.  At the end of this type of
inbreeding, the probability of any one individual having the favored
trait is nearly 100%.
 
The problem is that while you are breeding for your favored trait, there
are thousands of other genes on that one chromosome, not to mention all
the other chromosomes that make up the ferret's genetic code.  Inbreeding
decreases the amount of variability there as well.  If the genes are
healthy and there are no mutations, inbreeding can be done with relative
safety.  However, it is a different story if defective or mutated genes
are present.  In that case, as the percent of relatedness increases, the
chances of bad ones being expressed increases as well.  By inbreeding,
it is possible to accidentally generate a strain of ferrets that will
develop a specific cancer by a certain age, are deaf, or have clubfeet,
when all you wanted was a bigger or darker ferret.
[Posted in FML issue 4786]