Some of you, at some point in your lives, may have owned a dog. If it was a purebred dog, say, a dachshund, you're probably aware of the high prevalence of genetic disease in purebred populations. People often joke about how inbreeding produces weird, mutant offspring. This is true: every individual has (by some estimates) roughly 500 to 2000 alleles for a deleterious phenotype. Most of the time this is no problem, since you have two copies of every gene and the effect of most of these alleles will be masked by a compensatory "good" copy. However, your kids might have problems if you're unfortunate enough to mate with someone who has the same allele. If it's a relatively frequent allele, present in, say, 1 in 1000 people, each one of your offspring has about a 1 in 4000 chance of ending up with a genetic disease. Fortunately, the vast majority of those deleterious alleles are going to be VERY rare, what's called "private mutations" (meaning, essentially, only you have them), so that we don't end up being riddled with genetic disease. For best results, marry someone from the other side of the planet.

Now, consider what would happen if you mated with your parent. Don't consider it too hard, just keep it abstract and hypothetical, so we can get through this. This is regularly done by breeders, to reduce the number of variables in the breeding process. You can be sure that you're not introducing any new, unwanted alleles if you breed against a parent. In this case, since you share 50% of your genetic material with your parent, the odds of acquiring a genetic disease go way, way up. Suddenly, your offspring have a 1 in 8 chance of ending up with two copies of a deleterious allele, even if it's a very, very rare private mutation. Multiplied out over 500 to 2000 alleles, you're looking at some very sick puppies.

This can have drastic effects on a population. If genetic diversity is too low, disease can prevent a population from expanding. I read a study about a population of Norwegian (maybe Finnish) wolves crippled by low genetic diversity for years - until the population suddenly began to explode, thanks to the genetic diversity contributed by a single foreign interloper.

So: inbreeding, bad. Genetic diversity, good.

Unless you're a breeder. In that case, genetic diversity is bad. You're trying to create an archetype here, not maintain a healthy population. Variation is a nuisance.

This is especially true in industrial agriculture, where breeders wish to maximize productivity of meat, milk, fur, etc., and hence profits, and minimize variation, which interferes with the standardization of industrial processes. Agricultural populations, though enormous in census size (there are something like 1.5 billion cattle worldwide), can have the same level of genetic diversity as a vastly smaller normal breeding population - as little as a few dozen individuals. This is because a single stud bull's semen can be used to impregnate thousands of cows artificially.

And here's the punchline: apparently, this level of uniformity is not good enough. No, now we're going to start cloning our meat. 22,000 identical rump-roasts, anyone?

On the other hand, maybe a fragile industrial farm animal population is a good idea. If all the industrial cattle in the world are suddenly wiped out by a bovine plague, think of all the boons: less pesticide and animal-waste runoff, less cow-flatulant contribution to global warming. And all the genetically modified corn and soybean we can eat. Woot!