Thursday, November 05, 2015

Substitution Costs In Dog Breeding

One of the reasons working dogs and show dogs have generally separated into two very different breeds is that it is almost impossible to breed "top of the line" performers in two arenas at once, and that is especially true for a slow-breeder (any large mammal) and one that has a small gene pool (all show dogs).

JBS Haldane pointed this out about 50 years ago when he was writing about "substitution costs" faced by breeders. This is now known as "Haldane's Dilemma."

Haldane, along with Sewall Wright (of COI or Wright's Coefficient of Inbreedining fame) and Ronald Fischer more or less created the math we still use to today when looking at population genetics.

Yes, you can try to "shoot for the middle" in terms of two or more characteristics, but the results, for the most part, will not be top-of-the-line in either arena unless you are selecting for hundreds, or even thousands of years, which no one has done in the world of dogs.

It should be said that Haldane may have inflated the true cost of substitution -- a point he made himself.  Nonetheless, there is very little doubt that a cost is there, and it is large and is most easily addressed by either having massive amounts of time, or really large populations, or very rapid breeding (as in mice and rats), none of which we have in the world of show dogs.

For a related post, see Islands of Wolves, Rats, Lions and Dogs.

By the way, today is JBS Haldane's birthday -- born November 5, 1892.


Richard Gilbert said...

Haldane was a genius, but based on my experience I'd tweak this, even though the principle is 100 percent correct that you get, more or less, what you select for over time and that you cannot select for more than a few qualities and make progress.

In breeding livestock, I learned that the heritability of appearance is fairly high, whereas performance traits often are lowly heritable. And unlike with appearance, with performance it can be hard to separate what's genetic from what's environmental.

By my reckoning this makes the purple ribbon crowd, in whatever species, even worse than Haldane's principle would have them. They are doing something fairly easy and it is their complete lack of interest in function that completely destroys their chosen breed's practical qualities. So it is not at all that their animals decline because multiple trait selection is hard; they are not even trying for practical balance.

Performance folk and show folk don't mix. Never have, never will.

Donald McCaig said...

Dear Patrick,

It seems to take about three generations of show breeding to disrupt/destroy the sheepdog's working behaviors==. The working genes are separated on the genome (not inherited as a bundle) and once you breed away from them you get fragments. Keenness, for instance, isn't much use without bidability and both won't help you bring in your sheep if the dog doesn't have the genetic urge to go to the heads of stock and fetch them to the shepherd.

While it is unlikely a 3 gens away from working stock dog will work, it is entirely possible that a fine working dog may resemble dogs winning ribbons in show rings.

A breeder friend of mine with such a dog was pestered to create a "Dual champion" - just a little show blood in his working gene pool.

He replied, "Suppose you had a gallon of ice cream. How much shit added would stop you from eating?"

Donald McCaig

PBurns said...

The folks that breed show dogs in closed registries without performance or health tests are SURE that if you breed for looks alone, the rest will follow. And yet, if you ask them for any working breed that is older than 50 years, where that is true, their voice drifts off. Well... there's a few dogs in (insert country in which they do not live) that work (left undefined).

Different breeds fall out for different reasons. Some lose temperament, nose, or the complete package. With terriers, size falls apart very fast, with dogs too big to get into small pipes. Temperament also falls apart pretty quickly, as it's a careful balancing point. Too hard and the dog wrecks itself; too soft and it fades or jibes, or simply does not enter. If you do not work your dogs regularly, they never achieve their potential and you never see what is there and what is missing; you end up breeding like a blind painter mixing paint.