How backcrossing works

One of the common arguments for maintaining the closed registry system is that as soon as one does an outcross it will be impossible to breed back to the original phenotype.

This is actually not true, and with dogs, it does not take many generations to breed back before the dog looks and behaves exactly like the purebred.

I have already mentioned a nineteenth century program that introduced bloodhound into the British basset hound population to save it from an inbreeding depression. I have also mentioned Bruce Cattanach’s bobtailed boxer program, which used a naturally bobtailed corgi to introduce the trait into the boxer. In both cases, the dogs were able to return to their original type through relatively few generations of breeding back into the boxer gene pool.

Cattanach writes about how quickly he could breed dogs that looked very much like pure boxers:

The transformation in one generation can only be described as amazing. It suggests that very few genes are responsible for the main features distinguishing the Corgi and Boxer, except for the special Boxer head. The white coat colour, of course, was Boxer white and resulted from the doubling up on Boxer.

That’s right. Just a few genes separate all of these dogs breeds. Variation on only a single gene explains most of the wide variance in size among dogs.  Size is very easily selected for in breeding programs, which is why we have three widely varying sizes of poodle that all descend from essentially the same stock.

If a breed has a particularly specialized head, it may take few more generations to “fix.” The bull terrier in Britain was traditionally a white dog. After all, it was derived from the English white terrier. The English white terrier went extinct because it developed severe genetic genetic problems– among them deafness. Deafness also affected the white bull terrier breed, and it was feared that it would follow its white terrier ancestor into oblivion if something was not done to correct it. The English white terrier was extinct by 1900, and within just a few years of its extinction, crosses between bull terriers and a few select Staffordshire bull terriers occurred. One of the staffies used in the program was a first cross between a bull dog and a Manchester terrier.

Although it was easy to return to the bull terrier phenotype, it was very hard to breed the special bull terrier head in the colored lines. Bull terriers have a very specialized, egg-shaped head, and they also have triangular eyes, which were very hard to fix in the colored lines. The prevalence of “button eyes” in colored strains led some fanciers to denounce them as mongrel, and it is one of the reasons why the AKC has two separate varieties of bull terrier: white and colored.

However, even those button-eyed coloreds looked very much like bull terriers. Perhaps if the bull terrier fanciers had been more open to breeding from them, they would have become more or less like the white dogs much more quickly.

But through backcrossing and selective breeding, the colored bull terriers now have classic bull terrier conformation. One even won best in show at Westminster  in 2006.

Simply put, the argument that it is impossible to breed back traits after doing an outcross is simply bogus. The fact that it takes just a few generations of backcrossing to produce a dog that is virtually identical to the backcrossed breed shows how easy it is.

Albert Payson Terhune evidently knew of this fact when wrote the story of a collie named Buff. Nina, Buff’s mother, was an accidental cross between a collie and pit bull, and she accidentally mated with a top show collie. Buff resembled a perfect show collie, even though his grandsire had been a pit bull.

There actually was a collie named Buff. His photo appears in the frontispiece of the book, but I do not think that this collie actually had this ancestry. It is possible, but I doubt it.

The rapid effects of backcrossing in dogs have been well-established for quite some time. Stonehenge (John Henry Walsh) was quite aware of how easy it was to return to phenotype after just a few generations of backcrossing. He used an experimental crossbreeding between a bulldog and a greyhound. Writing about this crossbreeding and back crossing program in his The Dog in Health and Disease (1859), Stonehenge points out how quickly it easy to produce dogs that resemble greyhound from backcrossing:

It might naturally be supposed by any person who has not been convinced to the contrary, that it would take several crosses to get rid of the heavy form of the bulldog when united with the light and graceful shape of the greyhound. But on actually trying the experiment it will readily be seen that in the third generation very little trace remains of the bulldog, while in the fourth there is none whatever apparent in external form. My friend Mr. Hanley is the last who has tried the experiment, and having kept a daguerreotype of every individual used in it, which he has kindly placed at my service, I have been enabled to present to my readers perfectly trustworthy proofs of the correctness of this assertion. The bulldog “Chicken” used was a very high-bred animal, and of him also Mr. Hanley has preserved a daguerreotype, but as his blood is very similar to that of Mr. Stockdale’s “Top,” I have not thought it necessary to engrave him. The bitch “Fly,” put to “Chicken,” was also highly bred (pg. 179).

One must note that Stonehenge writes that Chicken was nearly identical to a bulldog named Top, which he depicts in his section on bulldogs. The first cross between Chicken and Fly, called a “Half- and-Half,” is a meld of features from the greyhound and this old type of bulldog:

The “Half-and-Half” was then bred to a greyhound named Blunder, and this breeding produced a white bitch named Hecate:

Stonehenge describes Hecate:

From these came the second cross, “Hecate,” a white bitch still presenting some slight characteristics of the bulldog breed, but by an ordinary observer this would be scarcely noticed. There is, however, a remarkable want of symmetry and true proportion in this bitch, which the portrait conveys exactly (pg. 181-182)

Hecate may not have been what greyhound fanciers wanted in an ideal specimen, but she doesn’t look all that much like a bulldog. She just looks like an “off” greyhound.

Hecate was bred to another greyhound named Preston. He was a very fast dog, and it was thought that he would pass on these traits to his offspring. One of the puppies produced from that breeding was a black bitch named Hecuba:

Stonehenge describes Hecuba as “a large black bitch of good shape, and, as I before remarked scarcely distinguishable from the pure greyhound” (182-183).

She was a very fast dog, but she lacked stamina. This finding suggests that certain working characteristics might be hard to breed back through backcrossing when the original outcross is between two very different breeds.

Hecuba was bred to another greyhound named Bedlamite, and the offspring that resulted from this litter were fast but were deficient in “stoutness.” Stonehenge shows a depiction of one of these dogs. Her name is Hysterics, and she is very clearly a greyhound.

Hysterics was then bred to Ranter, her full greyhound half brother, and the puppies that resulted from that breeding were not as good as the fourth cross. Perhaps such tight breeding caused these deficiencies.

So even in Stonehenge’s day, it was well-known among dog fanciers that it didn’t take many generations of backcrossing from an outcross to produce dogs that had the correct phenotype.

And in Stonehenge’s day, Gregor Mendel’s work was not yet accepted as science. Although Mendel was conducting his experiments at the time of Stonehenge’s writing, his work was essentially unknown to the British public.

Now, we have a much more complete understanding of genetics. It is not complete by any means, but we know how many traits are inherited dogs. Because we have this knowledge, it will be easier for us to engage in cross-breeding and backcrossing programs. We also know how to test for many genetic diseases, and we can test both breeds used in these programs for certain inherited diseases.

We know so much more than they did, and we could use crossbreeding and backcrossing programs to improve the health and diversity of many breeds.

However, institutionally, there are many barriers to these programs. The Dalmatian Backcross Project has produced Dalmatians with low uric acid concentration in their urine. Uric acid stones are major problem in the breed, so it was decided to make a cross with a pointer and then backcross to produce Dalmatians that have low uric acid concentration. After generations of backcrossing, these dogs are now 99.7 percent Dalmatian.

But the AKC and the Dalmatian Club of America have been resistant to allowing these dogs to be registered. The AKC recently deferred the decision to include these dogs to the Dalmatian Club of America. The DCA still refuses to accept them. The DCA has also started a propaganda campaign in its own literature, claiming that if such dogs are allowed in, the Dalmatian will no longer be purebred and the health of the breed will deteriorate.

All of these things are unlikely to happen, and if a dog has an old pointer ancestor but still looks and behaves like a Dalmatian, what difference does it make? The average dog owner might want a dog that looks like a Dalmatian and acts like one, but they also want one that is healthy. I don’t see what the big deal is– unless purity is such an overarching virtue that one “bastardization” several generations back negates the  validity whole strain. Such a position is actually quite hard to defend to the average person looking for a dog, and this might not be the best public relations step to take.

Backcrossing allows the breeder an opportunity to return to phenotype and working ability. It allows those genes to return to the bloodline, which also has the genes from the outcrossed breed. If those genes add something to dog– such as a healthy urinary tract or a naturally bobbed tail– then these outcrossing projects are worthwhile endeavors.

Because of the successes of these sorts of programs, the dog world should be more willing to operate with an appendix registry system. As we saw with the greyhound/bulldog project, not every dog produced was worthy of breeding– as is the case with purebreds. An appendix registry allows dogs that meet some of the breed requirements to be registered, and then these animals can be bred to the other dogs in the registry. When puppies are born, they are also checked to see if they fit the criteria and are registered accordingly.

The cat fancy has totally embraced the crossbreeding and backcrossing of many different breeds. The CFA allows certain breeds– such as Persians and Exotic shorthairs– to be crossbred. These two cats have very similar conformation, just one is long-haired and the other is short-haired. The animals are crossed and registered according to phenotype– something that was done in retrievers and spaniels for decades.  Cat fanciers are given a greater opportunity to selectively breed from diverse bloodlines than dog fanciers typically are.

Crossbreeding and backcrossing are tools that should be open to more dog breeders. They are tools that do require skill to use properly, but the skills can be learned. They require understanding the genetics of what makes up a particular breed or variety and that of the outcrossed breed. This sounds very scary to many traditionalists within the dog fancy. It needn’t be.

In the nineteenth century, they were doing these crossbreeding experiments with no real knowledge about health or genetics.

And the dogs didn’t fall apart.

We know so much more now. And it is time we get a chance to put that knowledge to work to build a better future for Canis lupus familiaris.