This article is from the rec.pets.dogs: Breeding Your Dog, posted to rec.pets.dogs newsgroup. Maintained by Cindy Tittle Moore with numerous contributions by others.
_If a purebred dog of breed X mated with a purebred dog of breed Y, both meeting health standards for their breed, is there a better chance the offspring would be healthier than a same breed mating because the gene pool is larger?_
In terms of health alone the first answer would be that in breeding two healthy dogs it shouldn't matter if they're the same of different breeds, you're apt to get healthy pups. But this doesn't take into account the question of recessives. Suppose you breed two dogs of different breeds that both have the same incidence of a recessive health problem. The pups would have the same odds of having that health problem as purebred pups of either breed. On the other hand, suppose the two dogs were of breeds that have no recessive health problems in common. This would reduce or eliminate the odds of the puppies of having the health problems of either breed. This is the classic explanation for the theory of first generation hybrid vigor. The resulting pups should not be bred though, since they'd have a good chance of having the recessives from BOTH breeds, so the grandpups would be inclined to be worse off than the purebred offspring of their grandparents. An excellent set of articles dealing with "hybrid vigor" can be found in _DogWorld_, Jan 1997 by George Padgett DVM. Another _very_ important point to keep in mind is that when a purebred carrying a genetic defect is crossed with another breed or mixed breed, the "bad" genes do NOT "go away" even though they may not be expressed in the offspring. If crossed with another dog carrying the same defect, the offspring of that breeding _will_ demonstrate the defect.
_Purebred dogs have all these diseases, though! It seems that you never hear about mixed breed dogs with problems._
Responsible breeders try to identify genetic diseases their dogs might be carrying and to eliminate them by careful breeding. It is ironic, though not surprising, that their efforts to identify and weed out genetic problems have lead some to cry "look at all the genetic diseases purebred dogs have!" A moment's careful thought will lead you to the conclusion that mixed breeds carry the _same_ harmful genes (their parents, or their parents' parents, _were_ purebreds, after all). The differences are * with some recessive disorders (though not _all_ genetic defects) the disease is less likely to be _expressed_ (though it can still be inherited by offspring) * you have lesser likelihood of ever identifying or eliminating any harmful genes your mixed breed may be carrying
Also, if you stop and think about it, many mixed breeds are simply not tested for most problems. When they get older and limp, it's just considered old age, although it could well be hip dysplasia. When they get older and start to go blind, it could be PRA, but the owners are unlikely to test for this. It's not that owners of mixed breeds are bad, by any means, but they are not looking for possible inheritable problems, either.
_When you breed two different breeds together, what kind of variation can you expect?_
Pfaffenberger's book has some interesting data on this. He did some experiments with four different breeds. They were dogs of approximately the same size, but very different physical appearance AND behavior. The results he saw in the first and in subsequent mixed generations are pretty interesting.
Let's look at a common crossbreeding: "cockapoos" (which are _not_ purebred dogs, nor registered with any registry). These are crosses between Cocker Spaniels and Minature or Toy Poodles. The dogs actually vary quite a bit, some being more poodle like than others, and some being more cocker like than others. However, they are generally all a small sized, buff colored shaggy dog. If you breed two cockapoos together (not generally done), you get an even wider variation of dogs -- some look like Minature Poodles, others like Cocker Spaniels. The reason for this is the recessive genes hidden in the first cross that came out in the second generation. This is actually a visual example of why "hybrid vigor" doesn't hold.
_What is outcrossing?_
Outcrossing is where the sire and dam are totally unrelated, preferably for three or four generations. The true form of an outcross is between two entirely different breeds because in reality the members of most registered breeds come from a common ancestor (althought it may be many, many generations back). It is very rare for outcrossed puppies to be uniform in appearance. Usually there are a very large ranges of sizes, coats, colors, markings, and other distinctive characteristics. Outcrossed litters are generally heterozygous, and do not reliably reproduce themselves, so even the nicest puppy in the litter may not later produce the best puppies.
Outcrossing is generally used to introduce something new to a line -- a better head, better colors, better front, etc. Usually the puppies retained from these breedings are bred back into the breeder's original line to standardize them back into the line's general characteristics and reproducibility -- with the one desired characteristic. The tricky part is that other characteristics may come along for the ride!
If you are dedicated enough, you can eventually continue breeding by outcrossing alone (but don't expect instant or quick results). You should pick dogs that complement eachother well and are similar in general appearance. This is a long hard road to eventually developing a line. Through outcrossing, many health problems can quickly be eliminated (or just as quickly added into your breeding), but usually you do sacrifice some show quality and producibility.
You have to remember that dogs that appear totally healthy may be carriers of genetic problems. To find this out, test mating is done to a dog that is affected with the genetic problem (resulting usually in puppies that are both affected and non-affected carriers) or by inbreeding to a related dog that also doesn't show the signs of being affected (usually littermates are used) this will usually result in some puppies free of the problem, some puppies as carriers, and some puppies affected if both dogs carry the problem gene (this is not as accurate as breeding to an affected dog, but you are less likely to have to put all the puppies down).
There are variations on outcrossing. A "true" outcross could be a dog that has totally unrelated dogs bred together throughout the pedigree. This is very rare. On the other hand, "linecrossing" is a form of outcrossing where dogs from unrelated lines are bred to produce a new line. The sire and dam are usually very linebred from their prospective lines and the resulting puppies are varied in appreance, some looking like the sire's line and some looking like the dam's line and some looking like mixtures of both lines.
_How about line breeding?_
Line breeding is when the sire and the dam are distantly related: e.g., grandsire to granddaughter, granddam to grandson, second cousins, half cousins, uncle to niece, aunt to nephew..... The general strategy is that there is a common ancestor that is being doubled up on both sides. So the desired dog appears several times in the pedigree.
This is probably the most common strategy in breeding purebred dogs (and in developing new breeds, for that matter). Though this method, new genes are slowly introduced and unwanted genes are slowly replaced. The actual rate varies by how strongly you line breed. It sacrifices little overall quality in terms of show quality. Usually the puppies are rather close in general conformation. The only problem with this method is that it often takes several generations to get poor genes out, (or adding desired genes in) resulting in many puppies that have the same genetic problems (or virtues) that their parents have. And then because some breeders are more interested in winning, they do not place the affected puppies on spay/neuter contracts. This is both a blessing and a curse for the breed. If the breeder is very careful, affected pups can be used wisely to prevent loss of quality, but still remove the affected genes by only breeding the affected pups to known non-carrier relatives. This way the breeder can again try to "edit out" the bad genes. It takes longer this way but less show quality is lost in the process. This process results in dogs that will often reproduce their same level of quality. This is refered to as reaching homozygous litters (more genes of the same kind apparent in the puppies).
Inbreeding and linebreeding really differ only in degree. Linebreeding is less likely to cause harm than inbreeding. Inbreeding is not for novices. Knowledge of genetics and the breed is required for success. For good results it must be well-planned and breeders must be ready for whatever problems it presents.
Inbreeding is where the sire and the dam are closely related: mother to son, father to daughter, sister to brother, half sister to half brother, cousin to cousin. People disgree about the exact point at which inbreeding becomes linebreeding. Inbreeding is the quickest way to find out what poor genes are in the line and what dominant characteristics are in the line.
Although many people are disgusted with the idea of this family incest, it is an extremely useful tool for diagnosing what genes are present. If the genes for bad eyes are present, but hidden or resessive, this will bring them out to their full extent. If there isn't any bad genes, then the puppies will be of very close uniformity and very able to reproduce themselves (theroretically). This is a homozygous breeding. The resulting puppies will have a lot of genetic material that is the same as their parents and grandparents and will be close genetically to each other.
Inbreeding doesn't introduce new genes and does not eliminate bad genes that the line already has. It only shifts them around like a rubix cube. This often results in litters with high show potential, if the quality was high to begin with. It shows you what recessives you have lurking in the dogs' backgrounds -- _both_ good and bad. But there are drawbacks. Besides the possibility of bad recessives, inbreeding exclusively will eventually lead to infertility. It's like a xerox machine. After so many copies, you have to renew the ink. The same with dogs, you have to introduce new genes. No reputable breeder will use inbreeding exclusively, and many breeders simply never use it. Usually, you will only find: very experienced breeders, ignorant breeders, and puppy mills making use of this technique.
Inbreeding increases the chance that a gene obtained from the sire will match one obtained from the dam, both stemming from the common ancestor(s) on which the individual was inbred. Thus, inbreeding tends to make animals homozygous rather than heterozygous. The inbreeding coefficient measures the resulting increase in homozygousity. All breeds have a given degree of homozygosity the mating of two dogs from the same breed would not produce a recognizable specimen of the breed!
Inbreeding increases homozygosity and decrease heterozygosity. So it can duplicate both desirable and harmful alleles, both of which can be unsuspected in the line, and may appear. Inbreeding does NOT create anomalies, it brings present anomalies to the surface. Even when the anomalies are present, inbreeding might not reveal them. However, once revealed, then the breeder can do something about them in the next generations of breeding.
An increase in harmful recessives is undesirable but it is not a major drawback if they are identified early. The effect of inbreeding on major polygenic traits is greater. Generally, traits that are highly inherited (ie largely additively controlled) are not adversely affected by inbreeding but, traits under non-additive control, especially those tied to dominance and thus not of high heritability, are often markedly harmed by inbreeding.
_OK, how do pedigrees figure into this?_
Remember that it is difficult to spot unaffected carriers. When an affected dog shows up, its pedigree is often examined for likely carriers. For example, PRA is a common problem in many breeds. There are dogs that come down with PRA that have a certain ancestor in common. That ancestor may then be considered a possible carrier and line breeding on him is avoided. This is a simplistic picture, obviously, since it's possible for an unaffected non-carrier of PRA to come from an unaffected carrier that came from an affected dog (therefore the affected dog is in the unaffected dog's pedigree). If a general blood test is ever developed that shows the presence of the recessive in an unaffected dog, then much more accurate breedings may be done; currently this is only possible for Irish Setters.
There is rarely only a single problem a breeder is trying to screen for. Suppose a suspected carrier of PRA is known for producing excellent hips. A breeder might therefore introduce that bloodline into theirs for the hips, and be willing to have the possibility of PRA show up in the line. In screening out one problem you might have to accept the possibility of another appearing.
Examining the pedigrees also lets you know what percent of ancestry the dogs share (since the relationships are often much more complex than simply cousins or aunt/uncle, the degree of common ancestry is often given as a percentage instead) and decide whether or not it's acceptable given your current goals.
_What are like-to-like matings and compensatory matings?_
Like to like mating implies the best to the best and the worst to the worst where the worst is not used at all. For most breeders, like to like matings are between dogs which resemble each other greatly and so similar type dogs are bred. These dogs may or may not be closely related.
The pups resemble their parents because of the genes in common with them. If those parents resembled each other their progeny would be even more like their parents. This tends to make the population look more uniform, however there is little increase in prepotency from this technique.
Compensatory Mating: This unlike to unlike mating is used by breeders to correct for a defect in an animal by mating it to another animal that might correct for the defect. The system is basically simple but the breeder must identify faults and virtues and it requires breed knowledge. The pedigrees of both dogs should be examined carfully to try to identify the ways in which the dogs differ and what the expected outcomes could be. A correct dog and not one who errs in the opposite direction is required. That is, if you want to improve structure, look for a dog with correct structure and not an overbuilt dog. This technique often results in only one or two pups with the combination desired.
_But this is all very vague and complicated!_
Yes, it is. There are no easy answers, and there are different things to consider in every breed. This uncertainty with respect to genetic inheritance is exactly the reason that breeding is so difficult to do right. It helps immensely to have a "mentor", someone who is familiar not only with the breeds, but the lines your dog belongs to -- advice from such a knowledgeable person is often extremely valuable.
If we knew everything about genetics, we wouldn't _have_ problems with our dogs any more. We'd eliminate Hip Dysplasia, PRA, heart problems, thyroid problems, seizures, etc. within a few generations if we knew everything. Unfortunately it's an art that few people are actually very good at.