Thoughts and Observations:
*I wanted to note at the beginning of this post that I am writing one post this month for the blog, which will also be used for my monthly article in the Bluegrass Hemerocallis Society newsletter as well as for Exhibition Poultry E-zine. This article is a general article on breeding techniques with a focus on inbreeding and is fairly equally applicable to animal breeding and plant breeding. I may use specific examples of a particular plant or animal, so if you want it to specifically apply to your plant or animal, just change the name in your head to what you are working with. The underlying principals are the same, with only one exception - you can't self-mate (most) animals as you can with many plants.
Breeding in many ways is as much an art as a science. Science (genetics especially) is always a valuable tool in the breeder’s toolbox, but I always like to point out that our ancestors created the vast majority of domestic plants and animals long before there was any knowledge of genes, beyond perhaps the most cursory understanding that "blood will tell". The art of breeding, like any art, has a lot to do with individual taste. What I may adore, you may despise and vice-versa, but in the end, we all use basically the same set of techniques to produce our desired outcomes.
There are really only so many approaches to breeding that one can take - outcrossing, backcrossing, line breeding, inbreeding, etc. We all use these techniques to one extent or another. It is interesting to me that in some circles though, one or the other of these techniques have taken on a superstitious veneer of 'taboo' status. For instance, in ornamental plant breeding, we see a majority of outcrossing and little inbreeding (or at least little that is openly admitted), generally due to the old notion that inbreeding is "bad". In animals, we see little outcrossing (again that is openly admitted) but lots of inbreeding, due to the equally superstitious notions of "crossbred" and "mongrel" or "un-pure" as opposed to "purebred". Each set of suppositions limits the breeders in those respective arenas.
There are no good or bad breeding techniques. There are only breeding techniques, and in some instances, each is of great value while in other instances, any given technique can be disastrous. It really all depends on the individual situation and the desired outcome. Inbreeding particularly gets a really bad rap, and there are some very negative potentialities to inbreeding if it is not applied carefully.
Inbreeding is used to concentrate genetic traits. It is that simple, and inbreeding will concentrate whatever is there, good or bad. Inbreeding can be especially valuable in determining what recessive genes are lurking in a particular population. If there are bad genes, deleterious genes, hidden in a population you can bet that inbreeding will bring them to the surface, but there is also no better way to determine what bad genes are lurking in a particular population. On the other hand, there is no better technique than inbreeding to determine what good genes are lurking in a population. As well, inbreeding can reveal how extreme a given trait can become and if a trait that is neutral in lesser expression will become dysfunctional or deleterious in higher concentrations.
I have used inbreeding a great deal in strain development both to concentrate desired traits and to reveal undesirable traits in both plants and animals. I always approach inbreeding with a very careful eye, quick to note when there are undesired traits and I am equally quick to eliminate a line or discontinue inbreeding when it becomes clear that there are many deleterious factors carried in a line. However, in some instances, inbreeding for generation may reveal no deleterious factors and in such instance, may only concentrate and strengthen the positive traits of the lineage.
In my work with daylilies, one of the first tests I like to run on a given cultivar is to self it. That is, breed it to itself. The first thing I am looking for is if the cultivar is self-fertile. The next thing I am looking for is how clone-like the seedlings of selfing may be, or how much variation they show. The former indicates a great deal of homozygosity and/or many dominant genes, while the later indicates heterozygosity of some or many genes. In the case of the former, I will then know that I may not be able to create anything very different from the parent in an F1 outcross, and so it becomes a producer of bridge plants for further breeding. If the plant is highly heterozygous for many traits, then one can expect to produce a wide range of phenotypes in the F1, and feasibly something different enough from the parent or unique enough to potentially be an introduction in its own right. However, I still consider such a cross more for the production of breeding material than for introductions, but that is my focus anyhow. For me, introductions are really the last thing on my mind with any cross at this time, while the production of new and improved breeding materials is always paramount in my efforts.
Chickens can't be selfed, but they can be crossed with full or half siblings and they can be backcrossed to either parent. To me, this is the only really tight inbreeding that is likely to reveal the hidden traits I wish to evaluate. The breeding of cousins, for instance, while inbreeding in the broadest sense, is to my way of thinking more in line with line breeding, and less pure inbreeding. While the purest inbreeding in plants is selfing, sib x sib matings as well as backcrossing to either parent (or aunt/uncle, grandparents) are useful techniques of inbreeding and can be equally as revealing for strengths or weaknesses as selfing. Backcrossing is particularly suited to revealing the hidden details of one or the other of the parent lines. The biggest pitfall with inbreeding comes when people get so tunnel-visioned on a particular trait that they ignore problem genes, sacrificing viability for phenotype extremes.
To put it more bluntly and in more common language, if you have bad traits in a line and you inbreed, you are likely to destroy that line as the bad traits become more and more concentrated. Conversely, if you have good traits, but no bad traits, inbreeding can concentrate those good traits and make them even more prominent. In reality, most lines have some of each, and so inbreeding must be approached cautiously and with care, being always aware to stop inbreeding when bad traits become glaringly obvious, no matter how "good" the line looks otherwise.
In animals, any given breed or line is likely to be the product of some inbreeding already. With animals, inbreeding can very quickly cause a lot of problems, and only the most vigorous and hardy lines are usually good subjects for intensive inbreeding. However, inbreeding is much more common in animals, thus the many problems we see in many, many breeds of domestic animals. The only instances where I would encourage intensive inbreeding in animals is where there are no obvious deleterious factors present (so, so unbelievable rare!) or where you have outcrossed and are working to make a new line, variety or breed and wish to concentrate good traits while working very hard to eliminate bad traits.
Plants may be another thing altogether, though. In daylilies, for instance, there are no know "pure" lines, as we see in so many domestic animals. Almost all daylilies will be heterozygous for a lot of alleles and are the results of much outcrossing, starting with the hybridizing of species about a century ago. It is my opinion that daylilies may be good candidates for some careful inbreeding, in some cases. Now we don't want to get carried away (as a cautionary tale, just look at what the animal people have done to their charges through "pure-breeding"!!), but there is much that careful inbreeding may accomplish. Of course, we want to be ever vigilant about inbreeding depression and the concentration of bad traits, but since most daylilies are so heterozygous, I doubt we are on the precipice of inbreeding collapse, as are so many domestic animal breeds. I would note however that certain traits seem to be questionable in daylilies already, so concentrating those traits may be unwise. For example, think edges so profuse they don't allow the flowers to open well or weak scapes with oversized flowers, etc. Even without intensive inbreeding in daylilies, there are problem areas and these need to be admitted and faced so that they don't get concentrated into 'ruined finery', so to speak.
To consider inbreeding in daylilies, let us consider an example. Let us say that there is a new, novel trait that has appeared in daylilies and we want to perpetuate that trait and perhaps even intensify it. If the trait is strongly dominant, then we can just cross to most anything, selecting those that express the trait most strongly and those, which combine the trait with other traits for new, novel phenotypes. But if the trait is recessive, then outcrossing it willy-nilly, crossing "pretty to pretty", is liable to give us very little that we want in the F1. Granted, that F1 can be interbred to bring the trait back out, but I have encountered very few daylily breeders who are going through generations of bridge plants to get to the desired outcome. Most seem to chuck the whole batch if they don't have an intro in the F1, though that doesn't apply to everyone and there are already some notable examples of just this type of inbreeding in the history of daylilies.
So viable options for breeding such a trait would include selfing, backcrossing to either/both parents and/or creating an F1 through outcrossing and then interbreeding those F1 or backcrossing the F1 to the parent with the desired trait and/or the grandparents that produced the trait in the parent showing the trait. While I would hesitate to do anything other than outcross and interbreed the F1 if there are problems in the parent with the new trait, I would also want to look at its parents to see how strong each of them are.
Since daylilies are highly heterozygous (generally), a weak plant or one with a given problem may not be concentrated for that trait. If the parents of that plant are easily obtainable and/or much is known about them, you may find that the problem trait is a dominant and may be heterozygous. If that is the case, then selfing or any other inbreeding may allow you to select out the desired trait in combination with other good traits and to eliminate the undesired bad trait or weakness. Genetically, things are not always as they seem, or more technically, phenotype is not always a full indicator of genotype. It is only through some experimentation that you can really know what is what genetically, the true breeding value, and thus have a fairly clear idea of the real limitations and strengths of a given cultivar or seedling, and one outcross that produces 10 or 15 seedlings really isn't going to tell you much.
The instance that I describe above is very analogous to animal breeding where outcrossing has been done (except for the selfing, of course) and a new combination of traits is being sought; i.e., a new breed or a new variety of an existing breed. Pure breeding is rather unique to the animal people. Can you imagine the ornamental plant breeder that would be striving to produce a cultivar that never threw offspring that looked any different from itself? (While this does occur in some commercial plant breeding, especially with some of the old, true breeding food plants, it is by far the exception in breeding for ornamental plants.) The thought is laughable and ludicrous, generally speaking, but it is a deadly serious affair to animal breeders where the brainwashing tends toward the notion that each individual of a given breed or line should look like identical twins or clones. Domestic plant breeding is much more in line with the notion of 'landraces' in animals. I think there is a lot to be said for landraces, especially the fact that it is nearly unheard of to find a landrace that is so concentrated for deleterious traits that their very existence is questionable. With that said though, there is a place for inbreeding in both the plant and animal worlds but the practice must be approached with caution and an understanding of the potential pitfalls, along with an eye for the slightest indications of problems. When used wisely, inbreeding is an indispensable tool for the breeder.