Friday, July 27, 2012
Genetics and Selection
The Use of Genetics and Selection in Breeding Poultry
In the hobby of breeding domestic fowl, we tend to focus on the issue of genetics as a one-trait/one-gene affair. This is not always the case and linkage can produce what appears to be one gene that is actually several genes coming together and interacting with the chemical pathways of pigment production to make the finished phenotype. Many other factors can combine to make the appearance of a “gene” that is in actuality several genes, linked or not, interacting. What we do in the hobby when we speak of “genes” is to actually describe heritable factors. A heritable factor is often observed to show a distinct Mendelian genetic segregation ratio. In many case, the heritable factor is likely, in fact, a single gene with a noted dominance or recessiveness. However, in some cases a heritable factor is not a single gene. Pattern gene is a noted example of multi-gene interactions involving linkages. We say that there is one pattern gene (Pg). Carefoot showed good statistical evidence that this is so and that there is one major gene, interacting with a handful of other well-known heritable factors or “genes” to produce the range of patterned varieties we see in the modern exhibition breeds. I agree with his assessments. I have however noted that “pattern gene” can show a range of expressions. Some of these are of course examples of various levels of heterozygosity, both for needed factors and unneeded factors that are detrimental to the desired visual phenotype. In other cases, testing suggests that birds can be homozygotes for all the needed genes and yet they do not express that same type of expression. Lacing is a prime example of this. Lacing requires the linkage group of Pg/db+/Ml with Co and preferably on ER or eb. Er birds also require the gene Db, which is in a linkage pattern with Ml and Pg. However lacing ranges from a very fine lace to a heavy, muddy lace on both e-alleles. I have seen cases of lacing that were on the same e-allele and tested at homozygosity for the required genes, but produced drastically different lacing and outcrosses of the two, in generations past the F1 tended to segregate into one type or the other; fine laced or heavy laced. Are we seeing two different alleles of Pg? Two different genes completely that do very similar things, or perhaps we are seeing the effect of another gene such as a melanizer or even a factor that merely modifies the expression of the lace? Could there be a factor that changes the expression of Pg at some point in the developmental pigment pathway? I don’t know, but there are clearly such differences across the entire expression of the heritable factor Pg and my observations suggest that the differences are in some way heritable, and segregating, as one would expect for a Mendelian gene. Such questions are fascinating and may in time be answered.
In the hobby, we would seek to test mate to produce statistical models of segregation pattern. We then name a gene, using our statistics to prove the unseen presence of a strand of DNA somewhere on a chromosome. This is the research model of one hundred years ago. It is usable, as far as it goes, but it may be a poor model for naming and designating genes. It is more attuned to diagnosing heritable factors. Heritable factors may not always have a basis on a one-gene/one-trait scenario. However, that does not mean they are not useful.
In the real world of laboratory research, where actual genes (groups of DNA base pairs working in unison to produce specific proteins) are mapped and traits linked to such, we see a real capacity to name a “gene”. When they say that they have mapped the gene, they mean that literally. In the hobby, when we say that we have discovered a gene, we have actually observed a heritable factor. Beyond really large and long range data collection using huge numbers, it is not very credulous to put too much faith in the small numbers we work with to definitively declare a “gene”, because we in no way have observed such. With that said though, the observation of a heritable pattern is still useful and factual. A hobbyist doesn’t need to know where the gene is or the designation given by the scientists who have mapped a given gene. What the hobbyist who wishes to pursue any given phenotype or trait needs to know is heritable probabilities. Some factors have very good statistical data to confirm that they are truly a gene, in the literal sense, and in some cases, our friends in the laboratory are showing us where those genes actually are in the DNA. To me, this is extremely fascinating and engrossing. I suspect that in time, such research will become extremely useful to the hobbyists of the future. I just hope I live to see the hand held chicken egg scanner that tells you if the embryo is alive and its entire genetic structure so you can decide whether to set the egg or not. I visualize it looking much like a tricorder from Star Trek, or perhaps even as an app for your Iphone. Until then, however, for the most part, we rely on observed and observable data concerning heritable traits. There is probably a high correlation between observed traits and DNA segments, but the important part is to understand how the factor behaves and if it is a simple factor or a complex factor. The former can be followed with a simple Mendelian ratio while the later requires a quantitative approach.
So how does the hobbyist apply the gathered knowledge of domestic fowl heritable factors? They decide what traits they are selecting for, first and foremost. Most will focus on color, while some will focus on form and a few will work on more intangible traits. The first step is to really, honestly, assess where you want to focus and then make a list of the traits that are most important. Here is where you are going to focus. However, it is not impossible to give a large range of traits some consideration. I would respectfully suggest that everyone could benefit their birds and themselves by selecting for hardiness and heritable disease resistance. Some deeply inbred lines could use an outcross here and there to something that isn’t nearly as bottle necked as they are. Aggressive birds are not always a joy to work with and birds with poor egg production make reproduction difficult and the possibility of a bottleneck become that much higher. If we first remember that no matter which “breed” we are working with we are breeding domestic fowl, then we can select for a healthy and functional domestic fowl as well as the traits that make it the given “breed”.
Selection for many traits requires a quantitative approach. Select a few traits that are most important to you and these are given the most emphasis, scoring higher in your evaluation. The absence of said traits can be cause to eliminate a bird if the heritable factor is generally dominant. The absence of a recessive gene in the phenotype can still make for a useful breeder, if it is a heterozygote for the gene. However, do not stop there and note a whole range of factors so you can know your line. Evaluate it honestly and when all else is equal in your priority traits, select the individuals with the best secondary traits on your list. In each generation, your goal is to merely increase the presence of as many of your selected traits as possible. The actual increase can be small, perhaps only five or ten percent, but if you are seeing any percentage of increase in trait expression then you are making progress on the population as a whole. That is how you form a strain. A strain is a highly selected line of relatives. They produce recognizable combinations in most instances. On the level of selection for heritable traits, concentrating the selected heritable traits into high expression in the phenotype is the hallmark of a ‘strain’. This also implies high levels of homozygosity for the heritable factors, such as genes, gene combinations, or other modifying interactions that may affect the phenotype.
The key to fast progress in heritable trait selection is to obtain very homozygous stock from a strain or to work with large numbers in order to apply selection pressure to increase the homozygosity for the targeted traits. The smaller the number raised, the slower your progress may be. However, even when working with small numbers, selection albeit limited, can still be applied. Genetic knowledge can be of great use to the person who can only work with a small number of individuals but wants to create some type of advancement with their stock. The best advice I can offer is to look for very hardy, fertile lines with good production and temperament and then bring in traits from outside of that line where needed through outcrosses to create your target phenotype. If you can find a strain with good domestic fowl qualities and expressing highly bred phenotype traits, they are jewels of great value and can both be bred as-is and used in improving other lines with similar heritable traits. There are a few well know varieties of certain exhibition breeds that are noted for being very well bred, with many good traits combined. They usually aren’t the most ‘fru-fru’ and frilly, but they are very good, sound birds that are a joy to work with and can be used to reinvigorate a great many breeds that share one or two major traits in common. When using this technique, a working knowledge of heritability traits can be valuable.