http://www.exoticfinches.com

Warning: All rights reserved. This article appeared in the National Finch and Softbill Society Bulletin Volume 22 No. 2. March/April 2005. p.8-10. Anyone wishing to reproduce this article for another bulletin, newsletter, article, journal, CD, or any other public forum needs express written consent of the NFSS and of the author michael@exoticfinches.com

In order to improve the quality of your birds through breeding it is necessary to understand what is meant by a normal distribution. When an entire population of a given species is measured for a quantitatively inherited trait (e.g. body size), the measured values within that population will be distributed in a certain pattern. When these values are displayed graphically, the result is a symmetric bell-shaped curve which exemplifies a normal distribution. Figure 1 below is a hypothetical example of a normal distribution for adult body weights of a wild population of zebra finches. The horizontal axis represents the value of measurement (in this example, weight in grams) and the vertical axis the frequency of birds in that weight class. Note that most observations are near the average 15.5g and very few observations are near the either end of the curve (e.g. 13.0 or 18.00 grams). It is a biological and statistical fact that, in populations that have not been manipulated, the vast majority of quantitative traits are distributed normally.

Why such a curve exists becomes evident when we study the genes that control the given trait. It is important to realize that individual genes behave in the familiar Mendelian segregation ratios that were discussed in previous installments of this series. Yet, because the trait is regulated by so many genes, a continuum, rather than discrete categories, results when one looks at that trait across a very large population.

Let's begin by looking at a trait such as bird size. If a trait such as bird size was simply inherited (e.g. AA = big, Aa = medium, aa = small) it would be very easy to develop a flock of big birds since there would be only 3 possible categories of young (Figure 2a). What if it took 2 genes being homozygous and dominant to attain the biggest size, but each dominant gene did help increase size? (Figure 2b).

That means that the AABB birds would be the biggest, AABb or AaBB not quite as big, and so on. The aabb birds would be the smallest. With two genes controlling size there are nine possibilities for the offspring and only the first is the biggest possible bird (AABB, AABb, AAbb, AaBB, AaBb, Aabb, aaBB, aaBb, and aabb). This does not mean that one of nine birds will be AABB because the frequency of AABB will be low (if you remember your Punnett square diagrams). Inside each bar in Figure 2 is the relative frequency of each genotype in a large population. The bar heights also represent the relative frequency. Note that we assume here that A and B have the same effect on size.

By now, you should see where I am heading. Traits such as bird size, roundness of head, breast width, size of the eyes, quality of the feathers, etc. are each controlled by many genes. The general formula for how many possible genetic combinations exist is 3n where n is the number of genes that influence the trait. This means that if 2 genes influence a trait there are 32 or 9 possible genotypes for the offspring (Fig. 2b). If 5 genes influence that trait there are 243 possible genotypes and if 10 genes control the trait there are 59,049 possible genotypes. If 10 genes must be homozygous dominant to achieve the maximum sized bird, you would need to produce about 60,000 offspring to isolate the largest possible specimen of that species. The truth is that we do not know how many genes influence size but judging from data of large populations it is certainly more than a few.

Note in Figure 2 that as we move from one gene (Fig. 2a) to two genes (Fig. 2b) we start to see the development of a bell curve. As the number of genes that control a trait increase, the distribution better fits a bell curve. Also, as the number of genes that regulate a trait increase, birds that fall into the categories at either end of the curve (i.e. the biggest or smallest birds) become relatively less frequent. A human size curve would show the same thing. We see very few 4 foot 5 inch or 7 foot 11 inch humans that do not have some genetic defect that caused the extreme. Human height, like bird size, is controlled by a large number of genes and fits a normal distribution.

It is important to realize that not all the traits must be physical. Even some behavioral traits may be impacted by a multitude of genes. It is likely that genetics plays some role in the willingness to breed in captivity, the ability to remain calm in a show cage, etc. but there are other non-genetic factors that also contribute to behavior.

A point worth mentioning is that the expression of some single gene traits, although controlled by a single mutation, can be impacted by a multitude of other genes which interact with the single gene of interest. For an example, I will use the pearl mutation of society finch. The extent of silvering on the bird is very variable. In Japan, the mutation birds are graded, "A" for best, "B" for next best, etc. Simply owning a pearl mutation bird will not guarantee that you obtain high contrast birds with lots of silver in the wings and head. Since there is no way of determining if a normal bird "has what it takes" to enhance the pearl expression in future generations, it is often a long struggle to improve pearl color by randomly mating pearls to normal birds. Another example of a single gene that is likely influenced by other genes is the dilute gene(s) in society finches where a range of color depth within a population is possible for a given single mutation (e.g. dark fawn, medium dark, etc.).

In this article you have become familiar with how genes that regulate quantitative traits are distributed within a randomly mated population. In my next installment of the Breeding for Quality series I will discuss how one can achieve overall improvement of your flock by using selection pressure to manipulate quantitatively inherited traits.

---