http://www.exoticfinches.com

Warning: All rights reserved. This article appeared in the issue of the National Finch and Softbill Society Bulletin. Vol. 22. No. 3. May/June. 2005. (p. 8-12). 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

The previous installments in this series attempted to give you enough information to learn the differences in modes of inheritance for different traits. Remember, most of the traits that lead to good conformation and size are not controlled by single genes but rather by a multitude of genes that act in concert or act to modify the major genes involved. If you want your birds to have rounder heads, smaller beaks, deeper colors, more vertical stance, wings that don't droop, large cobby chests, large size, big eyes, etc., you need to realize these are not traits controlled by single genes. Rather, these are quantitatively inherited. Single gene mutations are more common for color mutations and we become so accustom to understanding segregation ratios for color we make the false assumption that we can quickly recover all desired traits in nice clean ratios. Unfortunately, it takes many generations of intelligent pairing to achieve results with traits that are quantitatively inherited.

In this, my last in the "Breeding for Quality" series, I will compare strategies that can help a breeder reach his/her final goal - to obtain a uniform flock of superior birds. I'll begin by saying that random matings within a non-selected population rarely improve a flock if you goal is, in essence, to eliminate certain phenotypes and achieve a population of uniformly excellent birds. There is no one correct way of approaching population improvement but with certain guidelines the art of breeding has elements of science that can greatly assist you in reaching your goal.

There are several strategies for selecting birds to pair. Assortative matings are those where matches are made with a purpose, rather than random. Breeders classify assortative matings as either negative or positive.

With negative assortative mating, opposites are paired (e.g. a cobby bird to a leaner bird). It is important to note that we do not mean that the worst bird (e.g. the leanest) is mated to the cobbiest bird. Rather, among the birds of choice for other reasons (e.g. good color depth, nice wing carriage, etc.), we are mating the birds with opposite extremes for cobby or lean so that we can avoid generating extremes in body types in future generations.

Negative assortative mating tends to reduce the extremes in a large diverse population. This may or may not be good depending on the goal. For example, if optimum size is not yet achieved, this strategy is not useful for leaping forward in size. You would not leap forward in size mating smallest to largest. To leap forward in size one would need to mate the largest to the laragest. However, if all your big birds also have a flat head it would be wise to use negative assertive mating to try to correct head shape and not concentrate solely on bird size.

Negative assortative mating is a great strategy for fine tuning a fairly uniform population of birds. For example, if the only thing that bothers you about your flock is that many birds are too lean and many are too cobby, mating cobby to lean should even off the flock over a few generations.

With positive assortative matings similar individuals are paired (e.g. large bird to large bird). This tends to create "extreme" populations and flattens the normal distribution curve yielding more individuals than usual on each end of the distribution. Once the optimal trait is achieved (e.g. large size, cobby body, etc.) positive assortative mating is not recommended.

It is possible to combine both negative and positive assortative mating. For example, if optimum size is nearly obtained, choosing two birds of good size is a positive asortative mating. But if you have a choice within the population of good sized birds and you feel another trait is present in extreme forms (e.g. beaks too long and beaks too small) choosing one large bird with a long beak and another large bird with a small beak (negative assortative mating) should lead you quickly toward large birds with appropriately sized beaks. In my personal experience this strategy has given me a larger number of decent birds so when show time comes around I do not need to hope that one good bird is in condition for the show.

An actual example of how improper breeding strategy results in dead-end breeding might be helpful here. Long before the appearance of black brown society finches (these are of hybrid origin), I sought out someone with really dark Lonchura striata domestica. I obtained three pairs of birds of uniform extremely dark body color from one breeder. Unfortunately, they all had somewhat droopy wings. My guess is that the breeding selection program that resulted in dark body color was one where color was the only thing chosen, i.e. a positive assortative mating for dark color. Since wing deportment was uniformly bad, I had a feeling it would be difficult to correct. My prediction came true. I was never able to correct wing deportment without outcrossing to another line. Outcrossing resulted in a huge loss of color depth (a quantitatively inherited trait) and it was looking like it would be years of work to regain that color depth. I eventually gave up on the line and learned a hard lesson - anyone who ignores the entire package when selecting birds will wind up with birds with flaws that are difficult, if not impossible, to correct.

If one chooses to improve one trait until it is acceptable and then move to another trait, this method can be called tandem selection. For example, one might select for bird size for several generations until the desired size was reached and then select for egg size hoping that egg size will be correlated to adult size for future early culling. There are problems associated with tandem selection. First, you need to decide when to stop selecting for trait one. Is a 20 grams zebra finch big enough? Or do you try to push the weight to 25 grams before starting to cull birds for low egg size? In addition, this type of selection works best if the two traits are correlated. If however, the first trait is not related in some way to the next trait, there is a great risk of tapping out the necessary variability for the second trait, making advances difficult if not impossible. While tandem selection may yield the quickest advances for one trait, unless this is a trait so desirable that one is willing to risk overall improvement, it is not recommended.

A selection method for multiple-trait selection (i.e. improving more than one trait) is known as independent culling level. This method sets a minimum standard for more than one trait and that standard must be met for all of these traits before a bird is kept. With this method a breeder sets an acceptable level for one trait (e.g. 20 gram adult weight in zebra finch). If this weight is not met, the bird is culled even if its other traits rank well above a minimum selection value (e.g. it is a black face with a very dark belly). This method is attractive because the principles are simple and there is no chronological limitation of selection. For example, if the richest colored babies are first selected and the lighter ones culled, one can select the best marked birds from the remaining birds in that clutch. Therefore, you now have the richest color and best markings.

The downside to multiple-trait selection scheme is that we cannot be as "picky" with respect to the value of any given trait. This means that progress improving each trait is slowed as we head toward overall acceptability one small step at a time. If traits are favorably correlated, progress moves more rapidly. Unfortunately, this if often not the case and one must accept that someone else using tandem selection may end up with, e.g., larger birds more quickly. Yet, these large birds may fall short in several other traits. We see that one must set a value to each trait and guide one's breeding program accordingly. Wasting time trying to get the perfect toe length may sacrifice advances in conformation and feather quality, two traits of higher value for show breeders.

By now it should be no surprise that improvements in quantitatively inherited traits can take time. I cannot overestimate the importance of starting out with the best birds you can obtain. Birds that have been in domestication for a long period of time have had the opportunity to be slowly improved by a good breeder. It can take many generations to achieve a goal and it is foolish not to capitalize on the work of others and make improvements in good stock rather than start out ten generations behind the best available lines. Obtaining one great bird, loaded up with desirable traits can infuse a huge dose of 'good genes' into your gene pool. It is perfectly fine to obtain one great male and have him be the father of your entire flock if you have some unrelated hens to pair with him. This strategy has worked wonders for me with my society finches and more recently with my zebra finches.

I've often been asked what to do if one obtains two lines of birds of very good quality from different breeders. Should they be mixed? It depends. If the "average" phenotype of the two lines is something you think is better than the individual lines, then by all means mix. But keep something in mind. If both lines are somewhat inbred (very likely) then the first generation tends to be uniform and most likely of intermediate phenotype. However, maintaining this intermediate may be more difficult. Let's use dogs as an example. There is a very popular pet dog being sold that is the result of the crossing of Labrador Retriever and Standard Poodle. They are known as "Labradoodles" by owners and followers of the pet dog trade. They are not a new breed. They are a first generation hybrid of two breeds and technically, therefore, they are mutts. Since the parent breeds are inbred enough to breed uniformly typical offspring, the hybrids are intermediate and have good traits from both breeds (while losing some of the negative traits of the inbreds.). However, if one were to mate Labradoodles (which are highly herterozygous by nature) to each other, segregation of many traits would take place with unpredictable results. The most consistent way to make uniform Labradoodles would be to maintain Labrador Retriever and Standard Poodle lines and cross them every time you wanted the Labradoodle. This is what is being done with these dogs. Theoretically and with enough patience, however, you could cross Labradoodles to each other and eventually fix the phenotype resulting in a new breed that breeds "true". In fact, most new breeds of dogs do not start out with mutts or wolves but rather with two inbred lines. The above dog discussion holds true for finches. If you want to make a breeding line that is the average of two others it can be done with time but the numbers of birds you may need to produce may be high.

Another concern I often hear is that inbreeding may eventually result in offspring that are worse with each generation. We often read about the problems associated with inbreeding. The degree of inbreeding is defined as the probability that identical homozygocity occurs at a locus (e.g aabbccDdeeffGGHh is more homozygous than AaBbCCDdEeFfGgHh). The term "inbreeding depression" means that a general loss of vigor and fertility occurs as we inbreed and inbreed. This probably occurs because deleterious recessive genes are no longer masked by dominant genes and they express their negative trait. In the United States inbreeding has such a stigma that most breeders go to the opposite extreme which often results in increasing the number of generations it takes to get the desired result or never attaining the desired result. In order to avoid unnecessary inbreeding but still make rapid improvements, line breeding is recommended. Line breeding is a conservative form of inbreeding which results in slow but steady improvement and limits the risk of producing weaker individuals. Distantly related relatives (e.g. uncle and niece) are mated with sister-brother or father-daughter pairings avoided. The main purpose of line breeding is to transmit a large percentage of an outstanding ancestor's genes into future generation without causing an increase in the frequency of undesirable traits. A tricky balance but very doable.

To my knowledge, no one has calculated the level of inbreeding tolerated by different finch and softbill species, although this has been done in dogs, cattle, and other valuable domestic species. If you feel that you want to calculate the level of inbreeding (i.e. the amount of 'shared' blood between two potential mates) there are many software programs that make it easy. I am not recommending any particular software program but should you choose to buy software to keep pedigrees and calculate the level of inbreeding that would result from your pairings, the following companies offer such a service. Keep in mind that most were developed for dogs or livestock but will work just as well with birds. I found that those specific to birds were sorely lacking genetic analysis and were more for keeping track of dates of birth, etc. The following links were active at the time my article was printed.

http://www.avimate.com/

http://www.tenset.co.uk/

http://www.peds.co.uk/kcbs4a.htm

http://animalregister.net/index.php

If you are up for calculating inbreeding rates yourself, here is a web page that shows you how.

http://www.xprt.net/~pgrogers/Wrightf.html

I would also suggest looking up pedigrees for champion horses or dogs, where the science of breeding for quality has been around a long time. Studying long pedigrees shows you the relatedness of individuals and can give you a good idea of the level of inbreeding used. I am not assuming that the level of inbreeding that is tolerable is the same for dogs and birds, but it is a good start to look at canine pedigrees to see which pairings are considered acceptable and which strategy has been used.

I hope that my series on Breeding for Quality will help you take a more intelligent approach to pairing you birds. In order to make big gains in your flock's quality I will offer some final suggestions:

- Know what traits are controlled by a single gene and what traits are inherited quantitatively

- Keep pedigrees so that you can tract down the birds that are responsible for your greatest advances and your worst declines

- post a standard silhouette for your species and compare birds you are thinking of saving to this standard

- Learn how to line breed and use inbreeding carefully--but use it

- Learn the standards for your species and mutations

- Start out by obtaining the best stock

- Remember, a single terrific bird can go a long way to improving your flock

- Do not obtain a line of birds that is nice, but has one trait that is poor within the entire line

- Balance faults when selecting pairs - in other words never mate two birds with the same fault.

- Breed as many birds as you can handle and care for properly but keep as few as you need to maintain some genetic diversity. Judicious and harsh selection will get you a long way.

- Don't be afraid to sell whole clutches if none of them are better than the parents

- Be patient. It takes many generations to increase the proportion of winners in a flock

- Bring extra birds to shows and ask a judge to tell you which traits need improvement

- Ask judges questions at shows. Find out why the best bird was the best bird and why the others were not in the top ten

The best of luck to all of you. See you at the show bench!

Reference:

Bourdon, R.M. 2000. Understanding Animal Breeding. 2nd edition. Prentice Hall, Upper Saddle River, New Jersey.

---