The specific breeding techniques we use to maintain or improve a color strain vary from breeder to breeder and are strain dependent. Still, the basis for all breeding programs consists of the following three principles.

In breeding
Line breeding
Outcrossing

     Each of the above methods is commonly used by breeders, and can be broken down into varying degrees of complexity.

Inbreeding

     This is the method by which a breeder attempts to purify his strain of guppy, i.e. make the genetics homozygous. This means that all the siblings tend to look like the parents (good matching fish are developed this way).

     When employing this method, one usually breeds brother to sister (sibling-sibling), son or daughter to mother or father (backcrossing), or sibling to aunt/uncle. The best method will depend upon your strain characteristics, (genetics) and the quality of the inbred characteristics they possess. Since sib-sib, and backcross methods tend to purify the strain and set genetic characteristics, you have to be extremely careful that your strain possess mostly wanted characteristics and that the strictest of inbreeding won't concentrate unwanted characteristics as well.

     When setting up breeders, be very careful and observant in your choice of parents. Choose either for size and shape, or for color and shape--whichever favorable genetic characteristics tend to allele together in your strain. Do not use parents with flaws in color and shape when breeding for those particular features! Because size tends to be lost when inbreeding too closely, when breeding for size, do not inbreed too closely. Use sibling to Aunt /Uncle methods within the size line. This method is considered the furthest removed of the inbreeding methods.

     Always use at least two methods in maintaining lines (I'm talking lines, not strains), especially with strains that are pure because you may lose size or good color if you choose the wrong parents and/or inbreeding method.

     Genetics by nature tend to be relatively unstable and constantly undergoing minute to great changes to reflect changes to their environment the improper use of breeders is one of the most influential changes to the stock. In nature breeding parents are the most hardy and fertile of the strain overall, over time. Weaker parents and infertile parents will die off eventually leaving the former to perpetuate the species. The hardy and fertile characteristics are most wanted but they don't usually come with the other desirable characteristics like shape, size and color, most of which we have artificially concentrated in our show lines.

     If your chosen method is leading you astray from your primary objectives, be astute enough in your observations to catch the problem early and make the necessary adjustments to your breeding program. Always stay focused on your goal.

     Do not use inbreeding as the sole method of strain or line maintenance. If your line is straying away from the set objectives, you may be picking the wrong parents. Backcrossing may be necessary to re-inject the desirable genetics lost on one line, but possibly remaining on the other. Keep records (both mental and written) but also be especially observant.

Line breeding

     The inbreeding principle can be used to maintain set lines of a certain color strain. You should always try to maintain at least two lines of the same strain that are related, but bred for different characteristics, so they don't look too much like the same line. You can use the above described breeding methods for maintaining color lines for size, shape, color, fertility, hardiness, deportment, etc. Always keep in mind that the most difficult characteristic to fix or maintain in a strain is not size, as is commonly misunderstood by many hobbyists who inbreed too closely for too long, but shape & color. This includes all aspects of those features related to both body and finnage. Note how important shape is! If you loose a difficult characteristic to set, it could retard your progress many generations.

     The reason for keeping two or more separate, but related, lines is two fold:
You can spread your genetic "egg" characteristics into more than one "basket", which makes it easier to maintain them, and you can cross the lines to blend the related genetics and produce bigger semi-hybrids for show, which may still be useful as breeding stock. Since it is extremely difficult to maintain all desirable characteristics in one fish while removing all unwanted ones (no fish is perfect), this is the safest way to maintain or perform calculated and systematic improvements.

     When crossing these related lines, you can potentially obtain breeding stock from the line cross that can be used to improve one or all of your inbred lines. This is how you maintain inbred lines that have been inbred too long. Maintaining separate lines can potentially improve many aspects of a color strain; for example patterns, symmetry, injection of new background colors in bicolor caudals, increasing the density/intensity of half-black body colors, increasing fertility, tail/dorsal shape & size and fixing better body shape. In strains which cannot compete with show winners, like those with bleeding caudals such as HBAOC, HBYellow, and HBPastels, you can remove the bleeding colors by careful choice of parents, or gradually clean up the strain in incremental steps of progress.

Outcrossing

     Novice hobbyists without the help of an advanced breeder should never attempt outcrossing. As Dr. Alderson has stated, "It's like playing Russian Roulette with your strain". Lines should be inbred for several generations before crossing, and this will vary with the strain. For example, my black strain was not obtained as show fish, but with some problems, and had to be corrected. I started with an outcross to create the new strain. The strain had to be line bred carefully for at least 5-8 generations before I could outcross again or I would surely have ended up with hybrid junk. Lots of fish had to be culled and breeders had to be setup carefully as the strain was new to our fishrooms.

     I have to state here that many of our fine established strains do not and should not need to be outcrossed when you obtain them from a reputable breeder. These strains are usually worked on for many years by that breeder and have the genetics for a good strain already there so need only line breeding methods to maintain for quite some time. Keep records as to the generations and ask the source what you need to do to maintain the fish you obtained from them. They will know best. Don't try to make your own strain by messing up someone else's hard work! As Dr. Chang will always tell novice hobbyists: "Don't go trying to re-invent the wheel".

     Having stated the above, outcrossing can be very interesting and rewarding if properly done. This method is used to "jump start" your strain to produce big show hybrids or introduce new genetic "blood" into your inbred lines. When doing outcrossing one should be aware of some things. The chances of a favorable result usually run 10-20% with calculated crosses at best. You need to use compatible strains to obtain results in F1 or F2 generations.

     Dr. Alderson has already written on compatible crosses so it's not necessary to rehash. Just remember that when outcrossing you must have an objective before choosing strains to cross with. Never stray away from the original strain genetics that you are trying to improve. If the strain is so messed up that you need to deviate appreciably from the original genetic characteristics then it is not worth even working with so go to a different strain. After the initial cross, you should establish within the next 3-6 months whether or not the results are favorable. Look for early signs of favorability like growth rate, activity, color, shape, and size. Then you need to choose the right parents to perform the backcross to the original color strain you were trying to improve. If just going for big show hybrids then your objectives should be met in the first generation (this one). Backcross to both sexes if possible, as you don't know which result will be better. If both backcrosses work out well you can start with fresh new lines to your original strain. Always keep the original strain uncrossed for at least two or more generations during your experiments, in case of disaster.

     You should always attempt to share some of your work with a friend, breeding partner, or club hobbyist. Pass some of your strains on and let them help you to maintain the strain. Inevitably they will follow a slightly different, and interesting, path. Their results will, if they have followed the guidelines above, be as good as yours. Certainly different, possibly even better. It can add some genetic variance (new compatible blood) for enhancing your original strain, or even creating a new strain. You can exchange breeding stock every so often and by having several breeders working with the same strain, you can substantially increase the chances of strain improvement.

     Midge Hill's excellent article, "How To Outcross Guppy Strains," is timeless. It is the standard by which we measure our outcrossing techniques and results. Please refer to it for additional information.