Selection Methods

Selection Methods.

Selection for a single trait is the quickest way to make progress in that individual trait. However, because animals and the economics of livestock production are complex, this simple method is generally not practical. Correlated responses between traits can also cause problems in this type of selection. For example, if we simply selected for individual weaning weight alone, we would most often select kids born and raised as singles, reducing fertility and twinning rate in the herd.

Multiple trait selection is more practical and common today; however, many producers don’t realize they are doing this. When considering multiple trait selection, it is important to keep the number of traits as low as possible. The more traits being selected means that less progress is made in any one trait. There are two primary methods used for multiple trait selection: selection index and independent culling levels.

Selection indexes have been considered and utilized for many years in different forms. The use of an index requires the calculation of economic values or “weights” for each trait and then putting the individual’s performance into an equation. The result of that equation is an index value for the animal and that index is then ranked to identify the best animals for the combination of traits included in the index. The problem with this system is that economic values change continuously, and values for all traits must be collected before selection can occur. Because not all desired traits can be measured at one time, this requires keeping animals longer to collect all the necessary data.

Independent culling levels are a more common and practical method of selecting breeding stock. This method requires the producer to determine the number of animals they wish to keep at each selection interval (i.e., birth for birth type, weaning for growth to weaning, yearling or breeding time for total growth). The primary problem with this method is that a producer may eliminate an individual that is excellent in one trait due to poor performance in another. This is generally not a large problem if the number of traits is kept low, generally two to three traits. It is more practical than selection indexes because it does not require the use of complex equations. Decisions can be made early because complete data on an individual is not needed.

An example of independent culling levels conducted at birth would be to identify and not select replacement does from single births. Then, at weaning, select the heaviest doe kids, out of multiple births, to keep as replacements. A few extra (five or six head) could be kept in case they do not breed as desired. At breeding, expose all does and only keep as final replacements those does that bred (by marking or pregnancy check) and required the fewest treatments for internal parasites after the breeding season. The final replacements should be more fertile and more resistant to internal parasites than the previous population. At this point, you have selected twice on fertility and selected for growth rate to weaning and improved parasite resistance. The end result should be an increased twinning rate and heaver kids at weaning that require fewer treatments for parasites over time. To make the same selection using an index, you would need to keep all doe kids through breeding and then calculate the index.

Goat producers can improve production and health of their animals through selection. To do this, producers need to start keeping and utilizing performance records. Fertility is the most important economical trait generally followed by growth. For goats we need to also consider health traits especially resistance to parasites. The use of multi-trait selection is important but producers should not consider too many traits at once. Independent culling levels is the most practical way to select for multiple traits at one time without expected progeny differences being available. Comparisons between animals should be done under similar production practices to reduce differences between environments.