Broodmare Selection

©Voice, February 2007

Genetically, stallions and mares are equal — each contributes a random sample half of their genes to each foal. But, a stallion can sire many foals in a year. Popular stallions may be bred to over 200 mares in a breeding season. The mare, on the other hand, has only one foal a year; or with the use of embryo transfer, she may have two or three foals.

Horse breeding is a numbers game to a degree. A stallion has better odds of having some top foals compared to a mare because he sires more offspring. Since a mare produces fewer foals in a lifetime than a stallion, she has fewer opportunities to become genetically prominent. From this perspective, at least equal emphasis should be placed on mare and stallion selection if one wants to make genetic progress in a herd and also the breed. Outstanding breeders in every breed have superior broodmares herds.

Some mare owners think that breeding Plain Jane to champion Mr. Superman will result in the next world champion. How often have you heard an auctioneer state when selling a filly, “If she does not perform well, she’ll make a great broodmare.” This is a contradiction.
Let’s look at the genetic factors mare owners should consider before making the important decision to breed a mare. This decision impacts not only the breeder’s reputation but also the breed and can have a positive or negative outcome on the economy of the horse industry. Too many below average foals result in over supply and reduce the market value of horses.

Selection
Selection is the process that permits certain horses to produce more offspring than others. The objective in horse breeding is for each generation to be better than the preceding generation. The problem mare owners have is to identify and breed only superior broodmares.
Mare owners have three genetic selection tools to use: phenotypic selection, progeny-test selection and pedigree selection. These aid mare owners in estimating the breeding value of each mare. Unfortunately, many breeders do not use these tool or do not use them correctly, evidenced by all the below average foals born each year as a result of breeding poor-quality mares. It is important that mare owners use all three selection tools when evaluating which mares to place in their breeding herd and, over time, which ones to keep for breeding and which to cull.

Phenotypic Selection
Phenotypic factors are those that can be observed and measured, such as performance and conformation. All horses are bred for some type of performance: trail riding, work cattle, racing, hunter/jumper, the running walk, etc. In some cases, performance is easy to measure: the horse wins the race, earns the most points or earns the most money. Performance measures such as show ring placings are more subjective. Winning at one show is not very conclusive evidence of a horse’s performance ability.
Phenotypic selection is valuable for medium to highly heritable traits. Most performance and conformation traits fall into this category. The more highly heritable a trait, the more accurately the mare’s own performance record will predict her genetic merit. Heritability is the percentage of variation in a trait that can be attributed to genetics and passed on to the next generation.

Too often breeding decisions are based only on personal opinions, such as, “I really like Babe, she is the greatest trail horse.” Yet, Babe may be ridden only once or twice a season on short trail rides. Breeding decisions must be based on objective, measurable factors and not personal whims if they are to be successful and not result in undesirable, below average foals.

Performance events (e.g., cutting, three-day events, barrel racing and major breed capstone shows) have become big business today. Most have major corporate sponsorships and offer several thousand dollars in prize money.

There are several advantages to using selection based on performance. As noted, the current market emphasizes performance. Performance (and conformation) can be measured earlier in a mare’s life than in her foals (progeny). A mare usually will be 9 – 12 years old before she has two foals with performance records.

Most horse genetics research in the world today comes from Europe. One recent study noted that the greatest genetic progress was made by selecting mares based on their performance at 3 to 4 years old. It has been suggested that only 30-40 percent of the top performance-tested mares should be kept for breeding if the greatest genetic progress is to occur.

Conformation also needs to be evaluated, as it relates to both performance and longevity. Halter horses are not as popular today as in the past. More emphasis is placed on what a horse can do. Not evaluating a broodmare’s conformation before she is bred contributes to the number of undesirable foals produced annually.

From a genetic perspective, a mare’s own performance record is nearly as valuable as that of her progeny, and it can be measured much earlier in life. This is extremely important since broodmares have longer reproductive lives compared to other farm females. The King Ranch performance tests its fillies before they are placed in the broodmare band. Testing mares in the same environment provides more valuable information.

Progeny-Test Selection
In selecting mares to breed, it makes sense that they should have produced performance foals. If your objective is to win at the Walking Horse Celebration, select a mare that has produced one or more champions. Checking performance records of a mare’s foals is a good method for determining her genetic merit.

However, progeny testing cannot be used on young fillies or mares. Mares will be 9-12 years old before having one or two foals old enough to perform as 2-year-olds. This fact works against the progeny-test selection process because once mares reach about 16 years of age, their reproductive ability declines.

The use of embryo transfers and the fact that some major horse breeds allow more than one foal to be registered from a mare annually places greater emphasis on proven matriarchs.

Selection based on the performance of a mare’s progeny is more accurate as it eliminates environmental effects on the mares, but requires more time for the foals to reach performance age.

Pedigree Selection
Pedigrees are an important part of the total selection process. Unfortunately, some breeders base their selection entirely on pedigrees, disregarding performance, conformation and progeny’s performance. A good pedigree is not a license that a mare is good enough to be bred. As noted above, breeders should use all three selection tools to make intelligent breeding decisions. Pedigree selection is most valuable in evaluating young horses for traits that have not yet been expressed, such as performance. Pedigrees are also important in marketing horses, especially weanlings and yearlings.

Also, the use of pedigrees is not always understood. For example, saying that a mare is a great-great-greatgranddaughter of some truly outstanding stallion is really meaningless. Only 6.25 percent of the mare’s genes are from this outstanding individual. What is known about the other 93.75 percent of the inheritance? Using this mare as a broodmare will result in her foal having only 3.125 percent of its genes from this fabulous stallion while the other 96.875 percent may be from inferior ancestors.

If there are no well-known or high performance sires or dams on both sides of the pedigree within the first two generations, a mare is not a good candidate for breeding.

Multiple Trait Selection Systems
Most breeders want to improve more than one trait at a time. As the number of traits increases, less genetic progress will be made in any one trait so the number of traits used in selection should be kept to a reasonable level. A breeder must determine which traits are of greatest value and the importance of each trait relative to long range breeding goals. Selection index and independent culling levels are two systems that a breeder can use when selecting for several traits at a time.

A selection index system requires that the breeder place a value, or weight, on each trait deemed important and calculate an overall score for the breeding merit of each mare. The highest scoring mare(s) will be used. In the following example, an owner uses a selection index system to decide which two Tennessee Walking Horse fillies to show and keep as future broodmares. See Table I.

The owner has determined that action of gait, training ability, soundness and conformation are the major traits for selection. The traits are weighted as follows: gait = 8, training ability = 8, soundness = 6 and conformation = 4. All fillies have acceptable pedigrees.

Based on the selection index, the breeder would choose fillies 2 and 3. Filly 1 is lacking in most traits. Filly 4 has a great gait but is difficult to train.

Another type of selection approach is the use of independent culling levels. Here a minimum acceptable standard is established for each trait. If an animal does not meet the minimum for any one trait, it is culled.
Cutting horses are used in this example. See Table II. The breeder has established a minimum level of $10,000 earned by the mare in cutting, and her foal(s) must have earnings of $35,000. The mare must have a conformation score of 7 or above, not be taller than 14.3 hands and have a disposition score of at least a 6.

Mare A is culled because her foals have not met the minimum earnings. Mare B is culled for a low disposition score, and mare C is too tall. Of the four mares, Mare D would be kept or added to the breeding herd as she meets all the minimum standards. This example illustrates a problem with the independent culling level: an animal that does not meet the minimum for any one trait is culled. For example, Mare C’s foals have done well in competition, earning $65,000; she has good conformation (8) and disposition (8); but she is culled for being too tall. Breeding her to a 14.2-hand stallion could result in foal(s) that meet the standard of 14.3 hands.
If properly constructed, the selection index (Table 1) is more efficient than the independent culling level approach (Table II) and should yield more genetic improvement. An independent culling level with logical traits and realistic minimum levels can aid the breeder in initially selecting mares for the breeding herd or in selecting replacement mares.

Breeders need to objectively evaluate each mare in the herd annually. This evaluation should be based on genetic information. The following example uses the performance of the foals and the mare’s performance, conformation and pedigree. Each trait has been assigned a weight (noted in parentheses) and each trait is scored from 1 to 10. The objective is to rank the mares in the herd, culling the lowest scoring mare and replacing her with a higher scoring broodmare. The result will be a continual improvement in the overall genetic or breeding value of the herd. This calculation is similar to a rolling dairy herd average. See Table III.
The average score for this herd is 68.75. Mare B has the lowest score of 58.5. When mare B is culled, the herd average becomes 70.8. If mare B is replaced by a mare with a score of 72, the herd average goes up to 71. While these changes may appear small (68.75 to 70.8 or 71), genetic progress is permanent and will be passed on to the next generation of foals.

Summary
Before a mare is bred, breeders have an obligation to themselves, the breed and the horse industry to use proven methods to determine whether a mare is truly worthy to be a broodmare.

–Dr. Frederick Harper
Extension Horse Specialist
Department of Animal Science
University of Tennessee

References
Harper, F. and B. Bell. 1991. Genetics for the Horse Breeder (Part I). Tennessee Horse Express. 10:1. Agricultural Extension Service. University of Tennessee, Knoxville.

Harper, F. and B. Bell. 1991. Genetics for the Horse Breeder (Part II). Tennessee Horse Express. 10:2. Agricultural Extension Service. University of Tennessee, Knoxville.

Jones, W. E. 1982. Genetics and Horse Breeding. Lea & Febiger, Philadelphia. Equine Genetics and Selection Procedures. 1978. Equine Research, Inc. Carrollton, TX.

Viklund, A., J Philipsson, A. Wikstrom, Th. Arnason, E. Thoren, A. Nasholm, E. Strandberg and I. Fredricson. 2005. Testing Young Swedish Riding Horses for Sport and for Genetic

Evaluations. 56th Annual Meeting of EAAP. Uppsala.