Epsilon Archive for Student Projects

Abstract

The Swedish Warmblood horse (SWB) is a sport horse bred for dressage and jumping with a breeding goal to breed a sound and correct horse with a competitive temperament, excellent gaits and/or jumping ability. The selection of breeding stallions is the most important part of the breeding programme as the stallions get a lot more offspring and contribute more to the genetic progress than the mares. Many of the European breeding associations for sport horses have similar breeding goals and several stallions are used in more than one country and also have breeding values estimated in more than one country. The use of imported semen for SWB mares have increased a lot the last 20 years and the majority of the mares are covered by a stallion that originate from another country. In addition, a lot of horses are imported for competition and the Swedish Warmblood association have had a suspicion that these horses are preselected on their expected talent and that this preselection lead to biased breeding values of their sires. Breeding stallions themselves get better training from better riders in comparison to other groups of horses and that may also lead to biased breeding values. The aim of this thesis was to investigate how the stallions’ breeding values are affected by (1) the imported offspring and (2) their own results from competition and young horse tests. Data was provided from the Swedish Warmblood association and consisted of the pedigree and result files, including results from young horse tests and competition, used in routine genetic evaluation 2015. For each horse, three different breeding values were estimated using the DMU-package and a multi-trait BLUP animal model, in each of the two parts of the study. They were estimated with either different information going in to the evaluation or with different models used for the evaluation. The breeding values for stallions with at least 15 offspring evaluated at young horse test were compared. The results showed that the breeding values changed between the different estimations depending on the information or model used but there were still high correlations (calculated with Spearman rank correlation) between the different breeding values. There were also a large percentage of stallions co-selected among the top 40 between the different breeding value estimations. The first part of the study concluded that the imported horses do affect the breeding values and the values may be biased in advantage for stallions with a large proportion of imported offspring. Though there is not enough information to conclude how to handle the possible bias and the study should be performed again when more imported horses are registered in Sweden; as it is mandatory from 2016 to register every horse that will be in the country for more than three months. The second part of the study concluded that results from competition and young horse test both affect the breeding values to a great extent. In the routine estimation of breeding values today only the competition results are adjusted for by moving them to a dummy stallion and thereby giving them half of the effect that they would have if the real stallions kept them during the breeding value estimation. The dummy stallion is a fictive twin to the real stallion created in the data system only to be used in the breeding value estimation. This study suggests that the breeding values should be adjusted for – or not adjusted for – results both from competition and young horse test by moving none of them or both to the dummy stallion in the estimation.