Integrated genetic evaluation of breeding field test traits, competition traits and test status in Icelandic horses

The main goal in breeding the Icelandic horse is to produce aesthetically appealing and capable riding horses with five gaits and good spirit, suited both for leisure and competition. Selection of Icelandic breeding horses is based on breeding values calculated from the breeding field test records. It has been speculated that assessed horses are not a random sample of the population, a situation that would lead to bias in estimated breeding values and retards genetic improvement. The main aim of this thesis was to study integrated genetic evaluation of competition traits and breeding field test traits for Icelandic horses where the effects of preselection in the data were accounted for. Genetic parameters of the competition traits, and of genetic relationships between competition traits and the breeding field test traits were analysed. The breeding field test data included individual records of Icelandic horses evaluated in 11 countries. The competition data included records of horses that had competed in Iceland and Sweden. The competition traits and breeding field test traits were analysed using linear animal models. The competition traits analysed were both original and combined ones covering the competition aptitude of four-gait, five-gait, toelt and pace. The combined traits were formed in order to describe the competition traits in a simpler manner. The estimated heritabilities were low to moderately high for all competition traits and the genetic correlations estimated among competition traits were generally strong and favourable with few exceptions. Moderate genetic correlations were estimated between most of the competition traits and some of the conformation traits assessed at breeding field tests. High genetic correlations were generally estimated between the competition traits and most of the riding ability traits recorded in breeding field tests. Competition traits were concluded to be suitable for genetic selection. Breeding field test traits and test status, an all-or-none trait describing attendance at breeding field tests, were genetically analysed using bivariate animal and sire linearthreshold models. The presence of preselection in breeding field test data was verified by the estimated genetic parameters for test status: this trait had a significant genetic component and related strongly to the breeding field test traits, especially those that had high weight in the selection index. This applied to both conformation and riding ability traits, although breeders seem to pre-select horses to attend breeding field tests more strongly on good riding qualities than on aesthetic conformation. The emphasis on riding ability as the criterion for preselection was further supported by a larger increase in estimated genetic parameters for these traits when analysed simultaneously with test status. Environmental covariances between test status and breeding field test traits are not estimable as all individuals with test status equal to zero lack phenotypic values on the tested traits. The effect of assuming zero environmental covariances between test status and the breeding field test traits was studied in a simulation. It did not lead to serious bias in the estimated genetic parameters unless the non-estimable true environmental correlation deviated largely from zero. Moderately biased genetic parameters had only relatively small effects on the genetic evaluation. Estimation of breeding values where test status was included in the model always led to improvement in genetic progress and in higher correlations between true and estimated breeding values. The benefits from integrating test status, competition traits or both into current genetic evaluation based on breeding field test traits were estimated. In general there were trivial differences between models. Estimated breeding values were largely unbiased for all traits, although models including test status showed minor deviances from this. The current breeding evaluation system seems therefore to be well established. However, breeding values were more accurately estimated with the inclusion of the new traits, especially the test status. The ranking of sires, based on combined indexes, changed (between 10% and 20%) with inclusion of the competition traits and test status. This indicates how large the effect of adding new traits to the genetic evaluation will have on the selection of sires. Integration of test status and competition traits increases the accuracy of the genetic evaluation and influences the ranking of sires, and therefore selection of sires for future breeding. Genetic parameters for the test status trait should be re-estimated where the trait is re-defined for all horses and all traits including both competition and breeding field test data. Competition traits can be included directly in the genetic evaluation and will give breeders an effective tool on which to base their selection.

Author dc:creator

• Elsa Albertsdóttir 1975-

Contributors dc:contributor

• Landbúnaðarháskóli Íslands

• Hestar
• Kynbótamat
• Kynbótadómar
• Gangtegundir
• Prófanir
• Keppnisíþróttir

Language dc:language.iso

en

Handle dc:identifier.uri

http://hdl.handle.net/1946/10835

OAI identifier oai:identifier

oai:skemman.is:1946/10835