Communauté Française De Belgique Université De Liège – Gembloux Agro-bio Tech Improvement of Genetic Evaluation Systems for Maternally Influenced Traits and Multi-breed Livestock Populations

Animal breeding programs are designed to genetically improve livestock populations over many generations to enhance farm sustainability and competitiveness. Genetic improvement is achieved by selecting genetically superior animals, based on estimated breeding values (EBV), to be the parents of the next generation. These EBV are calculated by solving mixed model equations characterizing appropriate statistical genetic evaluation models. To guarantee effective genetic selection, genetic evaluation models must be tailored to the specific characteristics of the traits and population under evaluation. This PhD thesis focused on the development of genetic evaluation models suitable for categorical maternally influenced traits and for multi-breed populations. Appropriate genetic animal models were developed and assessed: (1) for two categorical maternally influenced traits based on calving ease scores from Walloon Holstein dairy cattle and on lamb survival data from a New Zealand sheep population; (2) for two multi-breed populations based on milk yield records of New Zealand purebred and crossbred dairy cattle, and on purebred and crossbred calving ease scores from Walloon Belgian Blue and Holstein cattle. Results showed that (1) fitting maternal effects was required to avoid biasing the EBV, and there was no clear advantage in using non-linear mixed models instead of linear mixed models for the genetic analysis of the two categorical maternal traits studied; (2) breed-dependent EBV could be estimated using the proposed multi-breed models, and that combining purebred and crossbred data had a positive influence on the accuracy of the EBV of purebred animals. Finally, part of the research presented in this thesis contributed to the development of the genetic evaluation systems currently used in Walloon Region of Belgium and in New Zealand. VANDERICK SYLVIE. (2017). AMÉLIORATION DES SYSTÈMES D’ÉVALUATION GÉNÉTIQUE POUR DES CARACTÈRES MATERNELS ET DES POPULATIONS MULTI-RACES D’ANIMAUX D’ÉLEVAGE. (THÈSE DE DOCTORAT EN ANGLAIS). GEMBLOUX, BELGIQUE, GEMBLOUX AGRO-BIO TECH – UNIVERSITÉ DE LIÈGE, 223P., 29 TABL., 7 FIG.