Use of genetic, genomic and phenomic approaches to improve livestock performances

Traditional methods to improve animal productions are derived from the quantitative genetics theory. The estimation of variance components, genetic parameters and genetic gains are essential steps of the selection process based on quantitative methods. The use of linear mixed models, algorithms for their solutions and computer to solve complex system of equations has made possible to estimate reliable genetic merits also from complex structures of relationships. More recently the high throughput technologies has provided a great increase of information both at genomic than phenomic level. This high informativeness has led two main consequences: improved accuracies and faster genetic gains from one hand, big data handling and unbalancing from the other hand. The present thesis will deal with these aspects. Two experimental contributions will focus on the estimation of genetic parameters for new criteria of selection on species of particular interest, with traditional and innovative phenotyping methods. A third contribution will deal with the research of new morphometric indicators that could have an economic impact if applied for the selection of the species of interest. Finally, a fourth contribution will address the issue of the data unbalancing derived from the current information gap generated by the speed of genotyping and the capacity of the phenotyping practices. All these contributions will have the same aim, the improvement of livestock performances.