Biometrical methods for analysis of multi-harvest forage (Urochloa spp., Panicum maximum and Medicago sativa) breeding trials

Abstract

This study focuses on the optimization of statistical methods in forage breeding trials, with a goal to improve the efficiency of the breeding process and increase the rate of genetic gain. The study is conducted on three forage species: Medicago sativa, Panicum maximum, and Urochloa spp. The first chapter of the study evaluates the use of spatial analysis in breeding trials, taking into consideration the spatial variation and correlations within a trial and between repeated measurements. The results of this chapter provide insight into the effectiveness of spatial analysis in forage breeding trials. The second chapter of the study focuses on the application of random regression and factor analytic mixed models to deal with longitudinal data generated in forage breeding trials. These models account for temporal correlations between repeated measurements and allow for the appropriate modeling of genetic effects over time. The results of this chapter highlight the usefulness of these methods in analyzing data from forage breeding trials. In the final chapter of the study, genomic selection is performed in alfalfa, incorporating enviromic-based data. This chapter highlights the potential of genomic selection in reducing breeding cycles and increasing the rate of genetic gain in perennial forage species. The results of this study provide valuable information for forage breeders and plant breeders in general, regarding the use of various statistical methods in breeding trials, and their potential impact on the efficiency of the breeding process and the rate of genetic gain. The findings of this study have the potential to contribute to the improvement of forage production, which is crucial for the supply of nutrient-dense food such as meat and milk, particularly in developing countries where forages are a primary source of nutrition for most ruminant livestock.