Biometrical Methods for Analysis of Multi-Harvest Forage Breeding Trials
Genotype by harvest interaction; Forage breeding; Genomic selection; Spatial analysis
Perennial forage grasses and legumes are the principal source of nutrition for most ruminant livestock in developing countries. Raising yields of forage crops can increase the availability and affordability of livestock products. The forage breeding has an important role in the increasing of forage yield. However, the genetic gains for perennial forages are small when compared to annual crops. The lower genetic gains for perennial forage is mostly due to the fact that trials should be evaluated under several harvests seasons. Therefore, the use of proper statistical analysis to deal with longitudinal data set and spatial variation in field trials are crucial to select the best genotypes more accurately. Furthermore, the use of genomic selection into forage breeding programs can accelerate the breeding cycle, consequently increasing the genetic gain over time. This study was divided into three chapter. The first chapter, we evaluated the use of spatial analysis in three forage species breeding trials (Medicago sativa, Panicum maximum and Urochloa spp.). In the second chapter we investigated the use of random regression and factor analytic mixed models in dealing with longitudinal data set generated in forage breeding trials. In the last chapter we performed genomic selection in alfalfa by incorporating environmic-based data.