Combination of mixed linear model approach with selection indices in kale breeding programs

Abstract

Utilizing selection indices is an effective strategy for the simultaneous evaluation of multiple traits in kale breeding programs. This approach allows for the selection of kale genotypes that exhibit enhanced productivity and adaptability by combining desirable attributes for the crop. In this study, we employed a mixed model approach in combination with various selection indices to estimate selection gains and recommend the most suitable index for kale breeding. The experiment was conducted at the Center of Development and Technology Transfer, Federal University of Lavras, Ijaci, MG. Thirty-four experimental genotypes were assessed in a randomized block design with three replicates, featuring four plants per plot. We evaluated several traits, including total leaf yield, number of leaves, average leaf mass, number of sprouts and chlorophyll content. Data analysis was performed at both the plot average level and the average quantity of the five harvests. Statistical analysis of mixed models confirmed the presence of genetic variability among kale genotypes. We examined the Smith and Hazel, Mulamba and Mock, Z-index, and FAI-BLUP indices. Smith and Hazel, Mulamba and Mock, as well as Z-index, were found unsuitable for leafy kale selection in breeding programs. The FAI-BLUP index demonstrated superior performance, aligning with the specific objectives of the kale breeding program and offering desirable gains. Therefore, we recommend the use of the FAI-BLUP index in kale breeding programs.

Keywords

• Brassica oleracea sub acephala L.; biometry; genetic improvement; multivariate analysis; REML/BLUP.