Research Article

Selection index using the graphical area applied to sugarcane breeding

Published: September 16, 2016
Genet. Mol. Res. 15(3): gmr8711 DOI: 10.4238/gmr.15038711

Abstract

This study aimed to develop a multivariate selection index based on the graphical area of a polygon formed by standardized values, also known as radar chart. This methodology may be used to assist selection of superior genotypes in sugarcane breeding programs. Seven technological traits in 37 sugarcane genotypes were evaluated. An area index (AI) was constructed and the resulting polygon areas were used to rank genotypes under selection. In this study, we propose the use of restricted maximum likelihood to estimate genetic parameters and mixed model equations to predict genotypic and breeding values. The area of each polygon was calculated for phenotypic, genotypic, and estimated breeding values. Thereby, the genotypes with larger area can be selected based on a detailed a posteriori evaluation of the radar charts. The proposed AI can be adjusted based on the breeders’ specific interests, it is perfectly useful in other crops, and may also be applied to studies on genotype-environment interactions. Moreover, AI is a powerful tool that can evaluate trait stability of genotypes based on slight differences in the area formed by each genotype. Hence, this method is easy to apply and shows great potential for use in sugarcane breeding programs as well as in other breeding programs.

This study aimed to develop a multivariate selection index based on the graphical area of a polygon formed by standardized values, also known as radar chart. This methodology may be used to assist selection of superior genotypes in sugarcane breeding programs. Seven technological traits in 37 sugarcane genotypes were evaluated. An area index (AI) was constructed and the resulting polygon areas were used to rank genotypes under selection. In this study, we propose the use of restricted maximum likelihood to estimate genetic parameters and mixed model equations to predict genotypic and breeding values. The area of each polygon was calculated for phenotypic, genotypic, and estimated breeding values. Thereby, the genotypes with larger area can be selected based on a detailed a posteriori evaluation of the radar charts. The proposed AI can be adjusted based on the breeders’ specific interests, it is perfectly useful in other crops, and may also be applied to studies on genotype-environment interactions. Moreover, AI is a powerful tool that can evaluate trait stability of genotypes based on slight differences in the area formed by each genotype. Hence, this method is easy to apply and shows great potential for use in sugarcane breeding programs as well as in other breeding programs.