Premium
Estimation of Additive and Nonadditive Genetic Components of Sugarcane Families Using Multitrait Analysis
Author(s) -
Pisaroglo de Carvalho Melissa,
Gezan Salvador A.,
Peternelli Luiz Alexandre,
Pereira Barbosa Marcio Henrique
Publication year - 2014
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj2013.0247
Subject(s) - stalk , saccharum , heritability , biplot , principal component analysis , brix , cane , biology , genetic correlation , hybrid , sugar , genetic gain , agronomy , restricted maximum likelihood , breeding program , cultivar , mathematics , horticulture , statistics , genetic variation , genotype , food science , genetics , maximum likelihood , gene
Sugarcane ( Saccharum spp. interspecific hybrids) cultivars are a polyploidy and heterozygous complex of genotypes with a large amount of genotypic variability. This study evaluated several commercial traits from a series of trials to provide a better understanding of the genetic structure of this population with the aim of identifying future breeding strategies. Data were analyzed using univariate and bivariate methods on single‐ and multiple‐site linear mixed models. In addition, the top‐yielding families using the best linear unbiased prediction were selected and further studied by performing a principal component analysis to facilitate understanding of their morphological characteristics. The traits evaluated were: stalk number, stalk weight, stalk diameter and height, juice percentage of soluble solids (Brix), cane yield, and sugar yield. The results of the single‐site analyses showed that stalk height, stalk diameter, and Brix were the most important variables because of the higher potential gains in yield based on their heritability values. In the multiple‐site analyses, relevant nonadditive effects (i.e., dominance) were detected for the variables cane yield, sugar yield, stalk weight, stalk diameter, and stalk height. The selection of the top‐yielding families by principal component analysis produced two groups of diverging morphological characteristics or ideotypes. Finally, bivariate analyses showed, among others, a high additive genetic correlation between stalk number and stalk weight.