Abstract

RUEDA RESTREPO, Jairo Alberto  and  COTES TORRES, José Miguel. EVALUATION OF TWO METHODS OF PHENOTYPIC STABILITY THROUGH CROSS-VALIDATION. Rev. Fac. Nac. Agron. Medellín [online]. 2009, vol.62, n.2, pp.5111-5123. ISSN 0304-2847.

One of the most important topics of plant breeders is to evaluate the phenotypic stability through regional trials or multi-environment trials. There are many methods proposed to analyze those trials and to estimate the phenotypic stability. This paper compares the regression method proposed by Eberhart and Russell and the components of variance proposed by Shukla, according to a cross-validation methodology. Data from 20 multi-environment corn tests, each one with nine genotypes, planted under a randomized complete block design with four replications, were used. It was found that the best model to predict the future performance of a genotype is the method of Eberhart and Russell, showing a root square value of the prediction medium 2,21% less than Shukla's method , which a prediction consistence of 90.6%.

Keywords : Genotype-environment interaction; phenotypic adaptability; Shukla's variance; regional trials.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )