STATISTICAL MODELS IN AGRICULTURE: BIOMETRICAL METHODS FOR EVALUATING PHENOTYPIC STABILITY IN PLANT BREEDING

Main Article Content

Daniel Furtado Ferreira
Clarice Garcia Borges Demétrio
Bryan Frederick John Manly
Amauri de Almeida Machado
Roland Vencovsky

Abstract

This paper reviews the main concepts of several methods of phenotypic stability analysis and points out their advantages and limitations. It was concluded that the simple linear regression method of Eberhart & Russel (1966) and the bi segmented regression method of Silva & Barreto (1985) have only historical importance nowadays. Moreover, based on factors discussed in the paper, it is recommended that the regression models of Toler & Burrows (1998) and the Additive Main effects and Multiplicative Interactions (AMMI) model should be used simultaneously to study and to estimate phenotypic stability effects.

Article Details

Section
Article