Vol: 33 Page: 207

Authors:
Wiesław Mądry

Title:
 Statistical analysis of stability measures using genotype x environment data. Part II. Mixed model of Shukla and a joint regression model

Language: Polish

Keywords:
ordinary approximated least estmiators (OALSE)
aproximated F tests
stability analysis
genotype by environment data
MINQUE and REML estimators
mixed model of Shukla
mixed joint regression model of Eberhart-Russel-Shukla
multi-environment trials (MET)

Summary:
The aim of the paper is to present the newest results (including those developed by the author) concerning backgrounds of Shukla’s model (Sh model) and a joint regression model that is known as Eberhart-Russell-Shukla’s model (E-R-S model). The best from among known estimators and tests for stability measures in the both models are given. The main attention is turn to MINQUE and REML estimators of stability variance. Propositions of estimators, using ordinary approximated least squares method, for stability measures in E-R-S model are given. Properties of these estimators are not optimal but in some experimental situations, often found in practice, they are good, close to the optimal. Approximated F tests for study on significance of stability measures are presented.