Evaluation of Maize Varieties through Data Analysis of Multi-Environment Trials: Application of Multiplicative Mixed Models

Abstract

Ethiopia is a significant maize producer in Africa Over the last two decades, Ethiopia's maize sector has undergone significant transformation. Farmers in Ethiopia require a consistent supply of new and improved varieties to meet their ever-changing production and marketing challenges. Breeders can no longer function without multi-environment trials (MET) analysis for varietal evaluation. To accurately select superior varieties that contribute to agricultural productivity, efficient statistical methods for maize variety evaluation must be used. The goal of this study was to identify better maize varieties based on yield performance by analyzing data from multi-environment trials using multiplicative mixed models. In this study, 32 maize varieties, including four checks, were sown across seven major maize growing areas in Ethiopia using RCB design, with three replications during the main cropping season in 2020. The results revealed that under the linear mixed model, the factor analytic models were found to be an efficient method for maize MET data analysis. The investigated FA models exhibit improved data fitting, resulting in a significant improvement in heritability. SXM1910008 and 3XM1920126 showed good yield performance over correlated locations, including Ambo, Bako, Hawasa, and Wondogenet, and were therefore identified as potentially useful stable genotypes with a wide range of adaptability. This is because the improved analysis technique we used here showed that correlated locations were the basis for genotype selection. Through the use of more effective statistical models, the analysis of data from multi-environment trials can offer a more robust framework to evaluate maize varieties with increased confidence in choosing superior varieties across a range of environments. Therefore, expanding the use of this effective analysis technique is essential for improving the choice of superior varieties in maize breeding program.

Keywords: factor analytic model, MET analysis; BLUP, mixed model, maize

DOI: 10.7176/JNSR/15-2-01

Publication date: May 31st 2024