Evaluation of Stability Analysis Methods for Maize Improvement in Nigeria.

Nigeria produces more than 50% of the total maize (Zea mays L.) produced in West and Central Africa and has taken the lead in the whole of sub-Saharan Africa. The demand for maize in the different agroecologies of the country continues to be insatiable; therefore, new varieties and hybrids are being produced consistently. There is a great need to release varieties to farmers in the ecologies where they perform best; that is, it is necessary to take advantage of the genotype x environment interaction always present in evaluation of new varieties developed on station. The objectives of this study were to (i) evaluate the different stability analysis methods and (ii) identify the most reliable for release of varieties into the rainforest agroecology of Nigeria. Seventeen OPVs were evaluated in 22 environments and the grain yield data were subjected to 15 stability analysis methods classified into four groups: variance analysis, regression, non-parametric, and clustering methods. Results showed that some of the methods were similar in classifying the varieties while some were quite different. The clustering techniques such as AMMI and GGE that are capable of displaying the GEI in biplots appeared to be the best in the study.