A comparative study on single and multiple trait selections of equatorial grown maize hybrids
Muhammad Azrai, Muhammad Aqil, Roy Efendi, Nining Nurini Andayani, Andi Takdir Makkulawu, R. Neni Iriany, Suarni, Muhammad Yasin, Suwardi Suwardi, Bunyamin Zainuddin, Salim Salim, Trias Sitaresmi, Bahtiar Bahtiar, Paesal, Willy Bayuardi Suwarno
Abstract
Maize ( Zea mays L. ) production in tropical equatorial regions faces significant challenges due to agroclimatic and soil fertility variability, necessitating the evaluation of maize hybrid adaptability and phenotypic stability across diverse agroecosystems. This study compares the effectiveness of the additive main effects and multiplicative interaction (AMMI) and multi-trait genotype-ideotype distance (MGIDI) models for identifying superior maize hybrids well-suited to the equatorial climate. Fifteen genotypes, including 13 hybrid candidates and two popular commercial varieties (BISI 2 and NASA 29), were analyzed in 10 distinct environments in Indonesia over three consecutive years (2018–2020). The ANOVA method used in the AMMI model analyzed variance into three major components, with PCA analysis indicating that environments (E), genotypes (G), and their interaction (G × E) had a highly significant effect on yield ( p < 0.001). Two hybrids, HM04 (CI301032/G102612) and HM02 (CI272022/G102612), displayed high adaptability and stability across various environments, with significantly higher yields than the grand mean by AMMI analysis. Additionally, HM10 (MAL03/CLYN231) and HM09 (G102612/CLYN231) were narrowly adapted to the ME-1 and ME-2 mega-environments, indicating they are best suited for these specific environments. Similar to AMMI, the MGIDI model suggested HM04 (MGIDI index = 1.74) and HM02 (MGIDI index = 1.76) as the two highest-performing hybrids, determined by their yield and nine other traits. Using the multiple trait combination index as a tool to assess the performance of these hybrids enabled researchers to determine the most effective traits for each genotype. The two models are recommended and may be integrated for comprehensive data interaction analysis, which simplifies the process of delineating genotypes with the environment and enables stakeholders to select desired traits while considering their strengths and weaknesses.