Document Type : Original Article
Authors
1
Seed and Plant Improvement Institute, Agricultural Research Education and Extension Organization (AREEO), Karaj, Iran.
2
Assistant Professor, Crop and Horticultural Science Research Department, Fars Agricultural and Natural Resources Research and Education Center, AREEO, Shiraz, Iran
3
Assistant Professor, Crop and Horticultural Science Research Department, Kermanshah Agricultural and Natural Resources Research and Education Center, AREEO, Kermanshah, Iran
4
Associate Professor, Crop and Horticultural Science Research Department, Kerman Agricultural and Natural Resources Research and Education Center, AREEO, Kerman, Iran
5
Assistant Professor, Crop and Horticultural Science Research Department, Khorasan Razavi Agricultural and Natural Resources Research and Education Center, AREEO, Mashhad, Iran
6
6- Assistant Professor, Crop and Horticultural Science Research Department, South Kerman Agricultural and Natural Resources Research and Education Center, AREEO, Jiroft, Iran
7
7- Assistant Professor, Crop and Horticultural Science Research Department, Safiabad Agricultural and Natural Resources Research and Education Center, AREEO, Dezful, Iran
8
Researcher, Crop and Horticultural Science Research Department, West Azerbaijan Agricultural and Natural Resources Research and Education Center, AREEO, Urmia, Iran
9
Assistant Proffessor, BioControl Research Department, Iranian Research Institute of Plant Protection, Agricultural Research, Education and Extension Organization (AREEO), Tehran, Iran
10.48308/envs.2024.1421
Abstract
EXTENDED ABSTRACT
Introduction: Multi-environment trials are essential for evaluating the performance of genotypes across diverse environments and identifying stable genotypes with high performance, either as varieties or parental lines in breeding programs. These trials are often conducted across multiple locations and years, where the combination of year and location is considered as the environment. Most previous models, such as the widely used AMMI (Additive Main Effects and Multiplicative Interaction) and GGE biplot (Genotype + Genotype-by-Environment Interaction) models, calculate genotype-by-environment interaction (GEI) based on the combination of location and year, without separating GEI into genotype-by-location interaction (GLI), genotype-by-year interaction (GYI), and genotype × location × year interaction (GLYI). Since year-induced variations are often random, less repeatable, and do not show consistent trends, using these previous models for interpreting interaction effects face significant challenges. As a result, genotype rankings differ depending on the years of evaluation, making it difficult to select superior genotypes for target locations. Therefore, this study was conducted to compare the efficiency of the HO-AMMI, AMMI, and GGE biplot methods for selecting stable and high-yielding maize hybrids in multi-environment trail.
Materials and methods: To evaluate the usefulness of separating GLI from GEI, an experiment was conducted with 14 promising maize hybrids along with two commercial control hybrids (Hybrids 15, and 16) in a randomized complete block design with four replications across eight locations (Karaj, Shiraz, Kermanshah, Kerman, Mashhad, Isfahan, Miandoab, and Dezful) over two years (2020 and 2021). Each plot consisted of four 6.5 m rows spaced 0.75 m apart, 0.36 m between hills with two plants in each. To eliminate the marginal effect, only the two middle rows were harvested. Grain yield was then measured at the moisture content of 14%. The AMMI and GGE biplot models were used to estimate GEI, while the HO-AMMI (Higher-order-AMMI) model was used to estimate GLI.
Results and Discussion: Based on the average grain yield across 16 environments, hybrids No. 10 and 14 had the highest grain yields, with 12.42 and 12.36 tons per hectare, respectively. According to the results, the HO-AMMI model successfully distinguished the high-yielding hybrids (Hybrids No. 10 and 14) with less yield variation than other hybrids. In contrast, the AMMI and GGE biplot model could not separate the other hybrids from the high-yielding group. It appears that the GGE biplot method more accurately identifies specific adaptability. In general, Hybrid No. 14 has less yield variability than other high-yielding hybrids, and hence may be recommended to farmers.
Conclusion: In the HO-AMMI model, selecting genotypes for the desired location is easier than with the AMMI model. The GLI biplot in the HO-AMMI model operates based solely on genotype and location, without the confounding effect of year. Conversely, the year effect in the AMMI model may overshadow the location effect, making genotype selection more difficult. The HO-AMMI model provides an accurate ranking of genotypes for a specific location without the confounding effects of GYI and GLYI, enabling breeders to identify high-yielding genotypes for target locations. Therefore, the HO-AMMI model can be effectively used in multi-environment trials to select stable, high-yielding genotypes.
Keywords
)