Identification of sorghum hybrids with high phenotypic stability using GGE biplot methodology

Assessing yield performance and stability of local sorghum genotypes: A methodological framework combining multi-environment trials and participatory multi-trait evaluation

Sorghum grain is a vital staple cereal crop for food and nutritional security for rural households in Burkina Faso. However, its yields are regularly affected by environmental and socio-economic constraints. Here, we aim to assess the performance and grain yield stability of local sorghum genotypes, as well as their acceptability by smallholder farmers. Nine genotypes were assessed across five environments in the North-Sudanian zone (700-900 mm) in Burkina Faso during the 2019, 2020, and 2021 rainy seasons. A randomized complete block with three replications was used to establish the experiments, from which data were collected on five quantitative traits as well as five participatory assessment criteria chosen by sorghum farmers and users. Grain yield analysis for each environment revealed significant differences between genotypes. The combined analysis showed highly significant differences between genotypes, environments and years, as well as their interactions. Most of the variation in grain yield was hexplained by the environment effect (29.0%), followed by the environment-by-year interaction (20.3%). The GGE biplot analysis identified two mega-environments (ME) with ME1 consisting of one environment (E1) and ME2, represented by four environments (E2, E3, E4, and E5). The E1 is a non-discriminating and poor environment with the lowest grain yield (1506 kg ha-1). The E5 and E2 were respectively, the most discriminating and representative environments, with also the highest grain yields (2406 and 2102 kg ha-1). In terms of stability, genotypes G6, G3, G5, and G9 exhibited the highest stability but lower performance, while G4 was the most unstable. G2 and G8, which produced respectively 2240 and 2072 kg ha-1, were better adapted to ME2. G2 was identified as the closest to the "ideal genotype". The principal component analysis showed that genotypes G2, G8, G7, G4, and G9 were the most selected in both individual and group assessments, owing to the panicle productivity, the good grain quality for storability, the grain attractiveness, and grain heaviness. This study highlighted the potential of genotypes G2 and G8 as promising varieties that could broaden the range of improved varieties and offer income opportunities for sorghum smallholder farmers in Burkina Faso.

Graphical analysis of multi-environmental trials for wheat grain yield based on GGE-biplot analysis under diverse sowing dates

BACKGROUND

Information on the nature and extent of genetic and genotype × environment (GE) interaction is extremely rare in wheat varieties under different sowing dates. In the present study, the GGE biplot method was conducted to investigate genotype × environment interaction effects and evaluate the adaptability and yield stability of 13 wheat varieties across eight sowing dates, in order to facilitate comparison among varieties and sowing dates and identify suitable varieties for the future breeding studies.

RESULTS

Considerable genotypic variation was observed among genotypes for all of the evaluated traits, demonstrating that selection for these traits would be successful. Low broad sense heritability obtained for grain yield showed that, both genetic and non-genetic gene actions played a role in the control of this trait, and suggested that indirect selection based on its components which had high heritability and high correlation with yield, would be more effective to improve grain yield in this germplasm. Hence, selection based on an index may be more useful for improvement of this trait in recurrent selection programs. The results of the stability analysis showed that the environmental effect was a major source of variation, which captured 72.21% of total variation, whereas G and GE explained 6.94% and 18.33%, respectively. The partitioning of GGE through GGE biplot analysis showed that, the first two PCs accounted for 54.64% and 35.15% of the GGE sum of squares respectively, capturing a total of 89.79% variation. According to the GGE biplot, among the studied varieties, the performance of Gascogen was the least stable, whereas Sirvan, Roshan, and Pishtaz had superior performance under all sowing dates, suggesting that they have a broad adaptation to the diverse sowing dates. These varieties may be recommended for genetic improvement of wheat with a high degree of adaptation.

CONCLUSION

The results obtained in this study demonstrated the efficiency of the GGE biplot technique for selecting high yielding and stable varieties across sowing dates.