Genotype plus genotype-by-environment (GGE) biplot analysis of a multi-environment trial (Yan et al., 2000). Gives the genotype and environment scores on the first two principal components, the winning genotype in each environment and the resulting mega-environments (which-won-where), and the average-environment coordination (AEC) ranking of genotypes for mean performance and stability.
Arguments
- data
Data frame with genotype, environment and trait columns (plot data or genotype x environment means).
- genotype
Name of the genotype column.
- environment
Name of the environment column.
- trait
Name of the trait column.
- scaling
"none"(default) or"sd"(each environment divided by its standard deviation, giving environments equal weight).- svp
Singular value partitioning:
"symmetric"(default, suited to which-won-where),"genotype"(genotype-focused, suited to comparing genotypes and the mean vs stability view) or"environment"(environment-focused).- verbose
Logical. If TRUE (default), prints a summary.
Value
An object of class aridagri_gge with
genotype_scores, environment_scores,
variance_percent, which_won_where (environment and
winner), mega_environments, aec (genotype AEC abscissa,
ordinate and ranks) and the settings used. Plot it with
plot().
Details
The genotype x environment table of means is centred on the environment means (and optionally scaled) and decomposed by singular value decomposition. With singular values \(d_k\), genotype scores are \(u_{ik} d_k^f\) and environment scores \(v_{jk} d_k^{1-f}\), with \(f = 0.5, 1, 0\) for symmetric, genotype- and environment-focused partitioning. The winner in an environment is the genotype with the highest value predicted by the two-component biplot; environments with the same winner form a mega-environment. On the AEC view the abscissa (projection on the average-environment axis) ranks mean performance and the ordinate (distance from that axis) measures instability.
References
Yan, W., Hunt, L.A., Sheng, Q. and Szlavnics, Z. (2000). Cultivar evaluation and mega-environment investigation based on the GGE biplot. Crop Science, 40(3), 597-605. doi:10.2135/cropsci2000.403597x
Yan, W. and Kang, M.S. (2003). GGE Biplot Analysis: A Graphical Tool for Breeders, Geneticists, and Agronomists. CRC Press, Boca Raton.
Examples
data(pearlmillet_met)
gg <- gge_biplot(pearlmillet_met, "genotype", "environment", "grain_yield")
#>
#> ======================================================================
#> GGE BIPLOT ANALYSIS
#> ======================================================================
#> Trait : grain_yield
#> Genotypes : 10
#> Environments : 6
#> Scaling : none
#> Singular value partitioning : symmetric
#> PC1 + PC2 : 82.7% of G + GE
#>
#> Which won where
#> Environment Winner
#> BKN-2025 PM-07
#> BKN-2026 PM-07
#> JDH-2025 PM-04
#> JDH-2026 PM-04
#> JSM-2025 PM-04
#> JSM-2026 PM-04
#>
#> Mega-environments (environments sharing a winner)
#> PM-04: JDH-2025, JDH-2026, JSM-2025, JSM-2026
#> PM-07: BKN-2025, BKN-2026
#>
#> Mean vs stability (average-environment coordination)
#> Genotype Mean AEC_abscissa AEC_ordinate Mean_rank Stability_rank
#> PM-04 2016.000 16.756 6.751 1 7
#> PM-07 1971.778 12.518 2.373 2 5
#> PM-01 1923.444 5.888 1.382 3 3
#> PM-09 1829.944 1.871 0.080 4 1
#> PM-02 1856.389 -0.525 8.389 5 8
#> PM-10 1826.722 -2.084 9.846 6 10
#> PM-06 1840.556 -3.250 1.356 7 2
#> PM-08 1757.389 -8.199 6.115 8 6
#> PM-03 1792.222 -8.225 8.648 9 9
#> PM-05 1694.444 -14.749 1.913 10 4
gg$which_won_where
#> Environment Winner
#> 1 BKN-2025 PM-07
#> 2 BKN-2026 PM-07
#> 3 JDH-2025 PM-04
#> 4 JDH-2026 PM-04
#> 5 JSM-2025 PM-04
#> 6 JSM-2026 PM-04
if (requireNamespace("ggplot2", quietly = TRUE)) {
plot(gg, type = "which_won_where")
}
