Increased yield stability of U.S. Midwest maize hybrids released during the last 40 years of breeding

Context

Yield stability is a desirable agronomic trait. While newer maize hybrids are considered more stable across environments than older ones, debate remains due to lack of consensus in assessing this trait and a lack of robust multi-environment datasets.

Objectives

This study aimed to evaluate yield stability in maize hybrids over time and its variation across environments.

Methods

Data included 39 hybrids released from 1980 to 2020, tested in 27 U.S. Midwest environments over three seasons (2020–2022). Because breeders seek hybrids with high mean yields and low variability across environments, linear mixed-effects models were applied to examine yield in relation to year of hybrid release, focusing on the yield deviation (variability around the regression line) instead of traditional slope-based metrics. The lower the deviation, the higher the stability. Self-organizing maps were used to visualize genotype-by-environment (GxE) patterns and stability trends.

Results

Results indicated significant positive yield genetic gains (123 kg ha⁻¹ year⁻¹; p < 0.001). Absolute yield deviation was not significantly affected by breeding. Relative yield deviation (i.e. deviation relative to the regression line) decreased at a rate of ∼1 % per decade (p < 0.001), with this reduction being more pronounced in the less productive environments. Patterns of GxE reinforced the higher stability of newer hybrids and allowed the identification of superior and stable genotypes, highlighting opportunities to explore physiological traits driving this improvement.

Conclusions

These results support the success of modern breeding programs in enhancing both productivity and resilience.