Multi-trait selection for mean performance and stability in maize

There is a need to strengthen maize (Zea mays L.) breeding strategies based on multivariate selection to obtain high-yielding hybrids that are more stable and resilient to contrasting environmental conditions. Here, we show how the multi-trait stability index (MTSI) can be used to select maize hybrids for mean performance and stability of multiple traits. A set of 10 traits, including grain yield (GY), yield components, and plant-related traits with negative and positive desired selection gains (SGs), were accessed in 90 F-1 hybrids conducted in multi-environment trials. Hybrid and hybrid x location interaction effects were significant (p <= .001) for all analyzed traits. The MTSI provided positive gains for all the four traits that were wanted to increase (2.52% <= SG <= 4.86; mean, 3.28%), including GY (SG, 4.86%), and negative gains for all the six traits that were wanted to decrease (-20.28% <= SG <= -0.09%; mean, -6.70%), including tassel branch number (SG, -20.28%) and plant height (SG, -1.2%). We also observed desired gains for the stability of all traits. Direct and univariate selection for GY solely was not efficient to provide desired gains for all traits. The MTSI provides a unique, robust, and easy-to-handle selection process that allows identifying the strengths and weaknesses of hybrids. The index was found to be a powerful tool to develop better selection strategies, optimizing the use of resources and time, thus contributing to the sustainability of maize breeding programs worldwide.