Evaluation of performance and stability in response to multiple environments in maize

Frequently occurring extreme weather events and environmental changes may significantly reduce corn ( Zea mays L.) yields. Thus, the selection of favorable traits and stable genotypes has emerged as a fundamental objective of breeding programs aimed at countering adverse weather effects. Field experiments in eight environments were conducted in 2019 and 2020 to evaluate the performance and stability of 93 inbred maize lines by multiple models and parameters. The genotype–environment interaction (GEI) plot and GEI effect functions in the Metan package were used to visualize the response patterns of different genotypes in multiple environments. Response patterns of 93 inbred lines with different traits across eight environments were constructed, revealing substantial GEI for anthesis–silking interval, days to 50% anthesis, and days to 50% silking, which were primarily influenced by environmental factors. Through evaluation by multiple methods, a total of 13 genotypes demonstrated excellent performance across four or more parameters or models, such as Zong31, Xz5426, and so forth. Based on the multi‐trait stability index (MTSI) model, all traits were positively selected. Grain yield had the highest selection weight at 25.8%, while ear barren tip had the lowest at 6.19%. Thirteen genotypes were selected, with DH509‐9 being the most stable (MTSI = 3.75). Cross‐validation revealed superior predictive accuracy in all additive main effects and multiplicative interaction (AMMI) models compared to best linear unbiased prediction (BLUP) models. The mean root mean square prediction difference was highest for AMMI0 (72.06) and lowest for BLUP_e (27.08), and AMMI0 model was the optimal model. The approach investigated in this research has the potential to significantly streamline the decision‐making process for breeders to identify genotypes characterized by both high average performance and robust phenotypic stability.