Foxtail millet (Setaria italica L.) is an important cereal for managing future water scarcity and ensuring food security, due to its strong drought and salt stress resistance owing to its developed root system. However, the molecular responses of foxtail millet leaves to salt stress are largely unknown. In this study, seeds of 104 foxtail millet accessions were subjected to 0.17 mol·L⁻¹ NaCl stress during germination, and various germination-related parameters were analyzed to derive 5 salt-sensitive accessions and 13 salt-tolerant accessions. Hong Gu 2000 and Pu Huang Yu were the most salt-tolerant and salt-sensitive accessions, respectively. To determine the mechanism of the salt stress response, transcriptomic differences between the control and salt-treated groups were investigated. We obtained 2019 and 736 differentially expressed genes under salt stress in the salt-sensitive and salt-tolerant accessions, respectively. The transcription factor families bHLH, WRKY, AP2/ERF, and MYB-MYC were found to play critical roles in foxtail millet’s response to salt stress. Additionally, the down-regulation of ribosomal protein-related genes causes stunted growth in the salt-sensitive accessions. The salt-tolerant accession alleviates salt stress by increasing energy production. Our findings provide novel insights into the molecular mechanism of foxtail millet’s response to salt stress.

Regulated deficit irrigation (RDI) is an important water-saving strategy in arid regions. To quantify the effects of RDI on foxtail millet yield and water use efficiency and determine an optimal RDI strategy, a three-year field trial was carried out over dry, normal, and wet rainfall years in the Loess Plateau. Full irrigation throughout the whole growth period served as the control, whereas mild, moderate, and severe deficit irrigation treatments were conducted at the jointing–booting stage, heading–flowering stage, and across the whole growing period, respectively. The results indicate that the effects of RDI on foxtail millet yield varied with crop growth stage and deficit severity. During the heading–flowering stage, mild RDI showed statistically similar grain yield and WUE relative to those under full irrigation. In normal and wet years, moderate and severe RDI had no statistically significant effects on grain yield and WUE. Additionally, moderate and severe RDI significantly improved irrigation water use efficiency by 19.94–28.50% and 34.35–47.72%, respectively. The primary reason is that RDI at this stage maintained root development and led to only limited suppression of plant growth. In contrast, moderate and severe RDI at the jointing–booting stage or throughout the whole growth period significantly inhibited root establishment and plant development, reduced dry matter accumulation, and consequently led to substantial yield losses. The inhibitory effect became more pronounced with increasing deficit severity. Specifically, severe RDI at the jointing–booting stage and throughout the entire growth period significantly reduced yield by 19.35–54.98% and 31.47–100%, respectively. Furthermore, to identify the optimal RDI regime adaptable to variable rainfall years, a multi-model comprehensive evaluation system based on yield and WUE was established by integrating three individual evaluation models, including the membership function method, TOPSIS, and grey relational analysis, with the Fuzzy–Borda combined evaluation model. The result showed that the heading–flowering stage is the critical period for implementing RDI in foxtail millet in the Loess Plateau. Mild RDI during this stage is preferred because it maintains stable yield and WUE while substantially reducing irrigation amount over various rainfall years. Additionally, moderate and severe RDI can also maintain stable yield while significantly improving irrigation water use efficiency in normal and wet years.