This study aimed to assess the sustainability of alfalfa-grain maize rotations in central-northern Shanxi Province， China， and to screeen the optimal management strategies. The Agricultural Production System Simulator （APSIM） model was calibrated using multi-site experimental data and employed to simulate long-term historical （1985—2014） and future climate scenarios （2021—2050； 2051—2080）. Three rotation systems were compared： 6A （6-year continuous alfalfa）， 5A1C （5-year alfalfa+1-year grain maize）， and 4A2C （4-year alfalfa+2-year grain maize）. The results indicated that the model performed well in simulating yield and biomass （Normalized Root Mean Squared Error， NRMSE<20%）. Under future climate scenarios， the equivalent yield of 6A was significantly higher than that of 5A1C and 4A2C （P<0.05）， which were increased by 9.45% and 23.02% compared to 5A1C in the 2020s and 2050s， respectively and increased by 23.02% and 28.49% compared to 4A2C. Furthermore， 6A achieved the highest single-cycle equivalent yield （59.98 t/hm²） under the historical scenarios. Regarding precipitation matching， values exceeded 0.6 for maize and were >0.8 for alfalfa during May–August. However， the matching degree was lower in April， September， and October， indicating substantial irrigation requirements. The precipitation matching were significantly improved by future scenarios （P<0.05）， though the values of 6A remained lower than 4A2C during April and May. In conclusion， 6A maximized land productivity under future climates with enhanced irrigation in spring conversely. 4A2C offers a more balanced combination of productivity and water use efficiency. Our results could support the strategies of optimized forage and adaptive rotation strategy in Shanxi.

Alfalfa  /  Maize  /  Crop rotation  /  Climate change  /  APSIM