Modeling long-term water use and economic returns to optimize alfalfa-corn rotation in the corn belt of northeast China

The long-term effects of alfalfa (Medicago sativa L.)-corn (Zea mays L.) rotation on soil water use and economic returns have not been evaluated and optimized in the corn belt of northeast China (CBNC). Crop modeling based on long-term weather data and short-term field data helps evaluate agricultural practices’ long-term effect. The Agricultural Production Systems sIMulator (APSIM) was calibrated and validated based on the data sets on soil water storage (SWS), aboveground biomass (biomass), and yield under continuous corn (CC), five years of alfalfa (5A), and three years of alfalfa followed by two years of corn (3A2C) cropping systems from 2014 to 2018 at Quanyangou-Lishu in Jilin province, China. The APSIM could simulate well SWS, evapotranspiration (ET), and annual biomass of alfalfa (5A) with normalized root mean square error (NRMSE) of 9.6%, 5.0% and 13.2%, respectively. This model was also confirmed to simulate well the SWS, ET, yield, and biomass of corn (CC) with the NRMSE of 9.6%, 3.4%, 6.5% and 9.8%, respectively and that of alfalfa-corn rotation (3A2C) with the NRMSE of 8.4%, 4.5%, 8.7% and 12.2%, respectively. The validated APSIM was used to optimize the alfalfa-corn rotation systems by simulating water use and economic returns under five scenarios, including CC, 3A2C, four years of alfalfa followed by two years of corn (4A2C), five years of alfalfa followed by two years of corn (5A2C) and five years of alfalfa followed by five years of alfalfa (5A5A) based on the weather data from 1984 to 2018 (35 years) at Quanyangou-Lishu (black soil with losing mollic epipedon, relatively warm and moist climate, in Jilin province, the center of CBNC), Fujia-Lishu (aeolian sandy soil, relatively warm and moist climate, in Jilin province), Chaoyang (cinnamon soil, warm and dry climate, in Liaoning province, the south of CBNC) and Hailun (black soil with moderately thick mollic epipedon, cold and wet climate, in Heilongjiang province, the north of CBNC). Dry soil layers were only found in 250–400 cm soil layer in 18-yr 5A5A or 20-yr 5A2C treatment at Chaoyang and in 350–450 cm soil layer in 20-yr 5A5A or 26-yr 5A2C treatment at Quanyangou-Lishu. At four sites, the highest and lowest mean annual biomass water use efficiency (WUEB) could be found in CC (30.91–38.93 kg ha⁻¹ mm⁻¹) and in 5A5A (20.53–26.31 kg ha⁻¹ mm⁻¹), respectively. The alfalfa-corn rotation treatments had similar WUEB ranged from 22.91 to 30.17 kg ha⁻¹ mm⁻¹. The NR of alfalfa-corn rotation treatments at four sites were 38–169% higher than that of CC and 6–35% lower than that of 5A5A. Alfalfa-corn rotation treatments improved the stability of NR at Fujia-Lishu and Chaoyang by the CV of 40–75% and 65–95%, compared to that of CC treatment by the CV of 79% and 101%, respectively. Among alfalfa-corn rotation treatments, the 5A2C had higher and more stable NR than 4A2C and 3A2C. Therefore, we suggested 5A2C at Fujia-Lishu and Hailun and 4A2C at Quanyangou-Lishu and Chaoyang to achieve dry soil layer’s prevention, relatively high WUEB and high and stable NR.