Integrating field experiments and DSSAT modeling to optimize soybean irrigation and planting density in arid northwest China

The trade-off between limited water availability and optimal population configuration poses a critical challenge to soybean production in the arid regions of northwestern China. This study investigated the regulatory mechanisms by which the interaction between irrigation and planting density under mulched drip irrigation affects soybean growth, yield, and water productivity (WP), and sought to identify a precise cultivation strategy that integrates high yield, high efficiency, and high economic return. Based on a two-year field experiment, three irrigation quotas (W) and three planting densities (D) were factorially combined to characterize irrigation–density responses in soybean canopy structure and biomass accumulation. On this basis, the DSSAT-CROPGRO-Soybean model was calibrated and validated using field observations, and subsequently used to construct 84 irrigation–density interaction scenarios for simulation-based optimization with economic benefit as the objective function. The field experiments showed that irrigation and planting density had significant coupled effects ( P < 0.05). The moderate-density and mild deficit irrigation treatment (D2W2) optimized canopy leaf area index (LAI) and assimilate allocation, thereby overcoming the problems of intensified population competition under high density (D1) and source limitation under moderate water deficit (W3). As a result, D2W2 achieved a two-year mean grain yield of 4488 kg·ha −1 , representing a 25.87% increase compared with the D3W3 treatment (low-density with moderate deficit irrigation). D2W2 also achieved the highest water productivity (WP) in both years, which was 27.51% higher than that of the D3W1 treatment (low-density with full irrigation). Although full irrigation (W1) increased crop water consumption, it significantly reduced WP because of luxury transpiration. Model evaluation showed good simulation performance for yield, aboveground biomass, ET, and WP, with overall normalized root mean square error (NRMSE) values below 10% for all four variables, indicating generally reliable prediction of soybean productivity and water-use responses under different irrigation and planting density conditions. Scenario simulations identified the optimal management strategy for soybean under mulched drip irrigation in the arid region of northwestern China as an irrigation quota of 435 mm and a planting density of 34.19 × 10 4 plants·ha −1 , achieving the maximum economic benefit of 13,957 CNY·ha⁻¹ while maintaining high yield and water productivity. This strategy can effectively balance population competition and individual plant development, providing a quantified irrigation–density synergistic management scheme and decision-making basis for water-saving and high-yield soybean cultivation in desert oasis regions.