Objective To study the effects of intercropping soybeans with different root architecture and sweet maize on crop phosphorus (P) uptake and growth under different P levels, and explore the relationship among root morphology, root architecture, P availability in rhizospheric soil and so on within the intercropping system.

Method Field experiments were conducted in August 2022 at the Teaching and Experimental Base of South China Agricultural University in Zengcheng District, Guangzhou City, Guangdong Province. Two different soybean cultivars, namely ‘Bendi 2’ (deep-rooted genotype) and ‘Yuechun 03-3’ (shallow-rooted genotype) were selected as experimental materials. The field experiment employed a split-plot design, with the main plots comprising two P application levels: With P (+P: 40 kg·hm⁻² for soybean, 120 kg·hm⁻² for sweet maize) and without P (−P: 0 kg·hm⁻² for both soybean and sweet maize). The subplots consisted of five different planting patterns as follows: Sweet maize monocropping, soybean ‘Bendi 2’ ‖ sweet maize, soybean ‘Yuechun 03-3’ ‖ sweet maize, soybean ‘Bendi 2’ monocropping and soybean ‘Yuechun 03-3’ monocropping. The yield, biomass, P uptake, root morphology and architecture of sweet maize and soybean were measured, and the land equivalent ratio (LER) and interspecific competitiveness of the intercropping system were calculated.

Result Intercropping soybean significantly increased the yield of sweet maize by average of 33.4% under both P levels. Under −P condition, intercropping soybean significantly increased the biomass of sweet maize by average of 62.7%. Intercropping sweet maize and soybean exhibited advantages, influencing by P application levels. Under +P condition, the average LER for intercropped sweet maize and soybean was 1.08, while under −P condition, the average LER increased to 1.21. Interspecific competition analysis indicated that in the intercropping system, sweet maize showed significantly stronger competitiveness than soybean (interspecitic competitiveness > 0). This competitive advantage was more pronounced under −P condition. In addition, intercropping with deep-rooted soybean genotype (‘Bendi 2’) significantly promoted P uptake in sweet maize under −P condition by average of 40.6%. Further analysis revealed that for deep-rooted soybean ‘Bendi 2’, intercropping with sweet maize under −P condition widened its root width, altered the proportion of fine roots (diameter ≤ 0.5 mm), and significantly increased P availability in the rhizosphere of intercropping maize. The correlation analysis of traits of soybean and sweet maize in the intercropping system revealed a significant positive correlation between the biomass of sweet maize and its total root length, root width of soybean, and available P content in the soybean rhizosphere soil. The biomass of soybean showed a significant positive correlation with its P uptake, total root length, root width, and available P content in the rhizosphere soil. However, it exhibited a significant negative correlation with its proportion of fine roots (diameter ≤ 0.5 mm).

Conclusion Intercropping with deep rooting soybean under −P condition can effectively promote the growth and P absorption of sweet maize, and increase its yield. These findings provide important theoretical basis for fully exploring the P utilization potential in the sweet maize‖soybean intercropping system, screening suitable soybean varieties for crop intercropping, and achieving reduced P fertilization rate with increased P use efficiency.