Soil water and phosphorus availability determines plant-plant facilitation in maize-grass pea intercropping system

Aims Plant-plant interactions are determined by resource availability. However, its responses to differences in soil phosphorus (P) and water and the rhizospheric driving mechanism have not been systematically revealed. Methods A pot-culture experiment was conducted in monoculture and in mixture with maize and grass pea. Resource stress treatments were created by organic and inorganic P applications (phytate and KH 2 PO 4 ) as well as drought and well-watered treatments to address the above issue. Results Intercropping increased system yield and biomass by 3.5% and 4.5% on average in soils with sufficient P and water, and by 12.0% and 11.7% under stressful conditions respectively. With increasing water and P availability, the yield-based interaction types transited from mutually facilitated to maize facilitated with grass pea as facilitator. Comparatively, the biomass-based facilitation shifting from maize facilitated and grass pea neutral to both being facilitated with increasing stress. This shift was probably mediated by complementary use of soil water in the two species, and rhizospheric acidification in grass pea. Also, rhizospheric interactions promoted phosphatase secretion by 5.3–22.2%, and mobilized organic P to buffer P limitation under P-deficient soils. Regardless of crop species, reproductive and vegetative biomass fell into a typical allometric pattern (α > 1). The rhizospheric interactions drove the differentiated biomass allocation patterns and the species-specific interaction transition. Conclusion High P and water environments intensified the asymmetric interspecific competition. The trait-dependent facilitation shift was mechanically driven by rhizospheric interaction in intercropping systems. The findings update the understanding on stress gradient hypothesis in a modified model.