Objective  The effects of rhizosphere precipitation on phosphorus (P) availability in the cereal-legume intercropping system mainly focus on the activation of root exudates. The effects of rhizosphere precipitation on P availability in red soil are not clear.

  Methods  Maize and soybean intercropping was used as the research object. Three root barriers, namely, complete separation of the root system, separation of nylon net, and no separation of the root system, were used in pot experiments at 0, 21.83, 43.67, 65.50 and 87.34 P mg kg^–1 (P0, P1, P2, P3 and P4) application levels. Maize and soybean intercropping with root barrier, nylon mesh barrier, and no barrier pot experiments were conducted. The experiment was carried out at the levels of microbial biomass carbon (MBC), dissolved organic carbon (DOC), rhizosphere soil organic carbon (ROC), acid phosphatase activity (ACP), alkaline phosphatase activity (ALP), available phosphorus and Hedley P fractions in intercropping maize and soybean rhizosphere soil.

  Results  Compared with a root barrier, no barrier can improve soil MBC in maize and soybean rhizosphere, significantly reduced soil DOC in maize rhizosphere, increased DOC in soybean rhizosphere at low P levels (P0, P1), significantly improved ACP activities in maize (only in low P) and soybean rhizosphere soil, and significantly increased soil ALP activities in soybean rhizosphere at low P level. In addition to the labile P fraction of maize, the available P and P fraction in rhizosphere soil of intercropping maize and soybean were significantly affected by root barrier methods. The labile Pi in maize rhizosphere soil was increased through decreasing labile Pi (except P0) and moderate Pi in soybean rhizosphere. In general, root interaction decreased the labile Po and moderate Po in maize and soybean rhizosphere. Regression analysis showed that root biomass was significantly positively correlated with MBC content in maize rhizosphere and DOC content in soybean rhizosphere (P < 0.01). Redundancy analysis showed that MBC accounted for 74.4% of available P and P components in maize rhizosphere soil, and DOC accounted for 18.3% of available P components in soybean soil. Mantel test showed that the relationships of labile carbon (C) fraction and phosphatase activities were significantly correlated with root biomass and soil P fraction in maize rhizosphere were much more than soybean.

  Conclusion  Root interaction can improve the P availability in red soil by changing labile C fraction (MBC and DOC) in rhizosphere soil, promoting the secretion of phosphatase by microorganisms, and improving the conversion of moderate P fractions (Pi and Po) to labile Pi. MBC and DOC are the most important factors affecting the rhizosphere P fraction of maize and soybean, respectively. Therefore, the mechanism in the influence of rhizosphere precipitation on the P availability in maize and soybean rhizosphere could be different.