Abstract:Cereal-legume intercropping pattern can promote the growth and yield of intercrops. Rhizosphere microbiome is believed to be closely associated with yield advantages of intercropping systems. Therefore, the pot experiments were conducted in this study for maize-peanut intercropping with three different root barrier treatments including plastic-film separation (namely complete separation, denoted as CM for maize and CP for peanut), nylon-net (50 μm) separation (denoted as SM for maize and SP for peanut) and no separation (denoted as NM for maize and NP for peanut). High-throughput 16S rRNA sequencing and PICRUSt functional prediction were applied to analyze the changes in bacterial community structure and functional categories in both maize and peanut rhizosphere under different intercropping patterns, in order to explore the ecological mechanisms of yield advantages of maize-peanut intercropping from the perspective of soil microbiome. The results showed that maize-peanut intercropping remarkably affected their bacterial community structure in rhizosphere. The bacterial communities in NM and SM were more similar and were considerably different from CM. However, the bacterial communities in SP were distinctly different from NP and CP. Furthermore, it was found that the relative abundance of Actinobacteria was significantly higher in NM and SM than in CM, and the relative abundance of Actinobacteria and Firmicutes were significantly higher in SP than in NP and CP. By contrast, the relative abundances of Gemmatimonadetes showed the opposite trend in both maize and peanut rhizosphere. These operational taxonomic units (OTUs), which were more abundant in nylon-net separation and no separation than in complete separation in both maize and peanut rhizosphere, were mainly assigned to the phylum Actinobacteria. Co-occurrence network analysis indicated that Streptomyces, Actinomadura, Arthrobacter, and Bacillus were the keystone nodes in maize bacterial network, showing a significantly higher abundance in NM and SM than in CM. Similarly, the keystone nodes in peanut bacterial network including Streptomyces, Actinomadura and Arthrobacter showed a significantly higher abundance in SP than in NP and CP. Lastly, PICRUSt-based functional prediction showed that maize-peanut intercropping significantly increased the relative abundances of functional categories including membrane transport and carbohydrate metabolism but decreased the abundance of the functional category related to replication and repair. In conclusion, maize-peanut intercropping selectively shaped the bacterial communities in both maize and peanut rhizosphere with an increase in the relative abundances of several potential probiotics, which improved the rhizospheric micro-ecological environment and functions and thus promoted the formation of yield advantages of intercropping.