Canodraw

The data of physicochemical analysis and taxonomic assignment were used for multivariate analysis. Canonical correspondence analysis (CCA) between microbial diversity (Species data) and physicochemical parameters (Environment data) was performed using the software Canoco 4.5 (Microcomputer Power, Ithaca, NY, United States) in association with CanoDraw®. Relative abundances (x) at class level (Species data) were transformed by arcsin (x/100)^1/2, while the environmental variables were transformed using log₁₀(x + 1). Chloroplast, unassigned sequences, and classes below 5% of relative abundance were removed for the species data matrix for further analysis. To obtain tests of independent effects of physicochemical variables, co-variables were eliminated in CCA. The effect of each environmental variable and their interaction on the sum of all canonical eigenvalues was tested by Monte Carlo permutation tests available in Canoco using 999 replicate runs to assess significance on the trace values.

The Shapiro−Wilk and Tukey’s honestly significant difference (HSD) tests (α = 0.05) were used to examine the normality and homoscedasticity of the residuals. Tukey’s HSD test (α = 0.05) was used to compare means. SAS 9.2 (SAS Institute, Inc., Cary, North Carolina, USA) was used to conduct statistical analyses using one-way analysis of variance (ANOVA). Origin Pro 2021 (Norhtampton, MA, USA) was used to evaluate correlations between belowground root morphological features, soil pH, biomass and N accumulation in the intercropping system. To detect significant changes between treatments, the international standard software CANOCO 5 and CanoDraw (Microcomputer Power, Ithaca, USA) were used for ranking procedures, and a linear model was chosen for redundancy analysis (RDA).