A. adenophora seed germination and seedling mortality

Seed germination time and rate after soil or leaf inoculation at the G0 (a-b) and seedling mortality rate after soil or leaf inoculation at the G0 or G21 and subsequent growth in Petri dishes (c). Seedling mortality rate after two weeks when seedlings were transplanted into soil (d). No: nothing inoculated, Ss: sterile soil; Sns: non–sterile soil; Ls: sterile leaf; Lns: non–sterile leaf; G0, inoculated on the day of germination; G21, inoculated on the 21st day after germination; G28, inoculated on the 28th day after germination; G21+28, inoculated on both the 21st and 28th days after germination. * P < 0.05, *** P < 0.001. Error bars depict the standard error. Different lowercase letters represent significant differences among the different inoculation time treatments (P < 0.05). No lowercase letters indicate no differences among the four inoculation time treatments under the same nutrient level (P > 0.05).

GLM analysis (a) and comparison (b) of the role of microbes in seedling growth based on total dry biomass (RI) between leaf and soil inoculated at low and high nutrient levels across four inoculation time points.

Source refers to the leaf or soil inoculation source, time refers to the four inoculation time treatments, and nutrition refers to the nutrient level of the soil. For G0, G21, G28 and G21+28, see Figure 1. Error bars depict the standard error. The difference in the RI among the different inoculation time treatments is indicated by different capital and lowercase letters for the soil and leaf inoculation treatments, respectively. The asterisks indicate that the RIs are significantly different from zero. “No” indicates no surviving seedlings at harvest. P < 0.05 is shown in bold.

Microbial community composition and potential functional differences between leaf litter (AAL) and rhizosphere soil (AAS) inocula (a-h), as well as correlations of microbial genera with seed germination and seedling mortality (i-j).

Core bacteria (a-b) and fungi (e-f) in the AAS and AAL groups. The potential functions of the core bacteria (c-d) and fungi (g-h) in the AAS and AAL. Correlations of the relative abundance of the top 30 bacterial (i) and fungal (j) genera with germination time (GT), germination rate (GR) and mortality rate (MR). Only the top 10 core taxa and potential functional groups are shown in the figures. The “un” in the figures is the abbreviation for “unclassified”, and the MM is Methylobacterium-Methylorubrum. Several bacterial functions classified as N circles related to the AAL and AAS are shown in Table S1-2. Several fungal guilds classified as plant pathogen-related guilds in the AAL and AAS are shown in Tables S3-4. Red and blue represent negative and positive Spearman’s coefficients, respectively. * P < 0.05, ** P < 0.01, *** P < 0.001. Different lowercase letters in the heatmap represent significant differences in the relative abundance of the same genus between AAS and AAL (P < 0.05).

Cultivable fungi associated with dead A. adenophora seedlings and their seedling-killing effects on A. adenophora.

Isolation frequency from one dead seedling (a) and cultivable fungal community composition at the genus level (b). The seedling-killing effects of 33 fungal strains on A. adenophora and their phylogenetic signals (c).

Enrichment of bacterial and fungal communities and functions by A. adenophora seedlings under different treatments.

Nonmetric multidimensional scaling (NMDS) ordinations of Bray–Curtis dissimilarity matrices with permutational analysis of variance (PERMANOVA) of bacterial and fungal communities and function (a-h). Contribution of plant compartment, inoculation source and nutrient level to the variation in bacterial and fungal communities and function at each inoculation time based on PERMANOVA (i-l). * P < 0.05, ** P < 0.01, *** P < 0.001.

Correlations of the enriched microbial community and function with A. adenophora seedling growth.

These genera accounted for more than 1% of the total relative abundance in seedling roots or leaves. (a) Correlations between 47 out of 214 genera enriched in roots with RI (left) and the relative abundances of genera under different treatments, i.e., different inoculation sources and time and nutrient levels (right). (b) Correlations between 18 out of 184 genera enriched in leaves with RI (left) and the relative abundances of genera under different treatments (right). (c) Correlations between putative bacterial functions enriched in roots with RIs (left) and the relative abundances of functions with negative and positive correlations under different treatments (right). (d) Correlations between fungal guilds enriched in roots with RIs (left) and the relative abundances of guilds with negative and positive correlations under different treatments. (e) Correlations between fungal guilds in leaves with RIs (left) and the relative abundances of guilds with positive correlations under different treatments (right). No bacterial functions in the leaves showed a significant correlation with seedling growth. “F_” represents fungal genera, “B_” represents bacterial genera, “un” represents unclassified; H: RI of seedling height; AD: RI of aboveground dry biomass; UD: RI of underground dry biomass; TD: RI of total dry biomass. Red and blue represent negative and positive Spearman’s coefficients, respectively. * P < 0.05, ** P < 0.01, *** P < 0.001. For abbreviations of fungal function guilds, please see Table S8.

Schematic diagram of soil or litter inoculation at different growth stages.

The abbreviations G0, G21, G28, and G21+28 are shown in Fig. 1.