3.2 Shifts in rhizosphere microbial diversity during forest restoration
There were 3035 bacterial and 647 fungal OTUs commonly identified at all forest ages. The numbers of unique bacterial OTUs identified in the RS15, RS25, RS35, and RS45 rhizosphere soils were 1027, 839, 783, and 979, respectively; the corresponding numbers of unique fungal OTUs were 395, 602, 1218, and 931, respectively (Fig. 1).
There was no significant difference in rhizosphere bacterial α-diversity among the four forest ages (Fig. S1A). The Shannon and Simpson indices for bacteria peaked in the RS15 rhizosphere soil, while bacteria richness did not change significantly among forest ages. Overall, fungal α-diversity increased first and then decreased with an increase in forest age, and the peak values were observed in the RS35 rhizosphere soil (Fig. S1B). As for fungi α-diversity, fungal richness was significantly different between the RS35 and RS45 rhizosphere soils, but not between the RS15 and RS25 rhizosphere soils. However, Shannon index changed significantly between the RS15 and RS25 rhizosphere soils, although there was no difference between the other two treatments. Simpson index did not differ significantly among the RS25, RS35, and RS45 rhizosphere soils.
Distinct variation in rhizosphere microbial β-diversity was observed among forests with different ages. In the case of bacteria, β-diversity first decreased and then increased, and finally decreased, with an increase in forest age (Fig. 2A). In the case of fungi, β-diversity increased from RS15 to RS35, and then deceased in RS45 (Fig. 2B). Subsequently, we visualized and quantified differences among rhizosphere microbial communities in terms of bacterial and fungal β-diversity using PCoA and PERMANOVA. The bacterial communities were clearly separated into two groups along PCoA 1 (explain ratio 51.98%), but the fungal communities were not (explain ratio 43.99%) (Fig. 2C-D).
Across all samples, Proteobacteria (27.7%),Acidobacteriota (21.0%), and Actinobacteriota (7.7%) were the dominant phyla in the bacterial communities (Supplemental information Fig.S2). The relative abundances of the taxa increased with an increase in forest age. Analysis of bacterial community composition based on Bray-Curtis distances across different forest ages revealed that the bacterial communities were clustered based on forest age; one cluster comprised RS15 and RS25 samples, while the other cluster contained RS35 and RS45 samples. In the case of fungi,Basidiomycota (57.4%) and Ascomycota (20.2%) were the dominant phyla among all the samples (Supplemental information Fig.S3).Basidiomycota abundance decreased with an increase in forest age, while Ascomycota abundance exhibited an opposite trend. Conversely, fungal community composition was not clearly clustered based on forest age.