Removing invasive plants is widely considered an effective approach for restoring invaded ecosystems. While the impacts of such removal on aboveground vegetation have been well-documented, its influences on soil microbial communities and multifunctionality remain poorly characterized, thus hampering a comprehensive assessment of its ecological impacts. Here, we assessed the impacts of removing invasive Moso bamboo (Phyllostachys edulis) on soil microbial communities and multifunctionality linked to carbon, nitrogen, and phosphorus cycling in a subtropical forest ecosystem, after five years of removal. Our findings revealed that bamboo removal significantly reduced microbial community complexity, altered community composition, and increased saprotrophic fungal diversity, while having minimal effects on bacterial and total fungal diversity. Moreover, the microbial communities in removal plots exhibited greater stability and lower sensitivity to environmental fluctuations. Notably, soil multifunctionality was higher in removal plots compared to both invaded and uninvaded plots. This improvement, although associated with several microbial community attributes, was primarily linked to greater stability in fungal rather than bacterial communities. Structural equation modeling (SEM) further revealed that fungal community stability mediated the linkages between other microbial community attributes and soil multifunctionality. These results underscore the ecological benefits of invasive bamboo removal for forest soil ecosystems and emphasize the fundamental role of microbial community stability in sustaining soil functions and restoring invaded landscapes.

The effects of nitrogen (N) deposition on forest soil organic carbon (SOC) are largely unclear, likely due to the divergent responses of particulate (POC) and mineral-associated carbon (MAOC). Conventional understory inorganic N (UIN) additions neglect canopy processes and the impacts of organic N, potentially misevaluating N deposition effects. This study was conducted in a long-term N addition experiment established in a Moso bamboo forest, which included six treatments combining canopy and understory N additions with organic (urea + glycine) and inorganic (NH₄NO₃) forms at a rate of 50 kg N·ha^-1·yr^-1. Litterbags were installed for a two-year decomposition experiment and collected at quarterly intervals, together with concurrent soil sampling under litterbags at 0–10 cm depth. We aimed to examine the effects of canopy vs. understory N addition and organic vs. inorganic N form on soil POC and MAOC concentrations. Our results showed that canopy N additions significantly reduced POC (-15.9%) but did not affect MAOC (P > 0.05). Conversely, understory N additions significantly increased POC (+30.9%) and decreased MAOC (-28.9%). Canopy N additions decreased POC by enhancing peroxidase activity and fungal diversity (FuD), while understory N additions promoted POC by inhibiting litter decomposition. Additionally, understory N addition-induced soil acidification decreased soil Ca²⁺ concentration, microbial carbon use efficiency, and bacterial necromass C, as well as the release of litter water-soluble compounds, thereby inhibiting MAOC. Moreover, nitrogen forms (organic vs. inorganic) had no effect on SOC fractions. Our findings underscore that canopy and understory N addition approaches differentially regulate SOC fractions by altering litter decomposition–microbial–mineral interactions, and the understory approach may overestimate soil POC gain and MAOC loss driven by atmospheric N deposition.