The long-term post-fire recovery phase is a critical stage for forest ecosystems to progress toward regeneration and mature succession. During this process, soil bacteria exhibit greater environmental adaptability, rapidly driving nutrient cycling and facilitating vegetation restoration. This study investigated the community structure and diversity of soil bacteria during long-term recovery after forest fires in the cold temperate zone, focusing on soils from the 2000 fires in Daxing’anling. Soil samples were classified into Low (L), Moderate (M), and High (H) fire damage intensity, with bacterial community composition and diversity analyzed using Illumina sequencing technology. After long-term fire recovery, the contents of soil organic carbon, black carbon, total nitrogen, alkaline nitrogen, available phosphorus, and available potassium were significantly higher elevated (p < 0.05), and water content was significantly lower, compared with that in the control check (CK) group. Soil urease, fluorescein diacetate, soil acid phosphatase, and soil dehydrogenase activities were significantly higher, and soil sucrase activity was significantly lower in H. There was a significant difference in the Alpha diversity index among the groups. Compared with CK, the Shannon index was significantly increased (p < 0.05) in L, while both Chao1 and Shannon indices were significantly decreased (p < 0.05) in M and significantly higher in H than CK. The results of the PCoA showed that there was a significant difference in the Beta diversity of the bacterial community among the groups (R² = 0.60 p = 0.001). The dominant bacteria groups were Proteobacteria and Acidobacteriota, while Actinobacteria became the new dominant group during the long-term post-fire recovery. AP, WC, DOC, MBC, S-DHA, and S-SC were significantly and positively correlated with soil bacterial diversity (p < 0.05). The results of the co-occurrence network analysis showed that all groups were dominated by symbiotic relationships, with M having the highest network complexity and strongest competitive effects. This study found that the physicochemical properties of soils recovered over a long period of time after fire returned to or exceeded the unfired forest condition. The Actinobacteria phylum became a new dominant bacterial group, with stronger network complexity and competition, in the process of forest recovery after moderate fire.

Forest fires critically disrupt soil ecosystems by altering physicochemical properties and microbial structure-function dynamics. This study assessed short-term impacts of fire intensities (light/moderate/heavy) on microbial communities in Larix gmelinii forests one year post-fire. Using phospholipid fatty acid (PLFA) and Biolog EcoPlate analyses, we found the following: (1) fire reduced soil organic carbon (SOC), dissolved organic carbon (DOC), total nitrogen (TN), and available nitrogen/potassium (AN/AK) via pyrolytic carbon release, while heavy-intensity fires enriched available phosphorus (AP), AN, and AK through ash deposition. (2) Thermal mortality and nutrient-pH-moisture stress persistently suppressed microbial biomass and metabolic activity. Moderate fires increased taxonomic richness but reduced functional diversity, confirming “functional redundancy.” (3) Neither soil microbial biomass nor metabolic activity at the fire site reached pre-fire levels after one year of recovery. Our findings advance post-fire soil restoration frameworks and advocate multi-omics integration to decode fire-adapted functional gene networks, guiding climate-resilient forest management.

Forest fires are key disturbance factors in forest ecosystems, and soil fungi play an irreplaceable role in post-fire recovery. This study focused on forest areas burned in 2000 in the Daxing’anling region of China, targeting long-term recovery sites with different fire intensities. Illumina MiSeq sequencing was used to analyze the structural characteristics of fungal communities and their environmental drivers. Results showed that compared with the control check (CK), the Shannon index of the low fire group (L) increased significantly (p < 0.05), while moderate (M) and high (H) fire groups reduced fungal diversity significantly. PCoA indicated significant differences in community structure (R² = 0.97, p = 0.001). In highly burned areas, the relative abundance of Ascomycota reached 94.17%, and Basidiomycota lost its dominance. Spearman analysis showed that pH, available phosphorus, available potassium, soil fluorescein diacetate hydrolase, soil dehydrogenase, and soil urease were significantly positively correlated with fungal alpha diversity. RDA revealed that total nitrogen, available phosphorus, soil water content, alkaline nitrogen, active potassium, and dissolved organic carbon had extremely significant effects on soil fungal community composition (p < 0.01). Co-occurrence network analysis indicated that symbiotic relationships dominated all groups. Networks in L and M groups were more complex, while that in H group was simplified and severely damaged. This study indicated that after long-term recovery, soil fungal communities in low fire areas returned to pre-fire levels; those in moderate and high fire areas did not recover, with high fire burns causing severe damage and community structure reorganization.