【Objective】This study aims to identify optimal agroforestry management patterns for Pinus koraiensis plantations and establish a scientific foundation for decision-making in compound management practices.【Method】Three agroforestry systems (forest-fungi, forest-medicine, and forest-vegetable) within P. koraiensis plantations were investigated. Quadrat surveys and high-throughput sequencing were integrated to assess species diversity, soil nutrient dynamics, enzymatic activities, and microbial community structures. Full-length 16S rRNA and ITS amplicon sequencing were utilized to characterize bacterial and fungal communities across soil depths of (0,10] and (10,20] cm.【Result】(1)During initial management stages, species diversity indices in agroforestry systems were lower than those in the control (monoculture plantation).(2)The forest-fungi and forest-vegetable systems exhibited higher soil nutrient accumulation (e.g., organic carbon, total nitrogen), whereas forest-vegetable and forest-medicine systems demonstrated superior nutrient conversion efficiency.(3)Soil catalase activity ranked as forest-fungi > forest-medicine > forest-vegetable, while cellulase activity peaked in the forest-medicine system, with significantly higher values in the (0,10] cm layer than in the (10,20] cm layer (P< 0.05). No significant differences were observed in sucrase or acid phosphatase activities among systems. (4)The α-diversity of soil fungal and bacterial communities was significantly different in richness,and not significantly different in biodiversity,microbial communities across all systems were dominated by mortierellomycota, ascomycota, actinobacteriota, and proteobacteriota, with ascomycota and actinobacteriota identified as key taxa driving nutrient utilization under agroforestry practices. (5)Partial Mantel test revealed than the dominant factors influencing bacterial and fungal communities were soil organic carbon,total nitrogen(TN)and plant Simpson index(P<0.001).In the forest-funus system,both catalase activity and cellulase activity exhibited significant effects on fungal communities,whereas in the forest-vegetable system,soil organic carbon content and C to N ratio,along with the plant Simpson index in control plots,showed notable impacts on fungal communities(P<0.05).Additionally,the plant Shannon index and soil sucrase activity in control plots were indentified as the primary drivers of bacterial community composition(P<0.05).【Conclusion】Agroforestry management initially disrupts understory communities and soil ecosystems, but gradually enhances species richness, nutrient cycling efficiency, and microbial functional stability.