Soil-microbial biomass stoichiometric characteristics serve as a robust indicator of ecosystem nutrient cycling efficiency， yet the mechanisms by which different artificial grassland restoration modes influence these traits remain unclear. This study investigated 3-year-old artificial grasslands dominated by tall-grass species （SF）， short-grass species （XF）， and a mixed planting of tall- and short-grass species （SFXF） in the Three-River-Source Region， with severely degraded alpine grassland （CK） as the control. Soil organic carbon （SOC）， total nitrogen （TN）， total phosphorus （TP）， microbial biomass carbon （MBC）， nitrogen （MBN）， and phosphorus （MBP） in the 0-10 cm layer were measured to analyze their stoichiometric ratios and microbial quotients. The impacts of different restoration modes on soil-microbial stoichiometric balance were further elucidated.The results demonstrated that all three artificial grassland establishment modes significantly enhanced SOC， MBN， and MBP contents. Specifically， the SF treatment increased soil TN， whereas the XF treatment reduced TN and decreased TP by 50.42%. MBN incresed by 24.36% and 35.59% under SF and XF treatments， respectively. The SFXF treatment decreased the microbial C-quotient by 78.40%， while the XF treatment elevated the microbial N-quotient by 49.61%， and the SFXF treatment reduced it by 11.33%. After restoration， the stoichiometric imbalances in soil-microbial biomass （Cimb：Nimb， Cimb：Pimb， and Nimb：Pimb） increased across all treatments， with the most pronounced escalation observed in the SFXF treatment. Redundancy analysis （RDA） revealed that soil N：P and C：P ratios were key drivers of microbial quotient variations. In conclusion， the SFXF mixed planting exerted the most substantial influence on soil-microbial C-N-P cycling. For practical management， supplemental nitrogen fertilization is recommended to optimize plant growth and improve soil quality.

Artificial grassland  /  Microbial biomass  /  Stoichiometric ratio  /  Microbial quotient  /  Stoichiometric imbalance