In order to provide theoretical and technical basis for returning different organic materials to field and fertilizing soil, soil samples from a 30 a long-term experiment were used to explore how the microbial keystone species affect the contribution of plant-derived lignin phenols to soil organic carbon (SOC) under different plant residue amendment conditions, in which, irrigated desert soils in Hexi oasis irrigation area selected as the research object under different treatments, including green manure (GM), wheat straw (SW), nitrogen fertilizer (N), green manure with nitrogen fertilizer (GMN), wheat straw with nitrogen fertilizer (SWN) and no fertilizer (CK). The results showed that SOC concentrations of GM and SW were significantly higher than those of other treatments. The treatments with wheat straw significantly increased the concentrations of total lignin phenols and three lignin phenol monomers. The total lignin phenol concentration in SW was 109%, 122%, 115%, 164%, and 57% higher than in CK, GM, N, GMN, and SWN, respectively. Compared to non-fertilization and chemical fertilizer treatment, plant residue amendment significantly increased relative abundance of keystone species F3 (Pezizomycetes), F26 (Mortierellomycetes), F173 (Sordariomycetes), B74 (Alphaproteobacteria) and B87 (Actinobacteria). A total of 15 keystone species were identified, of which key species B831 (belonging to Alphaproteobacteria) showed a significant positive correlation with the contribution of lignin phenols to SOC, and was significantly enriched under SW. This study provides a scientific reference for how microbial keystone species affect the contribution of plant-derived lignin phenols to SOC in future agricultural ecosystems under different fertilization methods, and also provides theoretical basis for understanding the impact of plant residue return on SOC turnover.