This paper sampled the soils of Pinus sylvestris var. mongolica plantations in the Hulunbuir, the Horqin and the Mu Us Deserts. Combining field survey and molecular biology methods, this study analyzed the diversity of soil fungal communities by Illumina MiSeq high-throughput sequencing, and soil multifunctionality was calculated by the Z-score method as well as its influencing factors. The results indicated that 1) Soil organic carbon and hydrolyzable nitrogen of P. sylvestris plantations showed significant differences (Hulunbuir Desert> Horqin Desert> Mu Us Desert, P<0.05). Soil available cobalt in the Mu Us Desert was significantly higher than those in the Hulunbuir and the Horqin Deserts, while the soil available copper, available iron and available manganese were opposite (P<0.05). The soil multifunctionality of P. sylvestris plantations showed significant differences (Horqin Desert > Hulunbuir Desert > Mu Us Desert, P<0.05). 2) The species diversity of soil fungi in the Horqin Desert was the highest, Chao 1 and species richness index were significantly higher than those in the Hulunbuir and the Mu Us Deserts, and Shannon index was significantly higher than that in the Mu Us Desert (P<0.05). The phylogenetic diversity index of soil fungi in the Horqin Desert was the highest, which was significantly higher than that in the Hulunbuir and the Mu Us Deserts (P<0.05). Soil pH of P. sylvestris plantations showed significant differences (Mu Us Desert > Hulunbuir Desert> Horqin Desert, P<0.05); the soil electrical conductivity and soil moisture content in the Mu Us Desert were significantly higher than those in the other two deserts (P<0.05). The soil temperature in Hulunbuir Desert was significantly higher than those in the other two deserts (P<0.05). 3) The random forest model showed that altitude, longitude, latitude, pH, soil temperature, fungal Shannon index and species richness were important predictors for identifying soil multifunctionality, with an explanation of 83.68%. Geospatial factors had extremely significant positive effects on soil factors and soil multifunctionality (P<0.001), soil factors had significant negative effects on soil multifunctionality (P<0.05), and soil fungal diversity index had no significant effect on soil multifunctionality (P>0.05). The soil multifunctionality of P. sylvestris plantations in the Horqin Desert was the highest, and its soil ecosystem was more stable. Soil fungal diversity and abiotic factors drove soil multifunctionality of P. sylvestris plantations, but abiotic factors were more important than biotic factors. These results provide a better understanding of soil multifunctional characteristics of P. sylvestris plantations from the perspective of biotic and abiotic factors, as well as the rational utilization and scientific management of P. sylvestris plantations ecosystem. 

以呼伦贝尔沙地、科尔沁沙地和毛乌素沙地樟子松(Pinus sylvestris var. mongolica)人工林为研究对象, 采用野外调查和分子生物学相结合的方法, 利用Illumina MiSeq高通量测序, 分析樟子松人工林土壤真菌群落的多样性, 采用均值法量化土壤的多功能性并分析其影响因子, 得到如下结果。1) 樟子松人工林土壤有机碳和水解性氮含量排序均为呼伦贝尔沙地>科尔沁沙地>毛乌素沙地, 且呈显著性差异(P<0.05); 毛乌素沙地土壤的有效钴含量显著高于呼伦贝尔沙地和科尔沁沙地, 有效铜、有效铁和有效锰含量则相反(P<0.05)。樟子松人工林土壤多功能性的排序为科尔沁沙地>呼伦贝尔沙地>毛乌素沙地, 且呈显著性差异(P<0.05)。2) 科尔沁沙地土壤真菌物种多样性最高, Chao 1指数和物种丰富度指数显著高于呼伦贝尔沙地和毛乌素沙地, Shannon指数显著高于毛乌素沙地(P<0.05); 科尔沁沙地土壤真菌系统发育多样性指数最高, 且显著高于呼伦贝尔沙地和毛乌素沙地(P<0.05)。樟子松人工林土壤 pH值的排序为毛乌素沙地>呼伦贝尔沙地>科尔沁沙地, 且呈显著性差异(P<0.05); 毛乌素沙地土壤电导率和土壤含水量显著高于其他两个沙地(P<0.05); 呼伦贝尔沙地土壤温度显著高于其他两个沙地(P<0.05)。3) 随机森林模型分析结果显示, 海拔、经度、纬度、土壤 pH、土壤温度、真菌Shannon指数和物种丰富度是识别土壤多功能性的重要预测因子, 解释度为83.68%; 地理空间因子对土壤因子和土壤多功能性均表现出极其显著的正效应(P<0.001), 土壤因子对土壤多功能性表现出显著的负效应(P<0.05), 土壤真菌多样性指数对土壤多功能性无显著影响(P>0.05)。科尔沁沙地樟子松人工林土壤多功能性最高, 土壤生态系统更稳定; 生物因子(土壤真菌多样性)和非生物因子共同影响樟子松人工林土壤多功能性, 但非生物因子比生物因子更重要。研究结果有助于从生物和非生物因子角度解析樟子松人工林土壤多功能性特征, 并为合理地利用和管理樟子松人工林生态系统提供科学依据。