中文摘要：

农田土壤-植物系统的氮素循环影响了生产力和环境， 但土壤微生物之间的相互作用对氮素循环的影响机制仍不清楚， 同时这种相互作用如何响应种植作物等管理方式也不明确. 本研究在中国东部3个气候带， 选择3种典型的地带性土壤类型（寒温带黑土、暖温带潮土和中亚热带红壤）设置不种植（裸地， non-cropping）和种植玉米（cropping）的田间试验， 基于高通量基因芯片测定不同土壤共有的氮转化基因（核心氮转化基因）， 利用随机矩阵方法建立土壤核心氮转化基因的分子生态网络， 揭示种植玉米对土壤核心氮转化基因网络结构的影响. 研究表明种植玉米增加了土壤中大部分核心氮转化基因的丰度， 显著提高了核心氮转化基因网络的复杂程度. 网络拓扑结构的模块数由裸地处理的8个增加到种植玉米的28个. 裸地土壤核心氮转化基因网络有2个模块枢纽， 其关键基因为固氮基因（nifH）； 种植玉米后网络有9个模块枢纽， 其关键基因包含固氮（nifH）和反硝化基因（narG和nosZ）. 土壤核心氮转化基因的功能分子生态网络结构与植物、气候、土壤等因素显著相关， 说明农田管理和环境条件的变化可以通过改变微生物的分子生态网络结构， 影响其驱动农田养分循环的功能.

英文摘要：

It is well established that nitrogen cycling in agricultural soil-plant systems greatly affects productivity and the environment. However, the impact of microbial interactions on nitrogen cycling and its response to management practices such as cropping and fertilization remain unclear. In this study, three typical soils were selected （a black soil located in the cold temperate region, a Chao soil in the warm temperate region and a red soil in the middle subtropical region） along a North-South transect of eastern China, and a controlled field experiment was set up with non-cropping （bare soil） and cropping （planting maize） treatments. The core microbial communities （with members shared across different habitats） associated with N transformations were identified from high-throughput functional gene array hybridization data. Functional molecular ecological networks （fMENs） were then developed by a random matrix theory （RMT）-based conceptual framework. Cropping increased the richness of most N transforming core genes and the complexity of N transforming fMENs. The number of modules in the topological structure of fMENs increased from eight under non-cropping to 28 under the cropping treatment. Two module hubs （key genes） of the nitrogen fixation gene （nifI4） were present under the non-cropping treatment, and nine module hubs including the nitrogen fixation gene （ntf/4） and the denitfification genes （narG and nosZ） were present under the cropping system. Additionally, changes in the network structure were significantly correlated with plant, climate and soil variables. In conclusion, agricultural management practices and environmental changes can alter the network interactions in soil microbial communities, and consequently affect their nutrient cycling functions.