氧化亚氮（N₂O）是大气中重要的温室气体之一，对全球气候变暖具有显著影响。土地利用方式的改变是影响N₂O排放的关键因素，尤其是在生态系统脆弱的半干旱地区，其影响机制更为复杂。然而，针对我国半干旱区复杂多样的土地利用方式如何影响土壤N₂O排放，其中影响N₂O排放的关键驱动因子，目前尚缺乏系统研究。为此，本文以陇中黄土高原四种典型土地利用方式：云杉林地（Picea asperata）、苜蓿草地（Medicago sativa）、撂荒地（Abandoned land）、小麦地（Wheat field）为研究对象，采用静态箱-气相色谱法监测土壤N₂O通量，结合土壤理化性质数据，揭示不同土地利用方式下调控土壤N₂O排放的关键驱动因子。 结果表明：（1） 与撂荒地相比，云杉林地和苜蓿草地显著提高了土壤含水量，而小麦地则增加了铵态氮（NH₄⁺-N）和硝态氮（NO₃^--N）的含量。（2） 与撂荒地相比，苜蓿草地和小麦地显著提升了硝酸还原酶（NR）和亚硝酸还原酶（NIR）活性，且各处理的NR和NIR活性均随土层加深而降低。（3） 在不同土地利用方式下，土壤N₂O通量随植被生长阶段呈先增后降的趋势。相较于撂荒地，云杉林地和苜蓿草地的土壤N₂O总排放量分别减少了34.2%、23.3%，而小麦地则显著增加了32.47%。（4） 随机森林结果表明，土壤温度对土壤N₂O排放通量的影响最大。相较于撂荒地和小麦地，人工林地和草地表现出更好的减排效应。在未来的植被恢复和生态修复过程中，应注重“农林草”土地利用方式的分配比例，适当提高人工林地和草地的占比，以实现生态效益与减排效应的双重目标。

Nitrous oxide (N₂O) a remarkable greenhouse gas in the atmosphere, exerts a pronounced effect on global climate warming. Changes in land use types critically affects N₂O emissions, particularly in ecologically fragile semiarid regions with more complex underlying mechanisms. However, there is still a lack of systematic research on how complex and diverse land use types affect soil N₂O emissions in semiarid regions of China and the key driving factors involved. To address this, this study focused on four typical land use types in the semiarid Loess Plateau of central Gansu Province: Picea asperata forest, Medicago sativa grassland, abandoned land, and wheat field. Soil N₂O fluxes were monitored using the static chamber-gas chromatography method, combined with soil physicochemical property data, to elucidate the key drivers regulating soil N₂O emissions under different land use types. Compared to the abandoned land, the Picea asperata forest and Medicago sativa grassland had significantly increased soil water content, while wheat fields exhibited elevated ammonium nitrogen (NH₄⁺-N) and nitrate nitrogen (NO₃^--N) concentrations. Compared to the abandoned land, the Medicago sativa grassland and wheat fields exhibited markedly enhanced nitrate reductase and nitrite reductase activities, which decreased with an increase in soil depth across all treatments. Soil N₂O fluxes under different land use types exhibited an initial increase followed by a decline during the vegetation growth stages. The total soil N₂O emissions decreased by 34.2% and 23.3% in the Picea asperata forest and Medicago sativa grassland, respectively, and increased by 32.47% in the wheat fields, compared to the abandoned land. Random forest analysis identified soil temperature as the most influential factor affecting the soil N₂O flux. Overall, compared to the abandoned land and wheat fields, the artificial forest and grassland systems in the study area demonstrated superior emission reduction effects. Therefore, future vegetation restoration and ecological rehabilitation efforts should prioritize optimizing the proportional allocation of “forest-grass-cropland” land use types and appropriately increasing the coverage of artificial forests and grasslands to achieve the dual objectives of ecological benefits and emission mitigation.