文献类型： 外文期刊

作者： Lai, Xiaoming ¹ ; Zhou, Zhiwen ³ ; Liao, Kaihua ¹ ; Zhu, Qing ¹ ;

作者机构： 1.Chinese Acad Sci, Nanjing Inst Geog & Limnol, Key Lab Watershed Geog Sci, Nanjing 210008, Peoples R China

2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China

3.Shanghai Acad Agr Sci, Shanghai 201403, Peoples R China

4.Shanghai Technol Engn Res Ctr Digital Agr, Shanghai 201403, Peoples R China

关键词： Carbon and nitrogen cycles; Model simulation; Rock fragment; Soil hydrology

期刊名称：CATENA （ 影响因子：5.198； 五年影响因子：5.594 ）

ISSN： 0341-8162

年卷期： 2021 年 203 卷

页码：

收录情况： SCI

摘要： Rock fragments (RFs, mineral particles with diameter > 2 mm) can substantially influence soil carbon (C) and nitrogen (N) cycles through different physical mechanisms. These physical mechanisms include changing soil hydraulic parameters (vSH) by reducing fine earth bulk density (vSHBD) and volume (vSHVo), and affecting soil C and N stocks (vCN) by reducing fine earth bulk density (vCNBD) and volume (vCNVo), and increasing fine earth C and N concentrations (vCNCo). In this study, based on soil and climate data in a typical hilly area of China, we construct scenarios by considering these physical mechanisms to investigate the responses of key soil C and N outputs (carbon dioxide or CO2, and nitrous oxide or N2O emissions, and N leaching) to RF content (RFC) in DNDC and DayCent models. Results showed that when considering vSH and vCN, parabolic relationships were observed between these soil C and N outputs and RFC, with maximum in RFC = 0.3-0.6 g g(-1). The vCN dominated the responses of soil CO2 emission to RFC, while vSH exerted comparative contributions to the responses of soil N2O emission and N leaching. When only considering vSH, opposite contributions of vSHBD and vSHVo were observed in DNDC model, and contributions of vSHBD overwhelmed those of vSHVo. However, vSHVo dominated these responses in DayCent model. When only considering vCN, the vCNCo dominated these responses, though considerable superimposed influences of vCNBD and vCNVo appeared under high RFC. The discrepancies between these responses to RFC in DNDC and DayCent models might be due to the different model complexities in simulating soil hydrology, biogeochemistry, and the role of bulk density in these models. Relatively, more distinct responses were achieved in DNDC model. These findings can extend our understandings of RFs and soil C and N cycles.