您好,欢迎访问湖南省农业科学院 机构知识库!

Factors Affecting the Vertical Distribution of Silicon in Paddy Soils in Mid-subtropical China

文献类型: 外文期刊

作者: Zhang, Pengbo 1 ; Ouyang, Ningxiang 1 ; Wei, Xiao 1 ; Zhang, Yangzhu 3 ; Hu, Bifeng 4 ; Lu, Zhaoyan 1 ; Peng, Hua 7 ; Zhang, Jiachao 3 ; Li, Xia 8 ; Xie, Modian 4 ;

作者机构: 1.Hunan Univ Finance & Econ, Changsha 410205, Peoples R China

2.Hunan Econ Geog Technol Dev Co Ltd, Changsha 410205, Peoples R China

3.Hunan Agr Univ, Coll Resources & Environm, Changsha 410128, Peoples R China

4.Jiangxi Univ Finance & Econ, Sch Publ Finance & Publ Adm, Dept Land Resource Management, Nanchang 330013, Peoples R China

5.Zhejiang Univ, Inst Agr Remote Sensing & Informat Technol Applica, Coll Environm & Resource Sci, Hangzhou 310058, Peoples R China

6.Zhejiang Univ, Key Lab Environm Remediat & Ecol Hlth, Minist Educ, Hangzhou 310058, Peoples R China

7.Hunan Acad Agr Sci, Hunan Inst Agroenvironm & Ecol, Changsha 410125, Peoples R China

8.Shimen Cty Cultivated Land Qual Monitoring & Prote, Changde 415300, Peoples R China

关键词: Available silicon; Parent material; Machine learning; Plough layer; Subsurface layer

期刊名称:SILICON ( 影响因子:3.4; 五年影响因子:3.2 )

ISSN: 1876-990X

年卷期: 2023 年

页码:

收录情况: SCI

摘要: Silicon can promote rice growth and reduce the toxicity of heavy metals in rice. However, controlling factors for the vertical distribution of silicon in typical paddy soils in mid-subtropical regions of China remain largely unexplored. Here, we aimed to analyze the silicon content in plough and subsurface layers derived from six parent materials in Hunan Province (southern China) based on 62 soil profile samples. Multiple stepwise regression, random forest, and gradient boosting machine learning models were used and compared to predict soil available silicon content and quantify the contribution of different variables on silicon variation. We found that plough layers in the eastern (85% of points) and western (78% of points) Hunan Province were deficient (< 100 mg kg(-1)) in available silicon content. Total silicon content in paddy soil derived from slate (623 & PLUSMN; 100 g kg(-1)) and limestone (606 & PLUSMN; 74 g kg(-1)) was higher than in other soils. The available silicon content was the lowest in granite (51 & PLUSMN; 15 mg kg(-1)). It was closely correlated with available cadmium while being mainly influenced by soil pH, organic carbon, and bulk density. The prediction model of available Si established using a random forest had the highest accuracy (R-2 = 0.74). Widespread deficiency in available silicon content in plough layers of paddy soil in mid-subtropical China highlights the importance and value of higher silicon fertilizer input in soils.

  • 相关文献
作者其他论文 更多>>