Spatial patterns of nitrogen runoff from Chinese paddy fields

被引：40

作者：

Hou, Xikang ^{[1
]}

Zhou, Feng ^{[1
]}

Leip, Adrain ^{[2
]}

Fu, Bojie ^{[1
]}

Yang, Hui ^{[1
]}

Chen, Yan ^{[3
]}

Gao, Shuoshuo ^{[1
]}

Shang, Ziyin ^{[1
]}

Ma, Lekuan ^{[3
]}

机构：

[1] Peking Univ, Coll Urban & Environm Sci, Lab Earth Surface Proc, Sinofrance Inst Earth Syst Sci, Beijing 100871, Peoples R China

[2] European Commiss Joint Res Ctr, Inst Environm & Sustainabil, Climate Change Unit, Via E Fermi 2749, I-21027 Ispra, VA, Italy

[3] Chinese Acad Environm Planning, Dept Water Environm Planning, Beijing 100012, Peoples R China

来源：

AGRICULTURE ECOSYSTEMS & ENVIRONMENT | 2016年 / 231卷

基金：

中国国家自然科学基金;

关键词：

Nitrogen loss; Surface runoff; Paddy soils; Nonlinear response; Spatial variation; Climate change; AGRICULTURAL SOILS; PHOSPHORUS LOSSES; NITRATE; WATER; LAND; FERTILIZATION; VARIABILITY; EMISSION; EXPORT; MODEL;

D O I：

10.1016/j.agee.2016.07.001

中图分类号：

S [农业科学];

学科分类号：

09 ;

摘要：

Fertilizer additions (N-rate) have increased surface runoff of nitrogen (R-TN) from croplands, which is eventually transferred to rivers. This pollution further contributes to ecological degradation and and health risks of drinking waters. Despite this recognition, little is known about the spatial pattern of R-TN from croplands and the drivers of its variation. On the basis of 210 site-years of measurements at 41 sites in Chinese paddy fields, we examined the nonlinear response of R-TN to N-rate and the effects of environmental factors on R-TN, R-TN per unit nitrogen fertilizer additions (RR), and background N runoff (R-0). The results show that (i) R-TN-N-rate relationship deviates from linearity and the parameters vary by climate and soil attributes; (ii) Observed variation of R-TN is better explained by precipitation and clay content (48%) than N-rate and its interaction (17%); (iii) The R-TN is 1.09 +/- 0.36 Tg N yr(-1) for Chinese paddy fields in 2008, to which R-0 contributes more than 50%, and the corresponding average RR is 6.8 +/- 1.7% (1 - sigma), 30% lower than the linear model. This study therefore suggests that the future policies for agricultural N runoff need to account for local environmental conditions rather than solely attempting to reduce N fertilizer applications. (C) 2016 Elsevier B.V. All rights reserved.

引用

页码：246 / 254

页数：9

共 61 条

[1] NEW LOOK AT STATISTICAL-MODEL IDENTIFICATION [J].