Weakened Growth of Cropland‐n2o Emissions in China Associated with Nationwide Policy Interventions
Ziyin Shang,Feng Zhou,Pete Smith,Eri Saikawa,Philippe Ciais,Jinfeng Chang,Hanqin Tian,Stephen J. Del Grosso,Akihiko Ito,Minpeng Chen,Qihui Wang,Yan Bo,Xiaoqing Cui,Simona Castaldi,Radoslaw Juszczak,Asa Kasimir,Vincenzo Magliulo,Sergiy Medinets,Volodymyr Medinets,Robert M. Rees,Georg Wohlfahrt,Simone Sabbatini
DOI: https://doi.org/10.1111/gcb.14741
IF: 11.6
2019-01-01
Global Change Biology
Abstract:China has experienced rapid agricultural development over recent decades, accompanied by increased fertilizer consumption in croplands; yet, the trend and drivers of the associated nitrous oxide (N 2 O) emissions remain uncertain. The primary sources of this uncertainty are the coarse spatial variation of activity data and the incomplete model representation of N 2 O emissions in response to agricultural management. Here, we provide new data‐driven estimates of cropland‐N 2 O emissions across China in 1990–2014, compiled using a global cropland‐N 2 O flux observation dataset, nationwide survey‐based reconstruction of N‐fertilization and irrigation, and an updated nonlinear model. In addition, we have evaluated the drivers behind changing cropland‐N 2 O patterns using an index decomposition analysis approach. We find that China's annual cropland‐N 2 O emissions increased on average by 11.2 Gg N/year 2 ( p < .001) from 1990 to 2003, after which emissions plateaued until 2014 (2.8 Gg N/year 2 , p = .02), consistent with the output from an ensemble of process‐based terrestrial biosphere models. The slowdown of the increase in cropland‐N 2 O emissions after 2003 was pervasive across two thirds of China's sowing areas. This change was mainly driven by the nationwide reduction in N‐fertilizer applied per area, partially due to the prevalence of nationwide technological adoptions. This reduction has almost offset the N 2 O emissions induced by policy‐driven expansion of sowing areas, particularly in the Northeast Plain and the lower Yangtze River Basin. Our results underline the importance of high‐resolution activity data and adoption of nonlinear model of N 2 O emission for capturing cropland‐N 2 O emission changes. Improving the representation of policy interventions is also recommended for future projections.