Updated loss factors and high-resolution spatial variations for reactive nitrogen losses from Chinese rice paddies

Yiwei Shang,Yulong Yin,Hao Ying,Xingshuai Tian,Zhenling Cui

DOI: https://doi.org/10.1016/j.jenvman.2024.120752

IF: 8.7

2024-04-14

Journal of Environmental Management

Abstract:Anthropogenic reactive nitrogen (Nr) loss has been a critical environmental issue. However, due to the limitations of data availability and appropriate methods, the estimation of Nr loss from rice paddies and associated spatial patterns at a fine scale remain unclear. Here, we estimated the background Nr loss (BNL, i.e., Nr loss from soils without fertilization) and the loss factors (the percentage of Nr loss from synthetic fertilizer, LFs) for five loss pathways in rice paddies and identified the national 1 × 1 km spatial variations using data-driven models combined with multi-source data. Based on established machine learning models, an average of 23.4% (15.3–34.6%, 95% confidence interval) of the synthetic N fertilizer was lost to the environment, in the forms of NH 3 (17.4%, 10.9–26.7%), N 2 O (0.5%, 0.3–0.8%), NO (0.2%, 0.1–0.4%), N leaching (3.1%, 0.8–5.7%), and runoff (2.3%, 0.6–4.5%). The total Nr loss from Chinese rice paddies was estimated to be 1.92 ± 0.52 Tg N yr −1 in 2021, in which synthetic fertilizer-induced Nr loss accounted for 69% and BNL accounted for the other 31%. The hotspots of Nr loss were concentrated in the middle and lower regions of the Yangtze River, an area with extensive rice cultivation. This study improved the estimation accuracy of Nr losses and identified the hotspots, which could provide updated insights for policymakers to set the priorities and strategies for Nr loss mitigation.

environmental sciences