Spatial heterogeneity and driving factors of global nitrogen use efficiency

Enhancing nitrogen use efficiency (NUE) is essential for advancing food security and sustainable agriculture, yet vast disparities in NUE across environmental gradients impede precise global nitrogen management. Here, we combined 3479 observations worldwide with multi-source remote sensing datasets using machine learning to develop a high-resolution (5 km 2 ) global NUE database. We found that NUE varies strongly with crop type and region (40.86% for maize, 41.09% for wheat, and 49.18% for rice). For example, maize exhibits higher NUE in high-latitude and arid regions, wheat achieves optimal NUE in temperate climates, while rice performs best in tropical zones. Additionally, maize uses nitrogen less efficiently than wheat or rice, resulting in substantially higher nitrogen surpluses in South America, East Asia, and Central Africa. We further identified climate as a dominant regulator of NUE, with optimal hydrothermal conditions supporting greater nitrogen utilization. Taken together, this work provides a high-resolution spatial foundation for designing region-specific nitrogen management strategies, offering a spatial roadmap for optimizing fertilizer application while minimizing environmental harm.