Objective Nitrogen (N) fertilization leads to greenhouse gas (GHG) emissions from croplands, and the emission intensity is jointly mediated by fertilizer type, N application rate, and soil physicochemical properties. However, previous studies have primarily focused on surface soils, with limited efforts devoted to a systematic synthesis of GHG fluxes across soil profiles in rainfed croplands, particularly regarding deep-soil processes and their controlling factors. This study aimed to quantify the effects of nitrogen fertilization on soil-profile emissions of CO₂, CH₄, and N₂O in rainfed croplands worldwide, identify the major driving factors, and improve understanding of the underlying carbon and nitrogen cycling mechanisms in agroecosystems.

Methods Relevant peer-reviewed studies were retrieved from Chinese and international databases, including CNKI and Web of Science, using keywords related to nitrogen fertilization, greenhouse gases, soil profiles, and deep soil. Only field experiments that included both control and nitrogen treatment groups and reported soil-profile GHG fluxes with extractable statistical information were considered. In total, 32 studies conducted at 11 global sites between 1997 and 2020 were included, providing 643 paired observations. To ensure comparability among studies, soil profiles were standardized into five depth intervals: 0–10, 10–20, 20–40, 40–60, and >60 cm. A meta-analysis was conducted to evaluate the effects of nitrogen fertilization on soil-profile GHG fluxes and to identify the major regulating factors.

Results N fertilization exerted pronounced depth-dependent effects on GHG fluxes in rainfed croplands. Moderate nitrogen input (<200 kg N/hm²) increased CO₂ fluxes by 5%–11%, whereas excessive nitrogen application (>400 kg N/hm²) decreased CO₂ emissions by 8%, likely because of suppressed microbial activity. The stimulatory effect of nitrogen on CO₂ emissions was more evident in soils with higher soil organic carbon content (>5.12 g/kg) or near-neutral pH. N₂O emissions increased exponentially with nitrogen input, indicating a clear threshold response. Rainfed croplands generally acted as a sink for atmospheric CH₄, but the effect of nitrogen fertilization on CH₄ fluxes was bidirectional: CH₄ uptake was inhibited in soil layers above 60 cm but enhanced below 60 cm. This pattern suggests that nitrogen fertilization may reshape the deep-soil CH₄ sink by altering oxygen diffusion along the profile and methanotrophic activity. Nitrogen fertilization also significantly increased global warming potential, with N₂O contributing the largest share. In addition, combined application of organic and inorganic fertilizers reduced GHG emission intensity, with the strongest mitigation effect observed in neutral soils of arid regions.

Conclusion Nitrogen fertilization significantly increases CO₂ in surface soil and N₂O emissions across 0-60 cm profile, and reduce the CH₄ sink effect in the soil profile above 60 cm. Heavy nitrogen application exaggerate the green house emission effect. The contribution of deep soil to sustained N₂O release and to shifts in CH₄ sink function suggest the importance of controlling nitrogen fertilizer rate for achieving the dual goals of crop productivity and GHG mitigation.