China, possessing only 7.8% of the world’s arable land, accounts for approximately 22.7% of global N fertilizer consumption (FAO, 2024), reflecting a serious problem of excess N fertilization in its agricultural systems. Previous studies have estimated that nearly 50% of the N fertilizer is lost from agricultural soils through various pathways, including ammonia (NH₃) volatilization, nitrate (NO₃ ^–) leaching and runoff, and nitrous oxide (N₂O) emissions. These reactive N losses not only result in substantial agronomic and economic damages but also cause considerable environmental burdens (Galloway et al., 2008; Coskun et al., 2017). Among these N loss pathways, N₂O is of particular contaminant due to its dual role as a potent greenhouse gas and an ozone-depleting substance. With an atmospheric lifetime of approximately 116 years and a global warming potential 273 times greater than that of carbon dioxide (CO₂), N₂O is major contributor to both stratospheric ozone depletion and climate warming (Tian et al., 2020; IPCC, 2021). Globally, an estimated approximately 60% of anthropogenic N₂O emissions originate from agricultural soils, where microbial processes such as nitrification, nitrifier denitrification, denitrification, and dissimilatory nitrate reduction to ammonium (DNRA) drive N₂O production (Reay et al., 2012). Among these, nitrification and denitrification are widely recognized as the dominant biological pathways, accounting for around 70% of N₂O emissions from soils (Butterbach-Bahl et al., 2013; Wrage-Mönnig et al., 2018). Excessive N fertilization not only intensifies N₂O emissions but may also stimulate the release of other greenhouse gases (GHGs), such as CO₂ and CH₄. Therefore, the identification of effective N₂O mitigation strategies, along with a comprehensive understanding of their efficacy across diverse soil and environmental conditions, is critical for improving N use efficiency and mitigating the agricultural GHGs emissions from agricultural systems.

Nitrification inhibitors (NIs) reduce reactive N losses primarily by suppressing the nitrification process. Currently, the most widely commercialized NIs include 3,4-dimethylpyrazole phosphate (DMPP), 2-chloro-6-(trichloromethyl)pyridine (CP), and dicyandiamide (DCD). DMPP is the most extensively used due to its lower application rates, limited mobility in soil, and higher environmental safety compared to CP and DCD (Gao et al., 2021; Liu et al., 2021). DMPP is reported to inhibit ammonia oxidation process by competitively binding to the copper-binding site on the AmoB subunit, thereby influencing soil N transformations and related microbial communities (Zerulla et al., 2001). Numerous studies have shown that the application of NIs in agricultural soils can increase N recovery efficiency by 4~93% and crop yields by 6~13%, while reducing N₂O emissions by 8~94%. However, the mitigation efficiency of NIs varies considerably across agroecosystems due to differences in soil type, climate, and management practices (Qiao et al., 2015; Lam et al., 2017; Liu et al., 2022a). Urease inhibitors (UIs), such as N-(n-butyl) phosphoric triamide (NBPT), N-(propyl) thiophosphoric triamide (NPPT), and hydroquinone (HQ), function by delaying the hydrolysis of urea. This delay helps moderate the transient pH spike caused by rapid urea delaying, thereby reducing the NH₃ volatilization and other reactive N losses (Cantarella et al., 2018). UIs application has been shown to enhance N use efficiency by 14~29% and crop yields by 3~10%, while reducing N₂O emissions by 2~77% (Li et al., 2018; Liu et al., 2022b). Given the multiple pathways through which reactive N loss in field conditions, the combined application of NIs and UIs (double inhibitors, UNs) is gaining increasing attention. However, compared with the relatively extensive studies on NIs, investigations into the effects of UIs and UNs on N₂O emissions remain limited, possibly due to differences in their modes of action and design rationale.

Various inhibitors offer a simple and operationally feasible strategy for N₂O mitigation, especially when compared to conventional strategies such as deep fertilizer placement, precision fertilization, or integrated agronomic management, which often demand specialized equipment and technical expertise. These advantages render inhibitor-based strategies particularly suitable for large-scale promotion in China’s smallholder-dominated agricultural systems (Xia et al., 2017). Previous studies have demonstrated that NIs primarily function by suppressing the activity and abundance of ammonia-oxidizing bacteria (AOB), and under certain conditions, NIs also influence ammonia-oxidizing archaea (AOA) as well as the abundance of other N-cycling microorganisms and functional genes (Fan et al., 2019; Lu et al., 2019; Liu et al., 2021). However, DMPP has generally shown negligible effects on the overall structure and function of the soil microorganism (Liu et al., 2023a). In alkaline soils, NBPT has been reported to stimulate the relative abundance of nitrification-related genes while simultaneously reducing the abundance of genes involved in denitrification processes (Liu et al., 2022a). The variation in N₂O mitigation efficiency of inhibitors is primarily attributed to the complex interactions between climatic conditions (e.g., rainfall, temperature), soil physicochemical properties (e.g., pH, organic matter, total N, available phosphorus, and potassium) and microbial communities and functions, which influence both biotic and abiotic N transformation processes (Lam et al., 2017; Liu et al., 2023b, 2023c). These differences also reflect the ecological divergence among various types of N-cycling microorganisms (Yin et al., 2022). Previous studies indicated that the NIs efficiency in N₂O mitigation is governed by the interactions between soil physicochemical properties, microbial community composition, and functional gene expression (Shi et al., 2016; Liu et al., 2023b). However, the relative contributions and mechanisms of abiotic and biotic factors in regulating the N₂O mitigation efficiency of UIs and UNs remain poorly understood. This knowledge gap constrains the broader deployment of inhibitors application and limits their mitigation potential in large-scale agricultural systems.

To investigate the underlying mechanisms of variable efficiency of UI and UN on N₂O mitigation. This study conducted controlled microcosm experiments using four distinct agricultural soil types (i.e., dark brown, fluvo-aquic and red soil), which were selected based on pronounced differences in physicochemical properties, particularly soil pH, organic matter content, and particle size distribution, as these factors are considered potential drivers of variation in inhibitor efficiency. Through dynamic observation of inorganic N transformations and N₂O emissions, and integrating 16S-rRNA amplicon sequencing, metagenomic sequencing, and random forest modeling, we aimed to address the following objectives: (i) to evaluate the effects of UI and UN applications on N transformations and N₂O emissions across different soil types; (ii) to identify the key abiotic and biotic factors, as well as the underlying mechanisms, that regulate the N₂O mitigation efficiency of UI and UN.

Reducing reactive N losses from agricultural soils is an effective approach to improving crop N use efficiency, yield, and ecological-economic benefits. This study found that the inhibitors (UI and UN) efficiency significantly varies across different soil types, with both biotic and abiotic soil factors playing crucial roles in determining their efficiency. Such variability leads to inconsistent ecological and economic outcomes from inhibitor application, thereby limiting their scalability and benefits in large-scale agricultural systems.

Soil pH is one of the most critical factors influencing the inhibitors efficiency. Inhibitors (i.e., UI and UN) tend to exhibit greater efficiency in alkaline soils, whereas their efficiency is substantially reduced or even negligible under acidic conditions (Shi et al., 2017; Liu et al., 2023b). It is widely recognized that in alkaline soils, nitrification and nitrifier denitrification are the dominant processes driving N₂O emissions (Wu et al., 2018), while incomplete denitrification plays the primary role in acidic soils (Li et al., 2020; Khan et al., 2022). Consistent with these studies, our study shows that UI application effectively decreased N₂O emissions in alkaline soils, whereas its effect is diminished in acidic soils. In contrast, UN application significantly decreased N₂O emissions across all soil types, although the mitigation efficiency was lower under acidic conditions. For the UI treatment, the observed reduction in peak NO₂− concentrations suggest a potential mechanism of N₂O mitigation. This effect was particularly evident in HB and HN soils, where UI significantly decreased NO₂− peaks. These results support the proposed mechanism by which UI inhibits urease activity by binding to its nickel active site, thereby slowing urea hydrolysis and reducing the availability of nitrification substrates and hydroxylamine oxidation rates, ultimately mitigating N₂O emissions from nitrification (Krol et al., 2020). However, UI application exhibited less effects on the all kinds of inorganic N in JL and ZJ soils, indicating a weak influence on soil N cycling and fails to decrease N₂O emissions in acidic soils (Hube et al., 2017). Moreover, the invalidation of UI on N₂O mitigation also due to UI mitigates nitrification primarily through substrate limitation, and nitrification is not the dominant N₂O-producing pathway in acidic soils. Therefore, the limited nitrification in acidic systems likely constrains the effectiveness of UI-based mitigation strategies.

The UN application exhibits a stronger regulatory effect on inorganic N transformation and N₂O emissions compared to the application of UI alone. The UN treatment consistently increased NH₄ ⁺ concentrations while significantly decreasing NO₂− and NO₃− levels across all soil types. This effect is attributed to the combined mechanisms of UI, which limits the availability of substrates for nitrification (Cantarella et al., 2018), and NI which suppresses both ammonia and nitrite oxidation (Wu et al., 2018). Notably, the N₂O mitigation efficiency of UN in this study slightly exceeded that reported for NI alone in previous studies (Liu et al., 2023b), indicating a synergistic interaction between the two inhibitors. The mechanisms by which the two inhibitors function are complementary rather than conflicting. Such synergy appears to enhance the control of N transformations in soil and contributes more effectively to the suppression of N₂O emissions (Ni et al., 2018; Luchibia et al., 2020). However, previous study has reported that although the combined application of DMPP and NBPT can significantly enhance crop yield and N use efficiency, it may exhibit antagonistic effects on N₂O mitigation in field condition (Zhao et al., 2017). This discrepancy highlights the substantial influence of complex interactions between climatic conditions and soil physicochemical properties on the UN efficiency. Therefore, further studies are needed to investigate the effects of UN on soil nutrient cycling and N₂O emissions under more diverse and field environmental conditions.

SOM is also a key edaphic factor influencing the inhibitors efficiency. Higher SOM levels are typically associated with greater nutrient availability and stronger sorption capacity (Mehnaz et al., 2019). Elevated nutrient content can promote the richness and diversity of soil microorganism and potential function (Wardle et al., 2008; Farrell et al., 2013), thereby increasing the complexity and heterogeneity of microbially mediated biogeochemical processes, including nutrient cycling and greenhouse gas emissions (Chang et al., 2020). As these processes become more intricate, they are less amenable to regulation by single classes of chemical inhibitors. Moreover, the stronger sorption capacity of high SOM soils may lead to the immobilization of small-molecule inhibitors (i.e. DMPP or NBPT) by adsorptive action of organic matrices, thereby reducing their bioavailability and functional efficacy (Tian et al., 2016). Additionally, SOM provides readily available carbon and nitrogen sources that can stimulate microbial populations capable of degrading these compounds, further compromising inhibitor efficiency (Fisk et al., 2015). Consequently, soils with higher SOM contents tend to exhibit a lower inhibitor efficiency. Moreover, a large-scale meta-analysis also revealed a significant negative correlation between SOM content and the mitigation efficiency of chemical nitrification inhibitors (Gao et al., 2021). In this study, although HB and HN soils had similar pH values, both UI and UN were less effective in mitigating N₂O emissions in the HB soil, which had higher SOM content than the HN soil. This negative relationship between SOM content and inhibitor efficacy was also observed across soils with varying pH levels, suggesting that elevated SOM may exert a suppressive effect on the mitigation efficiency of chemical inhibitors.

Soil texture and WFPS are also important influencing factors of both N₂O emissions and the inhibitors efficiency (Volpi et al., 2017; Pelster et al., 2021). Previous studies have shown that, within a certain range, higher clay content and elevated WFPS tend to increase N₂O emissions (Volpi et al., 2017). This is primarily due to the promotion of anaerobic microsites under finer soil particles and high-moisture conditions, which promote incomplete denitrification processes (Balaine et al., 2013; Rochette et al., 2018). Consequently, high clay content and WFPS may reduce the effectiveness of nitrification-targeting inhibitors such as UI and NI, which primarily suppress AOB and nitrification pathways (Sanz-Cobena et al., 2012; Volpi et al., 2017). The findings of our study are consistent with aforesaid perspectives. We observed that increased clay content and WFPS were associated with reduced mitigation inhibitors efficiency, while soil properties that enhance aeration (i.e. high sand content) appeared to improve the mitigation efficiency of inhibitors. In addition, AP and AK have been shown to exert significant influences on soil N₂O emissions (Mehnaz et al., 2019; Li et al., 2020; Liu et al., 2023b). However, their roles in modulating the mitigation efficiency of inhibitors remain relatively underexplored. These nutrient factors need greater attention in future research to better understand their potential interactions with inhibitors performance under various soil and environmental conditions.

RF analysis revealed that soil physicochemical properties were more important predictors of inhibitors efficiency than microbial community α- and β-diversity ( Figure 6 ). Notably, RF models constructed using profiles of nutrient cycling functional gene exhibited higher explanatory weight and model fit compared to those based on soil properties or microbial diversity ( Figure 7 ). These results suggest that many abiotic factors may ultimately influence the inhibitors efficiency by modulating microbial functions and altering soil biochemical processes. Previous studies investigating the microbial mechanisms underlying inhibitor performance have primarily focused on key microbial taxa and functional genes (Wu et al., 2018; Lu et al., 2019). In this study, well-established abiotic predictors, such as soil pH, SOM, WFPS, and texture, were identified as significant variables contributing to inhibitors efficiency base on RF model, these factors frequently emphasized in earlier studies (Shi et al., 2017; Li et al., 2020; Liu et al, 2023b), thereby supporting the robustness and consistency of our analytical approach. Furthermore, functional genes related to nitrification (e.g., amoA, nxrA) and denitrification (e.g., narG, nirS/K, norB), which have been widely recognized for their roles in regulating N₂O emissions and inhibitors efficiency (Ni et al., 2018; Fan et al., 2019; Luchibia et al., 2020), were also found to be strongly associated with inhibitors efficiency in this study. In particular, the gene (hao) of hydroxylamine oxidoreductase, which has often been overlooked in previous studies, exhibited a strong correlation with inhibitors efficiency and was identified as a highly important predictor in the RF model. This underscores its potential as a key functional indicator for evaluating inhibitors efficiency.

In addition, the RF model identified several non-N cycling functional genes as important predictors of NI and UN efficiency, including those involved in carbon fixation/metabolism (por, cooS, ppgK, acnB and glmES), methane metabolism (mch and mcr), sulfur cycling (sat and dsrAB), and phosphorus cycling (pst). Based on previous studies on the ecological roles of elements such as C, P, and K, we speculate that these nutrients elements may influence N₂O mitigation of inhibitors efficiency primarily by limiting growth or death of microorganisms that serve as the direct targets of inhibitors. Nonetheless, current study exploring the influence of non-N nutrient elements on the inhibitor efficiency remains poorly. Further investigation into these interactions is warranted to more comprehensively elucidate the environmental drivers underlying the variability in inhibitors efficiency across heterogeneous soils. Such knowledge will be instrumental in improving the field applicability, functional stability, and agronomic effectiveness of inhibitors under diverse agricultural conditions.

This study aimed to evaluate the effects of UI and UN on N₂O mitigation and its relevant regulatory mechanisms across diverse soil types. The controlled incubation experiments eliminate the influence of meteorological and other environmental factors, providing a relatively stable environment to investigate the regulatory effects and mechanisms by which soil abiotic and biotic factors contribute to variations in inhibitors efficiency. Our results demonstrate that mitigation efficacy of UI and UN exhibited evident variation in agricultural soils, and both abiotic and biotic soil factors influence the efficiency of UI and UN. Among the abiotic factors, soil pH, SOM, and texture were identified as critical factors. Microbial community and functional structure diversity showed relatively weak effects on inhibitors efficiency, while partial functional genes involved in N, C, P, and S cycling, particularly those associated with greenhouse gas emissions and microbial growth, emerged as significant factors. This situation is due to these genes can directly affect GHG production or indirectly influence biotic processes through their impact on microbial activity. Therefore, future research should not only explore the interactive effects of other major nutrient element on the efficacy of chemical inhibitors, but also identify strategies to enhance their performance and overall agronomic benefits.