The field experiments were conducted in 2019 and 2020 on maize cultivated under flood and drip irrigation with water from the Yellow River in Wuyuan County, Inner Mongolia (Wuyuan; 41°4′N, 108°2′E), China. The experimental site is located in a temperate, arid, and semi-arid continental climate. The annual temperature was 6.1°C and 6.3°C; precipitation was 173 mm and 171 mm; and the frost-free period was 117 and 126 days of the field experiments for 2019 and 2020, respectively. The main soil properties of the 0–30-cm layer before the experiment commenced were as follows: pH, 8.8; organic matter content, 7.4 g kg⁻¹; nitrate-N content, 11.8 mg kg⁻¹; olsen-phosphorus (P) content, 20.1 mg kg⁻¹; and available potassium (K) content, 132.9 mg kg⁻¹.

The field experiments were established with a randomized block design with a plot size of 6.5 m× 10 m. Four N treatments with four replications were employed: control (no N fertilizer application), OM (the optimized N management), OMI (optimized N management incorporated with 0.05% urease inhibitor), and FP (farmer practice). The N rate of the OM treatment was determined based on the maize production studies conducted by our group at the same location in previous years. The N rate of the FP treatment was based on broad surveys of the local farmers’ common practices in the region. The maize growth stages were based on the Biologische Bundesanstalt, Bundessortenamt, and CHemical industry (BBCH) classification (Lancashire et al., 1991). The N fertilizer was applied before sowing (30%), BBCH 30–39 (30%), BBCH 51–59 (30%), and BBCH 61–69 (10%), except for the Control treatment. Details concerning N fertilizer rates (urea 46% N) and maize cultivars in the two growing seasons are presented in Table 1 . The maize variety was Xinyu 12 in 2019 and Jindan 42 in 2020 with a population of 75,000 plants ha⁻¹. The calcium superphosphate (90 kg P₂O₅ ha⁻¹) and potassium sulfate (120 kg K₂O ha⁻¹) were applied before sowing as basal fertilizers. The flood irrigation rates applied at specific growth stages were 120 mm at BBCH 30–39 and 90 mm at each of BBCH 51–59, BBCH 61–69, and BBCH 71–79. The drip irrigation rates applied at specific growth stages were 50 mm at each of BBCH 30–39, BBCH 51–59, BBCH 61–69, and 60 mm at both BBCH 71–79 and BBCH 83–89. The experiment’s tillage, herbicide, pesticide use, and other practices were the same as local farmers, except for N fertilizer application, irrigation, and maize harvesting.

The maize plants covering an area of 6.6 m² were manually harvested at maturity in the middle of each field plot to calculate yield. The N leaching, NH₃ volatilization, and N₂O emission were measured using the field percolation pond, glycerol phosphate-sponge ventilation method, and static chamber methods in the actual field experiments, respectively (Cui et al., 2024; Guo et al., 2022; Huang et al., 2017). The percolation water collection device is shown in Figure 1 , which was embedded in a pit of 90 cm depth with a plastic boundary. The soil was excavated in a 30-cm increment layer and then returned to its original depth after installing the leaching bucket. The leachate was collected using an electrical vacuum pump of 100 kPa to a triangular flask during the maize growing stages (before sowing, BBCH 30–39, BBCH 51–59, BBCH 61–69, and BBCH 71–79 under flood irrigation and BBCH 30–39, BBCH 51–59, BBCH 61–69, BBCH 71–79, and BBCH 83–89 under drip irrigation). The leachate samples were stored in a 200-mL polyethylene bottle and immediately frozen at −20°C prior to analysis. The NH₃ was measured by sponge tracking and the KCl extraction method in situ for the maize field (Cui et al., 2024). Two sponges (2 cm thick and 16 cm inner diameter) spiked with 15 mL of a glycerol-phosphoric acid (H₃PO₄, 85.0%) mixture with 40 mL of glycerol, 50 mL of H₃PO₄, and 910 mL of deionized water and then inserted into containers 10 cm long and 15 cm in inner diameter in experimental plots. One sponge was inserted 5 cm above the soil surface of the container and was used to trap the NH₃ volatilized from the maize field. The other sponge was fitted into the top of the container to avoid contamination by atmospheric NH₃. The NH₃ trapped in the lower sponge was extracted with 1 M KCl. The N₂O was collected between 08:30 and 11:30 on the morning of each sampling day by a stainless static chamber with a pedestal (50 cm long, 50 cm wide, and 70 cm high) during the maize growing season (Guo et al., 2022). The N₂O samples were obtained using a 100-mL plastic tight syringe at 0 min, 10 min, 20 min, and 30 min after chamber closure and transferred to sealable airbags. The NH₃ and N₂O samplings were made at 15–25-day intervals, and continuous measurements lasted for 6 days with a frequency of 2 days after fertilization and irrigation during the whole maize growing season. The leachate and NH₃ samples were analyzed by a continuous flow analyzer (TRAACS2000 system, Norderstedt, Germany), and N₂O samples were analyzed by a gas chromatograph (Picarro G2308, Shanghai, China).

The life cycle inventories include the production, transportation, and application of agricultural material inputs and outputs from the perspective of an LCA (ISO, 2006; Zhang et al., 2021). The Nr losses and GHG emissions of the entire growing period for the maize system were quantified from sowing to harvest based on LCA. The system boundary is shown in Figure 2 .

A one-way variance analysis (ANOVA) of different treatments was performed using the least significant difference (LSD) test with the SPSS software (26.0 version). Tables and figures were constructed using Excel 2021 (Microsoft Corp., USA) and Origin 2024 (OriginLab, Corp., USA). The date values were means ± standard deviation (SD).

Both N management and irrigation method directly influenced the maize grain yield ( Figure 3 ). The OM and OMI considerably reduced the N fertilizer rate by 55.0% without compromising grain yield during the 2-year experiments compared to the FP treatment. The maize grain yield of OMI treatment was higher than that of OM treatment; thus, 10.2% (13.4 vs. 12.1 t ha⁻¹) in 2019 and 13.6% (11.3 vs. 9.9 t ha⁻¹) in 2020 more under flood and 3.4% (13.2 vs. 12.7 t ha⁻¹) in 2019 and 4.4% (11.9 vs. 11.4 t ha⁻¹) in 2020 under drip irrigation, respectively. The mean maize yield of OMI (12.3 t ha⁻¹) treatment was 8.2% lower than that of FP (13.4 t ha⁻¹) treatment under flood irrigation but similar to that of FP under drip irrigation (12.5 vs. 12.5 t ha⁻¹). Compared to OMI and FP, the OM yield was significantly lower under flood but not drip irrigation ( Figure 3 ). The OMI and FP yields did not differ significantly. The highest maize yield was recorded in 2019 for FP (14.5 t ha⁻¹) treatment under flood irrigation and OMI (13.2 t ha⁻¹) treatment under drip irrigation. The cost and benefits of input and output materials [fertilizers, drip tapes, irrigation, and others (seed, fuel for machinery, labor, and pesticides)] for maize production were calculated using the field experiments dataset ( Table 3 ). As shown in the table, the benefits of OMI and FP treatments under both irrigation methods were not significantly different. The net benefit of OMI treatments was lower than FP treatments under flood irrigation. Comparatively, the net benefit of OMI treatments was slightly higher than that of FP treatments under drip irrigation. Additionally, the net benefit of the drip system was slightly lower than that of the flood irrigated treatment due to the cost of drip tapes.

The electricity for irrigation, N leaching, and NH₃–N volatilization were the prime contributors to the Nr losses and NF in flood and drip irrigated maize systems ( Figures 4 , 5 ). The N leaching and NH₃–N volatilization of the FP treatment were significantly higher than those of other treatments. Figures 4A, B show that the Nr losses from OM, OMI, and FP treatments were 79.9 kg N ha⁻¹, 75.2 kg N ha⁻¹, and 129.6 kg N ha⁻¹ in 2019 and 86.0 kg N ha⁻¹, 80.0 kg N ha⁻¹, and 146.9 kg N ha⁻¹ in 2020 under flood irrigation, respectively. The Nr loss amount of various N treatments in 2020 was slightly higher than in 2019. Compared with FP treatment, the Nr losses of OM and OMI treatments were significantly reduced by 38.4% and 41.9% in 2019 and 41.5% and 45.5% in 2020, respectively. In addition, the Nr losses of OMI treatment were 5.8% and 7.0% lower than those of the OM treatment in 2019 and 2020, respectively. Under drip irrigation ( Figures 4C, D ), the Nr losses of OM, OMI, and FP treatments were 61.2 kg N ha⁻¹, 55.2 kg N ha⁻¹, and 97.4 kg N ha⁻¹ in 2019 and 63.8 kg N ha⁻¹, 60.5 kg N ha⁻¹, and 106.6 kg N ha⁻¹ in 2020, respectively. The Nr losses of the OM and OMI treatments were notably lower than those of the FP treatment, which were reduced by 37.1% and 43.3% in 2019 and 40.2% and 43.3% in 2020, respectively. The Nr losses of OMI treatment were less than 9.8% in 2019 and 5.2% in 2020 of the OM treatment, respectively.

The NF of OM and OMI was lower than the FP treatment ( Figure 5 ) without compromising the grain yield ( Figure 3 ). The NF of drip irrigation in the different N treatments was lower than that of flood irrigation. The highest NF of different treatments in 2019 and 2020 was recorded in flood irrigation, which was FP treatment (9.0 kg N t⁻¹) in 2019 and control treatment (16.7 kg N t⁻¹) in 2020 ( Figures 5A, B ). However, the highest NF of drip irrigation was recorded in FP treatment of 2019 and 2020 field experiments ( Figures 5C, D , 7.8 kg N t⁻¹ in 2019 and 8.6 kg N t⁻¹ in 2020). Under flood-irrigated experiments, the NF of OM and OMI treatments were significantly reduced by 26.4% and 37.0% in 2019 and 27.3% and 34.8% in 2020 compared to the FP treatment, respectively. The NF of OMI treatment was 14.4% and 10.3% lower than those of the OM treatment in 2019 and 2020, respectively. The NF of the OM and OMI treatments was remarkably lower than that of the FP treatment.

The GHG emissions and CF were significantly affected by irrigation methods and N management containing urease inhibitors to reduce environmental impacts ( Figures 6 , 7 ). The GHG emissions of the FP treatment were markedly higher than those in the OM and OMI treatments ( Figure 6 ). The GHG emissions of OM, OMI, and FP treatments were 4,704 kg CO₂ eq ha⁻¹, 4,552 kg CO₂ eq ha⁻¹, and 8,298 kg CO₂ eq ha⁻¹ in 2019 and 4,600 kg CO₂ eq ha⁻¹, 4,448 kg CO₂ eq ha⁻¹, and 8,164 kg CO₂ eq ha⁻¹ in 2020 under flood irrigation, respectively. The GHG emissions under drip irrigation were slightly lower than those of flood irrigation. The OM and OMI treatments were also significantly different in terms of GHG and CF. The GHG emissions of the OMI treatment were 45.1% and 3.2% lower in 2019 and 45.5% and 3.3% lower in 2020 compared to those of the FP and OM treatments, respectively. The GHG of OM treatment was 43.3% in 2019 and 43.7% lower than the FP treatment. Meanwhile, fertilizers, electricity for irrigation, and N₂O–N emissions were the dominant components of the CF under both flood and drip irrigation in the maize systems ( Figure 7 ). The CF among different N treatments in maize systems under food and drip irrigation presented similar trend to NF. The highest CF was recorded in the FP treatment and the control treatment of 2019 and 2020 under flood irrigation, and the highest CF of the drip irrigation was only recorded in the FP treatment. The CF of OM and OMI treatments were significantly reduced by 32.3% and 40.5% in 2019 and 30.0% and 34.8% in 2020 compared to the FP treatment in flood irrigation, respectively. The CF of the OMI treatment also documented a striking effect more than the OM treatment due to UI application. The CF of flood and drip irrigation presented similar outcomes and trends among the OM, OMI, and FP treatments. Compared to the FP treatment, the CF of OM and OMI treatments were significantly reduced in 2019 and 2020 under drip irrigation.

In this study, the N application rate of OMI treatment was significantly reduced while maintaining maize grain yield under flood and drip irrigation compared to the FP treatment. Drip irrigation consumed less water than flood irrigation. The principal advantages of the OMI treatment were reflected in the decent input rate of N fertilizer containing UI and irrigation methods to match maize N and water demands. The appropriate N fertilizer rates based on improved N management practices could increase crop yields with lower inevitable N losses to the environment (Cai et al., 2023; Yin et al., 2021). Our results showed that the grain yield of OM treatment was significantly lower than those of OMI and FP treatment under flood irrigation. In contrast, the maize grain yields of OM, OMI, and FP treatments presented no significant difference under drip irrigation ( Figures 3C, D ). Both N fertilizer and irrigation method were vital factors for grain yields, indicating that N and water inputs must be optimized simultaneously in crop cultivation (Du et al., 2018). However, flood irrigation systems and broadcasting of N fertilizer on the soil surface resulted in water losses by evaporation and large nutrient flow to the environment via leaching and runoff. The UI application resulted in lesser N losses to the environment and more N utilized by crops under drip irrigation than flood irrigation (Silber et al., 2003). Thus, applying N fertilizer containing UI provides a perspective to increase NUE and reduce N losses in maize systems under flood irrigation in a water-limited region of Northwest China. Drip irrigation reduced N loss and water consumption and increased N uptake, biomass accumulation, and crop yield compared to flood irrigation (Chen et al., 2020; Guo et al., 2021). Drip irrigation has excellent advantages in fertilizer input and water conservation, which can maintain favorable moisture conditions in the roots zone and facilitate the movement of nutrients to roots via diffusion. The frequencies of fertilizer and drip irrigation improve NUE and water efficiency in cropping systems (Carrijo et al., 2017; Wang et al., 2019). Drip irrigation could also enhance crop nutrient availability and uptake by synchronizing the water and nutrient supply (Si et al., 2020; Singandhupe et al., 2003).

Nitrogen footprint is mainly affected by fertilizer application, electricity for irrigation, pesticide use, fuel for machinery, labor for planting, N losses from N leaching, NH₃–N volatilization, and N₂O–N emissions (Liang et al., 2018). In our study, the NF was calculated by the agriculture system input and output indicators per ton of standard yield in agricultural production based on LCA, which could be used to assess ecosystem sustainability (Brentrup et al., 2004; Liang et al., 2018). Several studies indicated that the NF could reach 9.1 kg N t⁻¹ in the summer maize system with Nutrients Experts in Northcentral and Northeast China, 15.8 kg N t⁻¹ in the USA, and 6.3 kg N t⁻¹ in the EU (He et al., 2021; Huang et al., 2021; Leach et al., 2016). Our study showed that the NF of OM and OMI treatments under flood irrigation was similar to those reported in the EU but much lower than that in other parts of China and the USA. Our results showed that the NF of OM and OMI treatments under drip irrigation were significantly lower than those of other studies (He et al., 2021; Huang et al., 2021). Drip irrigation is an effective technology that can help optimize N management and improve the utilization of water resources compared to flood irrigation. Previous studies confirm the results of our finding that the electricity for irrigation, N leaching, and NH₃–N volatilization were dominant contributors to the Nr losses and NF of maize systems under flood and drip irrigation. The other indicators of Nr losses and NF only accounted for a small percentage ( Figures 4 , 5 ). The N fertilizer application rate is a critical indicator for influencing the N losses from N leaching, N₂O–N emissions, NH₃–N volatilization, and the NF in crop production under flood irrigation (Huang et al., 2021, 2017). Inefficient irrigation regimes are another essential indicator to control the NF of intensive irrigated agricultural systems (Nemecek et al., 2011; Wang et al., 2017). However, multiple fertilization and irrigation applications were the primary reasons for ample Nr-Mi of the NF in maize systems under drip irrigation compared to other crop systems (Grassini and Cassman, 2012; Huang et al., 2021). The average Nr losses and NF of different N treatments in drip irrigation were 24.6%–27.7% and 21.8%–44.6% lower than in flood irrigation. The average Nr losses and NF of the OMI treatment were significantly lower than those of the OM and FP treatments ( Figures 4 , 5 ). The appropriate N fertilizer rates and intelligent drip irrigation regimes were necessary to reduce higher NF in maize production regions (Han et al., 2023; Xu et al., 2020). Therefore, integrating N and drip irrigation management measures to optimize agricultural production has excellent potential for minimizing environmental impacts (Król-Badziak et al., 2020; Maltese et al., 2024; Zhang et al., 2020).

Carbon footprint is calculated based on fertilizers input, electricity for irrigation, pesticides, seed, fuel for machinery, labor for planting, N losses from N leaching, NH₃–N volatilization, and N₂O–N emissions per ton grain yield (Huang et al., 2021; Liang et al., 2018). In the present study, the GHG emissions and CF of drip irrigation were significantly reduced compared to flood irrigation. Studies have revealed that the CF in maize systems with optimized N management under flood irrigation is 436 kg CO₂ eq t⁻¹ in Northcentral Northeast China, 261 kg CO₂ eq t⁻¹ in India, and 231 kg CO₂ eq t⁻¹ in the USA (Grassini and Cassman, 2012; Huang et al., 2021; Lenka et al., 2017). In the present study, the CF of OM treatment under flood irrigation was similar to those reported in China, and the CF of OMI was higher than that in the USA and India. Comparatively, the CF of OM and OMI treatments under drip irrigation was higher than that in the USA but much lower than that of previously reported cropping systems in China. In addition, another study reported that the CF of maize under drip irrigation was significantly lower than that of rice (657 kg CO₂ eq t⁻¹) but higher than that of wheat (166 kg CO₂ eq t⁻¹) for similar yield goals (Linquist et al., 2011). Our study showed that integrating flood irrigation and UI could notably reduce N losses from N₂O–N emissions and NH₃–N volatilization and CF in maize systems. The UI is widely used in flood-irrigated agricultural systems to decrease gaseous N losses and increase NUE (Cheng et al., 2022). Our results also indicated that applying UI in flood irrigation significantly reduced total N losses to the environment in maize systems. Applying N-fertilizer-incorporated UI provides a practical way for farmers to reduce N overapplication and decrease environmental impacts under flood irrigation systems in arid and semi-arid regions. Better environmental performance inevitably improves the sustainability and economics of maize production. Drip irrigation performed better than flood irrigation in decreasing negative environmental impacts in crop production systems. Drip irrigation reduces the amount of N and water requirements and increases N uptake, biomass accumulation, and crop yields (Chen et al., 2020; Guo et al., 2021), which is consistent with our findings. Drip irrigation has broader application prospects in water-constrained areas to conserve natural resources and sustain crop production. Drip irrigation of different crops has been proven to reduce N rate and Nr losses compared with flood irrigation in arid and semi-arid regions (Di et al., 2024; Wang et al., 2021). Although drip irrigation has better environmental performance, adopting drip irrigation should be based on the crop production region’s climate and economic conditions.

The calculation of CF is directly affected by many driven indicators in agricultural systems based on LCA. Most relevant studies indicated that N fertilizer, N leaching, and N₂O–N emissions are the leading indicators influencing the CF of crop production under flood or furrow irrigation (Huang et al., 2021; Zhang et al., 2021). Our study revealed that fertilizers input, electricity for irrigation, and N₂O–N emissions were the critical factors controlling the GHG emissions and CF in flood- and drip-irrigated cropping systems. We also confirmed prior findings that the N fertilizer and electricity for irrigation are important indicators in controlling the CF of crop systems under drip irrigation (Nemecek et al., 2011; Wang et al., 2017). The N₂O–N emission levels were also an important factor in controlling GHG emissions and CF because the N₂O–N emission is 298-fold greater than the CO₂ coefficient in terms of the effects on global warming (Huang et al., 2021; IPCC, 2019). Our results were consistent with previous studies that UI use significantly reduced N₂O–N emissions under flood irrigation (Cheng et al., 2022; Elrys et al., 2021). In addition, fertilization timing and irrigation rates of drip irrigation were the primary contributors to the large GHG-Mi of the CF in maize systems under drip irrigation compared with those of other crop production models (Grassini and Cassman, 2012; Huang et al., 2021). Therefore, agricultural production should consider rational N and efficient water management to match crop requirements and regional characteristics in the future.

Although the integration of optimal N management and drip irrigation technology in the present study showed significant effects for improving environmental impacts, there were also some inevitable limitations. The Nr losses and GHG emissions due to agricultural material input and output results were highly dependent on the completeness and accuracy of the life cycle inventories. Our study mainly used the emission factors of life cycle inventories based on agricultural materials input and output in China, which have been widely used in earlier LCA research (Huang et al., 2021; Liang et al., 2018). However, the life cycle inventories and emission factors involve many industry and agriculture sectors, which are quite challenging to collect. Despite these limitations and uncertainties, our study considered the characteristics of environmental footprints to provide meaningful analysis for distinguishing maize production under flood and drip irrigation. Improving N management and irrigation methods can eventually achieve a win–win outcome in mitigating environmental impacts. However, this process may be slow, as the required technology is not yet widely available in China.