Introduction

Nitrogen fertilizer input in China’s agricultural production is large, which not only consumes huge wealth but also brings huge environmental risks due to the low nitrogen utilization rate in the current season. A large part of it is discharged into the atmosphere in the form of ammonia and N2O, becoming a major source of agricultural non-point source pollution (Yang et al. 2015b; Yang et al. 2020a). As reported by Vilarrasa-Nogué et al. (2020), agriculture attributed to 10–12% of global man-made greenhouse gas (GHG) emissions. The N2O emissions from croplands are strongly controlled by (1) environmental factors such as atmospheric temperature, relative humidity, and precipitation (IPCC 2013); (2) soil edaphic properties such assoil temperature and water content as well as availability of mineral nitrogen etc. (Hu et al. 2013); and (3) agricultural resource such as nitrogen and manure utilization, tillage management, straw application, and also crop rotations (Qin et al. 2012; Yang et al. 2020a; Cheng et al. 2020).

Since 1990s, the overuse and the influence of nitrogen fertilizer application in China were begun to emerge on the ecological implications and the reason for nitrogen use (Kahrl et al. 2010). According to achievements from numerous researches, the range of 150–200 kg N ha−1 is considered the optimum yield of food crops in China (Xin et al. 2019). On the other hand, in past decades, straw produces reached hundreds of million tons per year in China (Yang et al. 2015b), and most farmers burn straw in their fields to reduce processing time and costs. Straw mulching for power generation faces a series of technical, economic, and management problems and risks. Straw mulching is an appropriate and effective treatment channel to alleviate the negative effects of straw mulching (Lu et al. 2010; Xin et al. 2019).

In addition, the study of long-term field has shown that application of suitable fertilizer and straw mulching could promote soil organic carbon content (Huang et al. 2013a; Yang et al. 2020b; Yang et al. 2020c). However, these measures might also give impetus to N2O emissions via the elevation of soil nitrifier and denitrifier substrates simultaneously (Zhang et al. 2019; Vilarrasa-Nogué et al. 2020). Furthermore, straw return increases annual mean N2O emissions by 27.9% from field research in suburban Beijing (Huang et al. 2013b). However, crop residue return has several advantages to increase carbon (C) sequestration in soil (Zhang et al. 2010; Huang et al. 2013a; Yang et al. 2020a) followed by increasing grain yields (Cheng et al. 2020), but some researches showed no impact on crop yield (Qiu et al. 2012). Qiu et al. (2015) reported the reverse effect of adding plant-derived dissolved organic matter (DOM) increased the carbon dioxide (CO2) and nitrous oxide (N2O) emissions, and accelerated the decomposition of soil organic carbon, as well as reduced the soil carbon sink. Therefore, it is very essential to evaluate the net mitigation effect based on greenhouse gas emission response of straw mulching.

In order to minimize the gaseous emission from farmland in China, the optimal nitrogen application should be adopted first (Shi et al. 2014; Yang et al. 2019a) and, secondly, applied mixture and synthetic fertilizers with nitrification inhibitors (Akiyama et al. 2010); then, slow release fertilizers should be considered (Jiang et al. 2010; Huang et al. 2013b). Due to the inhibition of the first stage of nitrification (NH4+ oxidation to NO2−) by suppressing the bacterial enzyme activities, the 10% use of DCD reduced N2O emission 42–82% and also can increased herbage production (Gillette et al. 2018; Thers et al. 2020). The slow release fertilizer (SRF) could promote the nitrogen uptake efficiency and increase the crop yields in several production systems. The results by Geng et al. (2015) stated that the seed cotton yields in the SRF treatments were increased by 15–18% compared with urea application from cotton production in China. Besides, the slow release fertilizers also contribute towards environmental pollution control by means of the mitigation of hazardous GHG emissions and water eutrophication (Trinh et al. 2015; Yang et al. 2019b).

Although the greenhouse gas emission scale and worldwide distribution from natural to agricultural were estimated, the emission fluxes of Chinese rainfed soil condition under monocropping systems in Northwest China region are still poorly defined. China has 57.6 million hectares of dry land, about 4% of the world’s total arable land (Cheng et al. 2020). The arable lands from the Northwest China obtained resembling climatic condition with the Northern China Plain. The unique planting system in this region is irrigation rotation between winter wheat and summer maize, and monocropping with winter wheat or maize also occurred under rainfed farmland (Yang et al. 2015b). Campanha et al. (2019) pointed out the N2O emission is non-significantly distinct between irrigated soil and rainfed sandy loam in semi-arid areas of northwest Kenya.

However, the precipitation instead of irrigation may become a vital additive factor controlling net greenhouse gas (NGHG) emission and net greenhouse gas intensity (NGHGI) in rainfed monocropping system. The unique characteristics of soil and climatic conditions associated with winter wheat monocropping system under rainfed condition became attractive to examine the dynamics in GHG emissions correlated by optimized handling approach. Therefore, present investigation aimed (i) to study the influential factors on N2O emissions and consequential factors affecting NGHG emission as well as NGHGI under rainfed farmland, (ii) to evaluate the value of precipitation in terms of climate change governed on N2O emission fluxes within the cropping cycle, and (iii) to anticipate a better agricultural management practice that can reduce NGHG emissions with maintained or increased grain yields under rainfed monocropping systems.

Materials and methods

Study site description

A experiment of a 3-year field begun at wheat growing season in October 2012 and ended in July 2015. This research experiment center site is located in Chinese National Soil Fertility and Fertilizer Efficiency Monitoring Base of Loess Soil in Wuquan Town, Yangling District, Shaanxi Province, Northwest China. In this site, the soil type was loess soil comprising 32% of clay, 52% of silt, and 16% of sand. The content of soil organic matter, total N, and Olsen-P were 17.1 g kg−1, 0.93 g kg−1, and 15.0 mg kg−1; exchangeable K was 191 mg kg−1 with pH 7.44 across all plots. Temperature and precipitation were 12.9 °C and 550 mm, in which the precipitation is mainly achieved from July to September.

Experimental design and field management

In this study, there were five treatments and they were no N application as control (N0), conventional N management (NCon), optimized N management (SNOpt) in which its distinct characteristic was an optimized N rate for crop coupling with straw incorporation, optimized N management adding 5% of dicyanodiamide (DCD) as nitrification inhibitor (SNOpt + DCD), and optimized N management of slow released fertilizer (SSRFOpt). It was conducted under completely random block design with three replications, and each plot was 30 m2 (for 6 m long and 5 m wide). The type of Triticum aestivum L. cv. Xiaoyan 22 was main winter wheat and planting according to 120 kg ha−1 sowing rate and the row spacing of 20 cm. All fertilizers were applied as basal fertilizer. For NCon, the fertilizer application amount was 220 kg N ha−1, with local conventional fertilization method, and the optimized N rates for SNOpt, SNOpt + DCD, and SSRFOpt treatments were 150 kg N ha−1 which was estimated in line with crop nitrogen demand. Urea (46% N) fertilizer was the main source of N fertilizer except for SRF in this study. Phosphorus fertilizer (superphosphate) was utilized with the rates of 90 kg P2O5 ha−1 in each treatment, and the rates of potassium fertilizer (potassium sulfate) to be applied were calculated due to that contained in SRF as 47 kg K2O ha−1. Due to NPK composition ratios (26:6:8) in SRF, it was calculated to reach optimized N content of 150 kg N ha−1 and then added a required amount of P2O5 for the SSRFOpt treatment. The planting system in this area is winter wheat-summer maize irrigation system. This experiment was performed under rainfed wheat-summer fallow monocropping system. In addition, weeding was carried out manually during vegetative growth stages in the spring and rouging before harvest.

Greenhouse gas emission measurements

N2O emission and CH4 uptake were in three winter wheat cultivation cycles from 2012 to 2015; closed static chamber method was used for manual measurement (Mosier et al. 2006). Stainless steel base rings (50 × 50 × 20 cm) are inserted into the soil in each plot before planting the wheat. They were removed and reset again only once in a year at the period of tillage for winter wheat. In order to collect samples, the chambers (50 × 50 × 50 cm) were set onto the base collars adding some water to ensure air-tight condition. Immediately after the chamber was enclosed, the initial air sample was obtained from the top of the chamber by using a plastic syringe (50 mL) through a three-way value and a ptfetube connected with the chamber. In total, four samples were obtained from each treatment at 0, 15, 30, and 45 min with time intervals of 15 min. Daily and frequent measurements were carried out after fertilizer application and precipitation. The time of collecting air was 08:30–11:00 am of local time to obtain more representative results of gas emission fluxes.

An improved gas chromatograph (Agilent, 7890A, USA) equipped with a 63Ni-electron capture detector (ECD) operating at 350 °C was adopted. High-purity nitrogen and 10% CO2 as the supporter and complementary gas for N2O analysis (Zheng et al. 2008) were used. Calibration was performed by standard calibration gases before each measurement was performed, and air conditioning was used in the laboratory to stabilize the temperature at the time of measurement to decrease the tendency of ECD to change with temperature. In order to calculate the N2O emission flux, linear regression or non-linear method was used to calculate the N2O flux according to the variation rule of the concentration of the air volume above the closed chamber. It is calculated through Eq. (1) below.

$$ F={k}_1\times 273/\left(273+T\right)\times M/V\times \mathrm{H}\times dc/ dt $$
(1)
$$ c=a+ bt\ \left( dc/ dt=b\right) $$
(2)
$$ c=a+ bt+{dt}^2\ \left( dc/ dt=b\right) $$
(3)

where F (μg N2O–N m−2 h−1 or μg CH4–C m−2 h−1) is the net flux; k1is the dimensional conversion factor (0.001); and T (°C) presents the average temperature of the room. M (28 g or 12 g of N2O–N mol−1 and CH4–C mol−1for N2O or CH4) is the molecular weight of N2 in the N2O, C in CH4; V is the volume of mole (22.4 L mol−1); H (m) is the height of chamber and c (μL L−1) is the mixing volume ratio of nitrous oxide and methane; t (h) is the chamber closing time; dc/dt (μL L−1 h−1) is the initial rate of changes in concentration for N2O or CH4 in the gas chamber; and a, b, and d are the parameters obtained by fitting the linear or non-linear change function of concentration with time (Gao et al. 2014). The fluxes of N2O and CH4 are expressed as the average of three repeated measurements.

Estimation of N2O emission

By adding up the daily fluxes of the sampled and non-sampled days, the annual cumulative N2O emissions were calculated, and then, the linear interpolation between the time intervals was used for estimation (Mosier et al. 2006). The NGHG emissions and NGHGI were estimated and calculated in line with Eqs. (4) and (5). Additionally, the N2O emission factor (EF) derived from N fertilizer was determined by mean of Eq (6).

$$ \mathrm{N}\mathrm{GHG}={\mathrm{N}}_2\mathrm{O}\hbox{--} \mathrm{N}\times 44/28\times 298+{\mathrm{CH}}_4\hbox{--} \mathrm{C}\times 16/12\times 25 $$
(4)

where N2O–N and CH4–C units are kilograms of N2O–N per hectare and kilograms of CH4–C per hectare, respectively. NGHG was calculated by CO2 equivalent (CO2-eq). In the 100-year time range, 1 kg of N2O and CH4 are equal to 298 and 25 kg of CO2 respectively for the global warming potential (Forster et al. 2007).

$$ \mathrm{NGHGI}=\mathrm{NGHG}/\mathrm{grain}\ \mathrm{yield} $$
(5)

where grain yield unit is tons per hectare.

$$ \mathrm{EF}=\frac{\sum \left({\mathrm{N}}_2\mathrm{O}-\mathrm{amendment}\right)\kern0.5em -\sum \left({\mathrm{N}}_2\mathrm{O}-\mathrm{control}\right)}{\mathrm{Total}\ \mathrm{Input}\ \mathrm{N}}\times 100\% $$
(6)

where EF is the N2O emission factor (N2O–N emission as a percentage of N application); N2O-amendment and N2O-control are the cumulative N2O emissions from the N application and the plots of control, respectively (kg N2O–N ha−1); and total input N is the amount of N fertilizer applied (kg N ha−1).

Soil analysis

Soil temperature (10 cm of topsoil) and indoor air temperature were measured by digital thermometer immediately before and after collecting air samples. To measure soil moisture contents, samples were collected at 0–20 cm of top soil in each plot, and about 20 g of soils was dried in an oven at 105 °C for 24 h. The soil water-filled pore space (WFPS) was estimated by Eq. (7).

$$ \mathrm{WFPS}=\frac{\mathrm{Soil}\ \mathrm{water}\ \mathrm{content}\ \left(\%\right)\kern0.5em \times \kern0.75em \mathrm{Soil}\ \mathrm{bulk}\ \mathrm{density}}{1\kern0.75em -\kern0.75em \frac{\mathrm{soil}\ \mathrm{bulk}\ \mathrm{density}}{2.65}}\times 100\% $$
(7)

Grain yield and N use efficiency

A representative sampling area of 1 m2 was assigned in each plot to harvest for wheat grain yield and harvest index (HI). The dry weights of straws were measured as above ground biomass, and the wheat grain yield was recorded at approximately 15% moisture content in grain. The N use efficiency (NUE) % was estimated by Eq. (8) as described by Ladha et al. (2005).

$$ \mathrm{NUE}=\frac{\sum \left(\mathrm{Plant}\ \mathrm{N}-\mathrm{amendment}\right)\kern0.5em -\sum \left(\mathrm{Plant}\ \mathrm{N}-\mathrm{control}\right)}{\mathrm{Total}\ \mathrm{Input}\ \mathrm{N}}\times 100\% $$
(8)

where NUE is the N use efficiency (%); plant N-amendment and plant N-control are the accumulative N uptake by plant (N from grain + N from straw) at harvest from the N application and control plots, respectively; and total input N is the rate of N fertilizer applied (kg N ha−1).

Statistical analysis

The raw data was calculated by Microsoft Excel 2010 spreadsheet. The difference among treatments was examined via ANOVA and the Sigma plot 12.5 (Syst at Software Inc., Erkrath, Germany) was used to compare the mean to calculate the minimum significant difference (LSD) at the 5% level and to draw a chart.

Results

Precipitation, soil temperature, and soil WFPS percentage

The temperature of air and 10 cm of topsoil as well as the precipitation during the study period are presented in Fig. 1. The precipitation obtained was 233.2 mm and 225.8 mm in 2012–2013, 296.3 mm and 398.8 mm in 2013–2014, and 237.0 mm and 283.8 mm in 2014–2015, respectively, during wheat growing and fallow season (Table 1). The relatively largest quantity of total precipitation occurred in 2013–2014 cropping cycle during this study. Temperature for 10 cm of topsoil varied with daily mean atmospheric temperature, and it ranged from − 1.2 to 30.0 °C with mean value of 14.4 °C. Although soil temperatures at 10 cm depth were certainly similar between straw and without straw management condition, surface soil temperatures under without straw return treatment relatively reflected to atmospheric temperature rather than those under straw return (data are not shown). It was observed that straw could maintain surface soil temperature warmer in winter as well as could somewhat reduce surface soil temperature in summer. As this field research was done under rainfed dryland, soil WFPS in 20 cm of topsoil was enhanced by precipitation and declined by evaporation and other physiological processes (Fig. 2). In this study, soil WFPS% at each sampling point was below 60%, i.e., the highest value of WFPS% was 55.99% and the lowest was down to 6.98% during July drought.

Fig. 1
figure 1

Atmospheric temperature (°C), soil temperature at 10 cm depth (°C), and precipitation (mm) during field study. Wheat cropping season and fallow season are denoted by W and F, respectively

Full size image
Table 1 Comparison of cumulative N2O emission by each treatment as associated with precipitation observed in winter wheat-summer fallow monocropping system under rainfed. The number represents mean ± standard error (n = 3); different letters within same column indicate significant differences (p < 0.05)
Full size table
Fig. 2
figure 2

Dynamic of soil water-filled pore space (WFPS%) in the top 20 cm depth of soil during the study period under rainfed: Wheat cropping season and fallow season are denoted by W and F, respectively; error bars represent standard deviation of the mean (n = 3)

Full size image

N2O emissions from soil

The fluxes of N2O emission for all treatments from October 2013 to September 2015 are shown in Fig. 3. The mean values of N2O emission fluxes varied from − 5.49 to 281.91 μg N2O–N m−2 h−1 throughout this study. The largest gap of the N2O emissions occurred as ranging from − 22.91 to 456.07 μg N2O–N m−2 h−1 in 2013–2014 winter wheat growing season due to greater quantity of precipitation. The relatively higher peaks of emission were observed within one and half months after basal fertilization in each circle, and some differences were significant among the treatments during the periods. Then, the emission was slightly low with lower temperature in the winter. The highest N2O emissions were produced from the SNOpt treatment during few days after fertilization, and the maximum peaks of mean emissions were detected in 5 November 2013 (20 days after basal fertilization) with 281.91 μg N2O–N m−2 h−1 and in 29 October 2014 (13 days after basal fertilization) with 84.27 μg N2O–N m−2 h−1, in first, second, and third cycles, respectively. As this study was done in rainfed field, some small peaks of N2O fluxes were formed in spring before wheat was harvested and in the fallow season of summer associated with warmer temperature and frequent precipitation.

Fig. 3
figure 3

Mean N2O emissions from winter wheat-summer fallow cropping system under rainfed: Wheat cropping season and fallow season are denoted by W and F, respectively; error bars represent standard deviation of the mean (n = 3); upward line arrows indicate basal fertilizer application and straw dressing for winter wheat; downward arrows indicate remarkable precipitation (≥ 8 mm) during this study

Full size image

The mean cumulative N2O emissions were ranging from 0.36 to 0.60, 0.44 to 1.60, and 0.15 to 1.11 kg N2O–N ha−1 year−1 in 2012–2013, in 2013–2014, and in 2014–2015, respectively (Table 2). The values of annual cumulative N2O emissions in 2013–2014 were relatively the largest among a 3-year investigation. However, the similar trends of results on cumulative N2O emission through different management practices were observed in later 2 years, i.e., the annual cumulative N2O emission by all treatments was in the order of SNOpt > SSRFOpt ≥ NCon > SNOpt + DCD ≥ N0 treatments, with an exception of first year comprehending without any significant difference among the treatments. In fact, N2O emissions during the fallow season had no significant difference (ranged of 0.14–0.33 kg N2O–N ha−1) among the treatments over a 3-year field study, regardless of the amount of precipitation obtained (Table 1). In 2013–2014 and 2014–2015 cropping season, compared with the NCon treatment, cumulative N2O emissions were increased about double in the SNOpt treatment, whereas the SNOpt + DCD and SSRFOpt treatments decreased N2O emissions with 72–86% and 11–39% compared with SNOpt among straw return treatments. The N2O emission factors (EF) derived from N fertilizer were ranging from 0.03 to 0.76% in this study (Table 2). Due to highest rainfall obtained in 2013–2014, the highest EF which was almost double of that over a 3-year period was observed. Among them, the lowest value was determined in the SNOpt + DCD treatment, and the highest value occurred in the SNOpt treatment. Additionally, the EF of the NCon treatment was greater than that of the SSRFOpt treatment.

Table 2 N2O emission factor derived from N fertilizer and N use efficiency depended upon different management practices. The different letters within same column indicate significant differences (p < 0.05)
Full size table

NGHG emission and NGHGI

The CH4 is very small sink for dry cropland in China northern plain, and CH4 uptake makes partial contribution when calculating the net global warming potential (NGWP) from all emission and sinks (Huang et al. 2013b). Mean atmospheric CH4 uptake by soil was approximately 7 and 14 μg CH4–C m−2 h−1in 2013–2014 and 2014–2015, respectively. No significant difference on either daily or annual CH4 uptakes by soil was found among the treatments (Fig. 4). The mean annual CH4 uptakes by soil were ranging from 0.96 to 1.36, 0.33 to 0.69, and 0.98 to 1.31 kg CH4–C ha−1 year−1in 2012–2013, 2013–2014, and 2014–2015, respectively (Table 3). In this study, it was discovered that soil CH4 uptakes were decreased by higher precipitation in 2013–2014. These values were very close to the uptake of 0.9 (Hu et al. 2013) and 1.1 to 1.3 kg CH4–C ha−1 year−1 (Gao et al. 2014).

Fig. 4
figure 4

Mean CH4 emissions from winter wheat-summer fallow cropping system under rainfed: Wheat cropping season and fallow season are denoted by W and F, respectively; error bars represent standard deviation of the mean (n = 3); upward line arrows indicate basal fertilizers application and straw dressing for winter wheat; downward arrows indicate remarkable precipitation (≥8mm) during this study.

Full size image
Table 3 Net greenhouse gas and net greenhouse gas intensity in winter wheat summer fallow cropping system under rainfed. The number represents mean ± standard error (n = 3); different letters within same column indicate significant differences (p < 0.05)
Full size table

The NGHG and NGHGI are exhibited in Table 3. NGHG of the SNOpt treatment was the highest and about 39% higher than that of the NCon treatment over the 3-year period while it was 57% higher in 2013–2014 and 48% in 2014–2015 although NGHG of straw management treatments (SNOpt, SNOpt + DCD, and SSRFOpt) did not increase compared with conventional practice (NCon) in the first year 2012–2013. On the other hand, compared with the SNOpt treatment among straw management practices, NGHG of SNOpt + DCD and SSRFOpt were reduced by approximately 26% and 1% in 2012–2013, 59% and 5% in 2013–2014, and 70% and 31% in 2014–2015, respectively, while they were decreased by 57% and 12%, respectively, over a 3-year period. Similarly, the NGHGI was reduced by 61% and 14% in the SNOpt + DCD and SSRFOpt treatments, respectively, compared with SNOpt among straw return treatment which has the highest increase of NGHGI withstanding 41% higher than the NCon treatment over a 3-year period.

Wheat grain yield

As there was no N fertilizer for the N0 treatment, the mean of winter wheat grain yields for N0 was significantly low at 3.84, 4.72, and 4.26 t ha−1 in 2013, 2014, and 2015, respectively (Table 3). The grain yields of the other treatments were not significantly different in this study, except in 2013, SNOpt + DCD produced significantly higher yield than the NCon treatment. The grain yield of the SNOpt + DCD treatment always showed the highest in each cropping cycle by 6.26, 6.70, and 6.59 t ha−1 in 2013, 2014, and 2015, respectively. Overall, when the mean N use efficiency (NUE) affected by the treatments was compared over a 3-year study, it was observed that the SNOpt + DCD treatment had the highest N use efficiency with 33.5%, followed by the NCon treatment with 32.4%, SSRFOpt with 29.4%, and SNOpt with 27.7%, respectively (Table 2).

Discussion

N2O emissions from rainfed drylands by monocropping system

The patterns and processes of N2O emissions have been explained by several studies that N2O emission peaks can occur within 1 or 2 weeks after N fertilizer application and irrigation or rainfall combined with high soil temperature and moisture content (Cui et al. 2012; Hu et al. 2013; Huang et al. 2013b). The N2O emission fluxes were mainly attributed by N fertilizer application throughout this study, and our results fell well within the range reported by those earlier studies. Annual cumulative N2O emissions were ranging from 0.29 to 1.64 kg N2O–N ha−1 and in line with the result of 0.20 to 4.54 kg N2O–N ha−1 from double-cropping cereal rotation system in North China plain (Huang et al. 2013a). In our study, the cumulative N2O emissions during the fallow seasons under a 3-year field experiment vacillated from 0.14 to 0.33 kg N2O–N ha−1without any significant difference among the treatments no matter what the temperatures, soil moisture, and frequency and amount of rainfall were. The emissions from fallow land by treatment without N fertilizer application could be reduced from 5 to 25% in 2012–2013, 47 to 130% in 2013–2014, and 45 to 88% in 2014–2015, respectively, compared with treatments with N fertilizers. It would be because the availability of soil mineral N (NO3−–N and NH4+–N) in N-fertilized soil was slightly higher than that in soil without N fertilizer.

However, soil mineral N content with N2O emissions were not presented with significant correlation in the present study and it was granted by Gao et al. (2014). On the other hand, the mean values of N2O emission fluxes from our rainfed field ranging from 10 to 25 μg N2O–N m−2 h−1in the winter wheat cropping season were slightly lower than those from 15 to 30 μg N2O–N m−2 h−1from irrigated winter wheat in North China plain by Gao et al. (2014). The cumulative N2O emissions during the winter wheat cropping season except in 2012–2013 had significant differences among the treatments depending mainly upon straw incorporation, amount of precipitation, and rates and types of N fertilizers including nitrification inhibitor. Compared with treatment without N fertilizer, the cumulative N2O emissions were 1–5 times increased by N fertilization and straw incorporation. More than half of annual cumulative N2O were emitted during the winter wheat cropping season and the similar statistical differences of annual emissions occupied in this period. It indicated that cropping season plays as a responsible period for the investigation of N2O emissions since the effectiveness of different management practices on emissions were mainly observed during the winter wheat cropping season (Tables 1 and 3).

In this study, the N2O emission factor derived from N fertilizer varied from 0.03 to 0.76% with the mean value of 0.40 ± 0.08% due to the different managements. This is well below the default emission factor, but within the range of fertilized fields in Canada, it is 0.03–1.45% (Ma et al. 2010), near the ranges from upland in China with 0.40–1.54% (Shepherd et al. 2015), and from unplanted and planted soil in China with 0.19 to 0.3% (Ni et al. 2012). In general, N2O emissions are typically based on the IPCC’s 2006 default emission factor, which is 1% of the amount of nitrogen applied to mineral fertilizers, manure, and crop residues (Zhang et al. 2019). The starting point for this method was a 2-year monitoring study of Velthof in which the emission factor of grassland on sandy soil with calcium ammonium nitrate was exactly 1%. However, there are significant differences in N2O emission factors due to distinct environment conditions, crops (grassland, cultivated land, crop residues), and management (fertilizers and fertilizers, application amount, application time) (Badagliacca et al. 2018; Campanha et al. 2019; Thers et al. 2020). However, Gillette et al. (2018) indicated the lower emission factor by 0.02% from a rainfed cropped soil in a semi-arid region.

Factors regulating N2O emission under rainfed with winter wheat-summer fallow monocropping system

N2O emission from rainfed dryland is weather-dependent as the annual N2O emissions from rainfed winter wheat monocropping system were largely correlated to the annual mean of daily maximum air temperature, annual rainfall, and the rates of N fertilizer application (Zhang et al. 2019; Campanha et al. 2019). In this study, we could not find significant relationship among soil temperature, WFPS, and N2O emission, and the similar result was reported by Gao et al. (2014). On the other hand, some scientists reported that N2O emission was closely influenced by rainfall if it is above 40 mm (Ma et al. 2010; Liu et al. 2011). Although some small peaks of N2O emission fluxes without any significant difference were performed during the fallow seasons associated with warmer temperatures and precipitation, the remarkable differences of emissions among the treatments were observed during winter wheat cropping seasons. In that periods, compared with control (N0) treatment, the mean cumulative N2O emissions from N-fertilized treatments of NCon, SNOpt, and SSRFOpt except SNOpt + DCD were significantly increased by 1.3 times, 3.6 times, and 3.2 times in 2013–2014, and by 1.8 times, 4.6 times, and 2.4 times in 2014–2015, respectively. Unfortunately, we could not find any significant difference for N2O emissions among the treatments in the first cropping cycle 2012–2013, but compared with N0 treatment emissions, they increased by 78%, 75%, 2%, and 49% from the NCon, SNOpt, SNOpt + DCD and SSRFOpt treatments, respectively. We observed that the N2O emissions could be effectively minimized by adding 5% of DCD into N fertilization as the SNOpt + DCD treatment under field condition.

It has been understood that the amount of N fertilizer governs the fluxes and cumulative N2O emissions. Huang et al. (2013a) reported that the 56% reduction of N fertilizer rate to an optimized level noticeably reduced 40% of N2O emission. In this study, the N fertilizer rates in optimized management practice reduced into 32% lower than conventional management practice. However, cumulative N2O emissions from optimized practice were considerably increased by over 2-folds in 2013–2014 and over 2 and a half folds in 2014–2015 higher than that from conventional practice during cropping season. It indicated that straw incorporation under optimized practice extremely enhanced N2O emissions instead of decrease in emissions associated with lower N fertilizer rates. The similar results occurred by 58% increase of N2O emission for summer maize after wheat straw return (Liu et al. 2011), by 27.9% increase of annual mean N2O emission in straw return and 4 times higher in straw incorporation than in straw removal treatments (Huang et al. 2013a). This phenomenon might be owing to accelerated of microbial nitrification and denitrification processes during straw decomposition together with mineral N and available carbon, as well as suitable temperatures and soil moisture (Vilarrasa-Noguéet al. 2020). Moreover, during cropping seasons, the slow release fertilizer somewhat reduced N2O emission by 15%, 11%, and 39% in 2012–2013, 2013–2014, and 2014–2015, respectively, compared with SNOpt treatment. The similar result was observed by a 33% decrease in N2O emission during the winter wheat cropping season when slow release fertilizer was used (Shi et al. 2014).

A 50-mm rewetting after an 18-day drought increased N2O fluxes by 24-folds in non-conservation reserve program (Gelfand et al. 2015). Therefore, frequency and amount of precipitation, followed by daily temperature, are substantial controlling factors on N2O emission under rainfed. In this study, rainfall events also regulated the N2O emissions particularly during cropping season after fertilizer application though the mean values of cumulative N2O emission within the fallow season which did not change evidently even when higher precipitation was obtained. The amount of rainfall (almost 300 mm) received during the cropping season in 2013–2014 was over 25% more than the first and third cropping seasons, and underwent into 1–2-fold increase in N2O emissions (Table 1). It could be interpreted that large amount of precipitation with the combination of N fertilizers and straw incorporation under 9–11 °C of soil temperatures can be assumed as a favorable condition for increasing N2O emissions. Eventually, it can be evaluated through this study that the effectiveness of the treatments for N2O emissions could be determined during the cropping season. Moreover, the amount of precipitation obtained during cropping season plays as important impact in controlling N2O emissions under a rainfed monocropping system.

Influences of straw and N fertilizer on grain yield and N use efficiency

In this study, the grain yield varied from 4.26 to 6.70 t ha−1, and it was within the range reported by Zhang et al. (2013) with 4.9 ± 2.0 t ha−1 from farmers’ practices regarding N fertilizer use in China. At present study, although the optimized practice saved 32% of N fertilizer, the grain yields were not reduced noticeably compared with a conventional practice. Moreover, the SNOpt + DCD treatment produced somewhat higher grain yield with the highest N use efficiency of 33.5% in each winter wheat cropping season. It thus seemed that adding 5% of DCD into N fertilization can inhibit nitrification to prevent losses from N fertilizer as N2O emissions and lead to have better N use efficiency from applied fertilizers for grain yield increase by 4–5%.

Adaptation of rainfed monocropping system to global warming

Agricultural management approach changes one type of source or sink of global warming potential may also influence other sources or sinks and thus change the NGHG and NGHGI (Shang et al. 2011; Huang et al. 2013b; Xin et al. 2019). In this study, compared with the NCon treatment over a 3-year period, the SNOpt treatment with a 32% decrease in N fertilizer brought about 38% increase in N2O emission by straw return, about 28% increase in soil CH4 uptakes, and eventually approximately 39% and 41% increase in NGHG and NGHGI, respectively. Despite the lower N fertilizer input, N2O emission was higher but not significant due to straw return, leading to increase in NGHG and NGHGI. However, the soil CH4 uptake influenced by N fertilizers and straw in Northern China Plain is very small (Liu et al. 2012) and the negative global warming potential from CH4 uptake represented less than 1% for all calculations (Huang et al. 2013a). Again, comparing among the optimized practices with different N management, adding 5% of DCD into urea fertilizer contributed over 55% decrease in N2O emission, over 10% increase in grain yield, and leading to about 57% and 61% decrease in NGHG and NGHGI, respectively. The use of slow release fertilizer conveyed about 13% reduction in N2O emission, approximately 3% increase in yield, and then concerned with about 12% and 14% decreased in NGHG and NGHGI, respectively.

The positive annual NGHG and NGHGI for all treatments in the present study was consistent with some previous field and modeling studies. The NGHGI values in our study were 86–223 kg CO2-eq t−1 grain, and definitely lower than the winter wheat-summer maize cropping system, where the values were 252–393-kg CO2-eq t−1 grain (Huang et al. 2013a) because of higher emission factor from irrigated cropland. However, our values were very close to 230-kg CO2-eq t−1 grain from irrigated maize in central Nebraska (Grassini and Cassman 2012) owing to low emission factors. The NGHGI over a 3-year period increased 50% to over 2-folds due to different N fertilizer applications coupling with straw return relative to N0, although SNOpt + DCD could completely minimize N2O emission factor. As this present study was carried out during winter wheat-summer fallow monocropping system under rainfed, the NGHGI is absolutely lower than the results of a 3-year field research from rice-wheat annual rotation systems in China which reported that straw management treatments due to different fertilizer applications compared with control were significantly increased by 4–12-folds in the NGHGI (Yang et al. 2015a).

The similar results were reported by Soltani et al. (2013) that 433–1612 kg CO2-eq ha−1 and 173–474 kg CO2-eq t−1 of NGHG and GHGI were obtained from wheat production in Iran. Jin et al. (2014) also suggested that application of manure and synthetic N fertilizers could remarkably decrease NGWP and NGHGI in arid agricultural systems. Therefore, since winter wheat-summer fallow monocropping system under rainfed may reduce considerably NGHG and NGHGI, our results suggest that straw incorporation in optimized management practice enhances the NGHG and GHGI through higher N2O emission factor dependence upon frequency and amount of precipitation.

The relationship between soil temperature, WFPS, N2O, and CH4 emission

Surface soil temperature of 10 cm fluctuates with atmospheric temperature (range − 1.2 to 30.0 °C, mean 14.4 °C). Soil WFPS increased with the occurrence of rainfall and then began to decline due to the soil texture, soil moisture evaporation, and crop absorption. It ranged from 14.21 to 58.29%. N2O emissions began to increase and then decreased with fertilization, and rainfall had a significant impact on N2O emissions, especially the SNOpt treatment and the SSRFOpt treatment. In this study, CH4 emissions decreased to a certain extent in response to rainfall.

Jiang et al. showed that there was no significant relationship between soil temperature, WFPS, and N2O release (Jiang et al. 2017); the peak of N2O emissions is after a heavy rain. Studies have found that when the soil WPFS is below 50% and the soil temperature is below 10 °C, the N2O emissions are relatively low (Hu et al. 2013). Studies have also shown that soil WFPS is higher than 60%, and soil temperature is higher than 10 °C, which is conducive for the release of N2O (Ma et al. 2010). In the corn-growing season, there was no obvious relationship between soil temperature and N2O release, probably because the high temperature in the corn season was not a factor limiting the production of N2O. At the same time, there is no correlation between soil WFPS and N2O release, because the main factor limiting N2O emissions is available N, so the peak of N2O emissions generally occurs after fertilization accompanied by rainfall (Ma et al. 2010). For CH4 absorption, moderate WFPS (20–60%) is conducive for CH4 absorption, and studies have shown that temperatures below 10 °C will inhibit CH4 absorption (Yang et al. 2020a). In dryland, the release of soil N2O and CH4 is mainly controlled by soil moisture (Xu and Hosen 2010), rainfall is the main factor of soil moisture, and high temperature can promote the release of CH4 (Zhang et al. 2015a; Zhang et al. 2015b). A research by Guo et al (Jiang et al. 2017) Showed that there was no significant correlation between soil temperature, WFPS, and CH4 absorption, but it increased from April to July and declined in winter.

Conclusion

The NGHG in terms of N2O plus CH4 fluxes in carbon dioxide equivalent (CO2-eq) was remarkably influenced by straw incorporation, nitrogen fertilization, and precipitation during a cropping season in rainfed monocropping system. Higher mean values of emission factors were obtained in 2013–2014 than 2012–2013 and 2014–2015 due to higher precipitation. The optimized methods can save about 32% use of N fertilizers without any significant decrease in grain yields relative to conventional farmers’ N fertilizer use; however, straw return enhanced N2O emission orientating to increase NGHG and NGHGI. Across over this study, N2O emission factor was the lowest at 0.05% derived from N fertilizer by adding 5% of DCD into optimized method under field condition. The use of slow release fertilizer could somewhat decrease in N2O emission with 0.5% of emission factor compared with optimized method by urea fertilizer with approximately 0.7% emission factor. Since the statistical differences among annual cumulative emissions were observed coinciding with that during the winter wheat growing season, it is a vital and important period for investigation of N2O emissions from winter wheat-summer fallow monocropping system under rainfed cropland. In addition, the soil in our field acted as a weak sink for atmospheric CH4 uptake, and further study should be considered to encounter the evaluation of soil carbon sequestration changes due to straw management. It, eventually, can be expected that in order to mitigate N2O emission, straw return before N fertilizer application as treatment was attractive to avoid residue decomposition processes attributed with N fertilizers.