ZHANG Chong ,WANG Dan-dan ,ZHAO Yong-jian ,XlAO Yu-lin ,CHEN Huan-xuan ,LlU He-pu ,FENG Li-yuan,YU Chang-hao,JU Xiao-tang#

1 College of Tropical Crops, Hainan University, Haikou 570228, P.R.China

2 College of Resources and Environmental Sciences, China Agricultural University, Beijing 100193, P.R.China

Abstract Ammonia (NH3) emissions should be mitigated to improve environmental quality.Croplands are one of the largest NH3 sources,they must be managed properly to reduce their emissions while achieving the target yields.Herein,we report the NH3 emissions,crop yield and changes in soil fertility in a long-term trial with various fertilization regimes,to explore whether NH3 emissions can be significantly reduced using the 4R nutrient stewardship (4Rs),and its interaction with the organic amendments (i.e.,manure and straw) in a wheat–maize rotation.Implementing the 4Rs significantly reduced NH3 emissions to 6 kg N ha–1 yr–1 and the emission factor to 1.72%,without compromising grain yield (12.37 Mg ha–1 yr–1)and soil fertility (soil organic carbon of 7.58 g kg–1) compared to the conventional chemical N management.When using the 4R plus manure,NH3 emissions (7 kg N ha–1 yr–1) and the emission factor (1.74%) were as low as 4Rs,and grain yield and soil organic carbon increased to 14.79 Mg ha–1 yr–1 and 10.09 g kg–1,respectively.Partial manure substitution not only significantly reduced NH3 emissions but also increased crop yields and improved soil fertility,compared to conventional chemical N management.Straw return exerted a minor effect on NH3 emissions.These results highlight that 4R plus manure,which couples nitrogen and carbon management can help achieve both high yields and low environmental costs.

Keywords: ammonia emission,crop yield,4R nutrient stewardship,partial manure substitution,winter wheat–summer maize cropping system

1.lntroduction

The application of nitrogen (N) fertilizer is crucial for boosting crop yields and sustaining population growth(Erismanet al.2008;Zhang Xet al.2021).However,the application of N fertilizer or manure to croplands may lead to the emissions of large amounts of ammonia(NH3),particularly from high-pH (alkaline) soils.High atmospheric NH3concentrations contribute to the formation of fine particulate matter,which exert adverse impacts on human health (Plautz 2018).Ammonia deposition in ecosystems can also lead to the loss of biodiversity and water pollution (Bowmanet al.2008;Suttonet al.2011;Hernándezet al.2016;Liuet al.2022).Croplands are one of the largest NH3sources,emitting as much as 9–16 Tg N yr–1of NH3globally because of N fertilizer application,this value is likely to double by 2050 (Maet al.2021;Zhanet al.2021;Xuet al.2022).Ammonia emission is one of the major pathways for N loss in cropping systems;it represents 10–19% of the N loss from applied fertilizer,reaching up to 64% in some cases (Panet al.2016;Zhanet al.2021).Therefore,proper management to achieve a high targeted yield and low NH3emissions is crucial in intensive cropping systems(Zhanget al.2015;Guet al.2021).

Improving N management is the key to reducing NH3emissions in croplands (Younget al.2021),and this can be summarized as the concept of the 4R nutrient stewardship (4Rs),i.e.,applying the right source of N at the right rate,right time,and right place.The 4Rs are widely used to reduce NH3losses (Huanget al.2016;Panet al.2016;Xiaet al.2017a;Liet al.2018;Tiet al.2019;Shaet al.2021) while increasing crop N acquisition and yields globally (Zhanget al.2015;Nkebiweet al.2016).For instance,the optimum N rates,deep placement of N fertilizer,split N applications,and replacement of ammonium bicarbonate with urea have reduced NH3emissions by 51,69,17,and 40%,respectively (Shaet al.2021).However,most studies to date have generally analyzed only one or two aspects of the 4Rs,and few have investigated the interactions among all the four aspects of the 4Rs (i.e.,full implementation of the 4Rs)in reducing NH3emissions in croplands.The application of organic amendments (e.g.,manure and straw) is a common practice to increase soil fertility and crop yield,especially in the long run (Zhanget al.2020).Recently,many researchers have called for the application of the 4Rs along with manure,which is called 4R Plus (4R Plus 2022),to reduce nutrient losses,and improve productivity and soil fertility (4R Plus 2022).However,few studies have examined the effects of 4R Plus on NH3emissions under field conditions.

We selected the North China Plain (NCP),one of the global hotspots for NH3emissions,as the study region,it emits 1.6–2.8 Tg N yr–1of NH3,and 45% of these emissions come from croplands (Liuet al.2018;Zeng and Li 2020).Its NH3emission factors (EFs) are twice as large as the global average,largely because of the alkalinity of its soils (pH>7.5),higher air temperature during summer seasons (Zhanet al.2021),and poor N fertilizer management,including surface broadcasting of a large amount of urea (Guet al.2020).The main cropping system in the NCP is winter wheat–summer maize (Zhanget al.2019),and the hot and wet summer season provides favorable conditions for NH3emissions.Soils in the NCP(primarily calcareous Cambisols) are characterized by low fertility because of the low content of organic matter(Duet al.2018).The use of organic manure or straw to increase soil fertility has received much attention in recent years (Zhang Cet al.2021).The soils and cropping systems in the NCP provide opportunities to study the following objectives: (1) to what extent NH3emissions can be reduced by implementing the 4Rs;and (2) the effects of manure (4R Plus) and straw application on NH3emissions and crop yield.We conducted this study under various fertilization regimes in a long-term field trial.

2.Materials and methods

2.1.Site description

A long-term field trial was established in October 2006 at the Shangzhuang Experimental Station (40°08.4′N,116°10.7′E) of China Agricultural University.The site has a humid continental monsoon climate,with a mean annual air temperature and rainfall of 13°C and 540 mm(1981–2015),respectively,and 60–70% of the rain falls between June and August.The soil is a calcareous Cambisol (WRB classification) or fluvo-aquic (according to the Chinese soil genetic classification).At the beginning of the long-term field trial in 2006,the soil had the following basic properties: a clay loam texture with 28% clay,52% silt,and 20% sand;bulk density,1.31 g cm–3;pH,8.1 (measured in water at a ratio of 1:2.5 of soil:water);organic carbon content,7.1 g kg–1;total N,0.8 g kg–1;ammonium N,1.2 mg kg–1;nitrate N,24.5 mg kg–1;Olsen-P,7.8 mg kg–1;and K,76.2 mg kg–1(Huanget al.2013).Winter wheat was sown in early October and harvested in the middle of June the following year,and summer maize was sown immediately after the winter wheat harvest and harvested in early October in the same year for a continuous winter wheat–summer maize rotation system.

2.2.Experimental design

The field trial included eight treatments with four N levels,namely,no N (N0),optimum N (Nopt),conventional N (Ncon),and balanced N (Nbal),as well as two straw management methods,namely,straw removal and return (Table 1).The N rates in the Nopttreatments were determined using the target yield,which was 160 kg N ha–1of urea for both summer maize and winter wheat.In the Ncontreatments,the urea N rate was 260 kg N ha–1for summer maize and 300 kg N ha–1for winter wheat,which are the typical N rates used by local farmers.The N rates in the Nbaltreatments were also determined based on target yields,which were 180 kg N ha–1for summer maize and 170 kg N ha–1for winter wheat.The N fertilizer used in the Nbaltreatments was cattle manure plus urea.The target yields of the Nbaltreatments were higher than those of the Nopttreatments because the combination of urea and manure N in Nbaltreatments could achieve better synchrony of crop N demand and soil N supply,and thus higher yields.Decomposed cattle manure (equivalent to 30 t ha–1of fresh weight) was applied annually before sowing winter wheat.We assumed that 40 and 20% of the manure N was available in the winter wheat and summer maize growing seasons,respectively (Qiuet al.2012),and that the gap between the available manure N and balanced N rates was filled by urea.Details on the N application rates in the eight treatments are shown in Table 2.

Table 1 Treatments in long-term field trial

Table 2 Synthetic and manure N inputs in field treatments from 2017 to 20191)

With respect to the N application method,urea was applied at a depth of 5–7 cm during four-and ten-leaf fertilization for summer maize.Urea and manure N were surface-broadcasted,and deep plowing was performed using a deep plow machine (approximately 20 cm);and topdressing urea-N was surface-broadcasted,and then flood irrigation was performed for winter wheat.In the maize season,half the urea-N was applied during four-leaf fertilization and the other half during ten-leaf fertilization.In the wheat season,half the urea-N was applied during basal fertilization and the other half during topdressing for the synthetic N treatments (Nopt,SNopt,Nconand SNcon),while one-third of the urea-N was applied during basal fertilization and two-thirds during topdressing for thebalanced N treatments (MNbaland SMNbal).Therefore,the N fertilizers were applied at the right place and the right time in all the fertilization treatments (Ju and Zhang 2017).A randomized complete block design was used with three replicates,and each plot was 64 m2(8 m×8 m)in area.Based on the above experimental design,the optimum N treatments (Noptand SNopt) represented examples of fully implementing the 4Rs,and the balanced N treatments (MNbaland SMNbal) could be examples of 4R Plus applications (4R Plus 2022).

The same amounts of synthetic phosphorus and potassium fertilizers were applied in all eight treatments:200 kg P2O5ha–1yr–1and 200 kg K2O ha–1yr–1,respectively,and half was applied during four-leaf fertilization of summer maize and the other half during the sowing of winter wheat.The irrigation timing and rates for winter wheat were determined using soil moisture measurements,and irrigation was conducted three to five times annually at rates of 28–70 mm each time.Summer maize was rain-fed and irrigated only when dry weather occurred at emergence.At the sowing of winter wheat,deep plowing(approximately 20 cm) was performed using a plow machine to incorporate the fertilizer and maize straw.There was no tillage after the winter wheat harvest,and the wheat straw was shredded and left on the soil surface.After 11 years of cultivation,the soil properties in the plots for the eight treatments were determined in October 2017 and they were found to be quite diverse (Table 3).

Table 3 Soil properties in eight field treatments1)

2.3.Measurement of ammonia flux

Ammonia flux was measured using the continuous-airflow enclosure method (Yao etal.2018).The apparatus included a pump,two NH3trappers (250 mL gas washing bottles,each containing 80 mL of 0.025 mol L–1sulfuricacid),a stainless-steel dynamic chamber (30 cm in length,15 cm in width,and 15 cm in height),a valve,and tubes.Two dynamic chambers in each plot were inserted to a depth of 10 cm in the soil,and 5 cm was left above the soil surface.When collecting NH3,a plexiglass lid with a silicone pad was placed beyond the dynamic chamber and held by four clamps;ambient air at 2.5 m above the soil surface was pumped into the chamber at a rate of 15–20 chamber headspace volumes per min (Yaoet al.2018).The NH3emitted from the soil was captured by the two NH3trappers.After collection,the sulfuric acid solution in the NH3trappers was transferred to the laboratory to determine the NH4+-N concentrationviathe indophenol blue colorimetric method at 625 nm using an ultraviolet-visible spectrophotometer (UVmini-1240,Shimadzu,Japan).Ammonia flux was measured daily for 1–10 days and then every 3–5 days for 10–21 days after fertilization.On the day of measurement,NH3was sampled from 8 a.m.to 10 a.m.

2.4.Auxiliary measurements

Soil samples were collected 2–3 days after N fertilization to measure the mineral N (NH4+and NO3) and moisture levels.Three soil cores,at a depth of 0–20 cm,were sampled from each plot using a 3 cm-diameter gauge auger,and they were combined to form a composite sample.Soil samples were passed through a 2-mm sieve,and a 24 g fresh sub-sample was then extracted using 1 mol L–1KCl with a 5:1 mass ratio of KCl solution to soil.The NH4+and NO3–concentrations in the extracts were analyzed using an AA3 continuous flow analyzer(BranCLuebbe GmbH,Norderstedt,Germany).A 20 g fresh sub-sample was oven-dried at 105°C for 24 h to determine the soil gravimetric water content and calculate the water-filled pore space (WFPS) using the methods proposed by Gaoet al.(2014).

At maturity,6 m2of winter wheat and 14.4 m2of summer maize were harvested and separated into grains and straw.They were weighed separately to calculate the biomass,and the straw was manually shredded.Subsamples of grain and straw were weighed immediately,and then again after drying at 70°C to determine the water content.They were then ground sufficiently to pass through a 0.15-mm sieve,and the N content was measured using a CN analyzer (Vario Max CN,Elementar,Hanau,Germany).

The air temperature was measured using a sensor(HMP155A-L,Campbell Scientific,USA) and rainfall was measured using a tipping bucket rain gauge (TE525MM-L,Campbell Scientific).The data were recorded using a data logger (CR1000;Campbell Scientific).

2.5.Calculations and statistical analysis

The NH3flux on the day of measurement was calculated using eq.(1).The NH3flux on non-sampling days was estimated using linear interpolation between two adjacent measurements.The cumulative NH3flux was calculated as the sum of the daily NH3fluxes during the entire observation period.

where C is the NH4+-N concentration of the field plot samples (mg N L–1),V is the volume of the sulfuric acid solution (L),S is the coverage area of the dynamic chamber (m2),T is the sampling duration (h),and 10–2is the unit conversion factor.

The NH3emission factor (EF) was calculated using eq.(2):

where NH3-Nfertilizationand NH3-Nnon-fertilizationare the cumulative NH3emissions in the fertilization and non-fertilization treatments,respectively (kg N ha–1),and N rate is the synthetic and manure N rate (kg N ha–1).As there were no significant differences in the cumulative NH3emissions between the N0and SN0treatments (refer to Section 3.2),they were averaged as the soil background NH3emissions in the calculation of the NH3EFs.

The N surplus was calculated using eq.(3):

where Nfer,Nman,Nstr,Ndep,and NBNFare the N inputs from synthetic fertilizer,manure,straw,deposition,and biological N fixation,respectively (kg N ha–1);and Nharis the aboveground N uptake (kg N ha–1).N deposition data were obtained from Yinet al.(2017),who reported that N deposition was 35 kg N ha–1for summer maize (from June to September) and 23 kg N ha–1for winter wheat (from October to May) in the NCP.Estimates of biological N fixation for maize and wheat (5 kg N ha–1) were obtained from Bouwmanet al.(2013).

All statistical analyses were performed using SPSS 20.0 (IBM Corp.,Armonk,NY,USA).The differences in NH3emissions,crop yield,N surplus,and other parameters were determined using Duncan’s test with a significance level ofP<0.05.Graphs were plotted using SigmaPlot 14.0.

3.Results

3.1.Ammonia flux

There was a large variation in the NH3fluxes for summer maize during different years and fertilization events(Fig.1-A).The flux was relatively high during four-leaf fertilization in 2017 and 2019,and the conventional N treatments (Nconand SNcon) produced the highest NH3fluxes of 0.05–12.76 kg N ha–1d–1(average 2.19 kg N ha–1d–1),whereas the other four fertilization treatments produced fluxes of 0.05–4.65 kg N ha–1d–1(average 0.88 kg N ha–1d–1).Ammonia fluxes in the other four fertilization events (four-leaf fertilization in 2018 and ten-leaf fertilization in 2017,2018,and 2019) were relatively low,with the conventional N treatments and the other four fertilization treatments producing fluxes of 0.04–2.03 kg N ha–1d–1(average 0.32 kg N ha–1d–1)and 0.05–0.58 kg N ha–1d–1(average 0.21 kg N ha–1d–1),respectively (Fig.1-A).The low NH3fluxes in these four fertilization events were associated with the occurrence of rainfall within 3 days after fertilization (Fig.1-B).The air temperature was 21–31°C (average 27°C),and the soil WFPS was 30–69% (average 51%) during the maize season (Fig.1-B).This implied that NH3emissions were primarily determined by N management and rainfall within 1 week after fertilization for summer maize.

Fig.1 Ammonia flux (A),rainfall,soil water-filled pore space and air temperature (WFPS;B) in summer maize during 2017–2019.N0,Nopt,Ncon,and Nbal denote no N,optimum N rate,conventional N rate,and balanced N rate,respectively;S and M denote straw and manure,respectively.Bars denote standard errors (n=3).

The NH4+-N concentrations of soil for the fertilization treatments was 0–28.83 mg N kg–1(average 4.68 mg N kg–1),except for the conventional N treatments which had higher soil NH4+-N concentrations of 0–41.21 mg N kg–1(average 7.82 mg N kg–1) (Appendix A).The soil NH4+-N concentration generally peaked within 7 days of fertilization and then decreased,indicating rapid nitrification in the studied soil (Appendix A).This corresponds well with the high NH3flux,which occurred within 1 week of fertilization (Fig.1-A).Soil NO3–-N concentrations in the fertilization treatments were in the range of 1.92–53.53 mg N kg–1(average 20.82 mg N kg–1),except for the conventional N treatments,which had higher soil NO3–-N concentrations of 0.04–81.90 mg N kg–1(average 26.82 mg N kg–1) (Appendix A).

The NH3flux of winter wheat was much lower than that of summer maize,with values between 0.02 and 0.35 (average 0.10) kg N ha–1d–1in the six fertilization treatments (Fig.2-A).The flux values fluctuated with air temperature,and the magnitude was similar in different fertilization events (Fig.2-A and B).The air temperature during the winter wheat season was 6–19°C (average 12°C),which was significantly lower than that in the summer maize season (Fig.2-B),and it was the key factor responsible for the lower NH3emissions from winter wheat.

Fig.2 Ammonia flux (A),rainfall,soil water-filled pore space and air temperature (WFPS;B) in winter wheat during 2017–2019.N0,Nopt,Ncon and Nbal denote no N,optimum N rate,conventional N rate and balanced N rate,respectively;S and M denote straw and manure,respectively.Vertical bars denote standard errors (n=3).

The NH4+-N concentration in soil for winter wheat was much lower than that for summer maize,with values between 0 and 8.72 mg N kg–1(average 1.63 mg N kg–1) (Appendix B) in the fertilization treatments;peak soil NH4+-N concentrations occurred within 7 days of fertilization (Appendix B).Soil NO3–-N concentrations in the fertilization treatments were in the range of 1.00–40.09 mg N kg–1(average 16.16 mg N kg–1),except for the conventional N treatments,which had higher soil NO3–-N concentrations of 5.73–74.93 mg N kg–1(average 32.28 mg N kg–1) (Appendix B).

3.2.Ammonia emission

Cumulative NH3emissions from summer maize varied significantly every year for the fertilization treatments,with averages of 13.57,8.05,and 30.89 kg N ha–1in 2017,2018,and 2019,respectively (Fig.3-A).The conventional N treatments produced the highest NH3losses of 25.30–27.59 kg N ha–1,whereas the optimum N treatments and manure plus synthetic N treatments (Nopt,SNopt,MNbal,and SMNbal) significantly reduced the losses to 12.29–14.08 kg N ha–1in the three summer maize seasons (Fig.3-A).Cumulative NH3losses in the winter were much lower than those in the summer (Fig.3-B).The values of the two-season average cumulative NH3losses ranged from 3.13 to 3.40 kg N ha–1in the six fertilization treatments(Fig.3-B).

Fig.3 Cumulative ammonia emissions from summer maize (A),winter wheat (B),and summer maize–winter wheat rotation (C)during 2017–2019.M,W,and M–W denote maize,wheat,and summer maize–winter wheat rotation (n=3),respectively;N0,Nopt,Ncon,and Nbal denote no N,optimum N rate,conventional N rate,and balanced N rate,respectively;S and M denote straw and manure,respectively.The means are the average NH3 losses during the three seasons of maize cultivation and are shown in Fig.3-A (n=9),and two seasons of wheat cultivation are shown in Fig.3-B (n=6),whereas the means in Fig.3-C are the sums of the means in Fig.3-A and B (n=3).Vertical bars denote standard errors,and lowercase letters denote the mean values among treatments,where different letters indicate significant (P<0.05) mean differences.

The annual cumulative NH3emissions also varied every year and were primarily determined by the NH3emissions from summer maize,which accounted for 79–89% (average 83%) of the annual emissions.The emissions were significantly higher in the two conventional N treatments (28.46–30.99 kg N ha–1) than in the other four fertilization treatments (15.47–17.21 kg N ha–1) (Fig.3-C).The emissions in the straw return treatments were comparable to those in the straw removal treatments,indicating that straw return had a minor effect on NH3emissions under the studied conditions (Fig.3).

The NH3EFs for summer maize were,on average,3.59,1.27,and 9.76% in 2017,2018,and 2019,respectively(Fig.4-A).The conventional N treatments exhibited the highest NH3EFs of 6.91–7.79% in a three-year average.The optimum N rate and manure plus synthetic treatments significantly reduced NH3EFs to 3.10–4.70%,and no significant differences were observed among the four treatments (Fig.4-A).The NH3EFs in winter were much lower than those in the summer (Fig.4-B) and showed the same magnitude between the two seasons.The two-year average of NH3EFs were 0.15–0.29% for the six fertilizer treatments (Fig.4-B).The annual average NH3EFs of the two conventional N treatments (3.29–3.74%) was significantly higher than those of the other four fertilization treatments (1.66–1.81%) (Fig.4-C).

Fig.4 Ammonia emission factors for summer maize (A),winter wheat (B),and summer maize–winter wheat rotation (C) during 2017–2019.M,W,and W–M denote maize,wheat,and winter wheat–summer maize rotation (n=3),respectively;N0,Nopt,Ncon,and Nbal denote no N,optimum N rate,conventional N rate,and balanced N rate,respectively;S and M denote straw and manure,respectively.The means are the average NH3 emission factors for the three seasons of maize cultivation,as shown in Fig.4-A (n=9),and the two seasons of wheat cultivation are shown in Fig.4-B (n=6),whereas the means in Fig.4-C are the sums of the means in Fig.4-A and B (n=3).Vertical bars denote standard errors,and lowercase letters denote the mean values among treatments,where different letters indicate significant (P<0.05) mean differences.

3.3.Crop yield

The manure plus synthetic N treatments produced the highest maize grain yields (8.21–8.46 Mg ha–1),followed by the four synthetic N treatments (6.36–7.65 Mg ha–1),whereas the no N treatments produced the lowest (5.40–5.59 Mg ha–1) (Fig.5-A).The two-season average wheat grain yields were the highest in the manure plus synthetic N treatments (6.14–6.76 Mg ha–1),followed by the synthetic N treatments (5.28–6.14 Mg ha–1),and thereafter the no N treatments (1.36–1.52 Mg ha–1) (Fig.5-B).Annual grain yields were 14.35–15.22,11.44–13.69,and 6.76–7.11 Mg ha–1in the manure plus synthetic N,synthetic N,and no N treatments,respectively (Fig.5-C).Straw return increased the grain yields in the long-term fertilization treatments,although no significant differences were observed (Fig.5).

Fig.5 Grain yields of summer maize (A),winter wheat (B),and summer maize–winter wheat rotation (C) during 2017–2019.M,W,and M–W denote maize,wheat,and summer maize–winter wheat rotation (n=3),respectively;N0,Nopt,Ncon,and Nbal denote no N,optimum N rate,conventional N rate,and balanced N rate,respectively;S and M denote straw and manure,respectively.The means are the average yields of the three seasons of maize cultivation,as shown in Fig.5-A (n=9),and the two seasons of wheat cultivation are shown in Fig.5-B (n=6),whereas the means in Fig.5-C are the sums of the means in Fig.5-A and B (n=3).Vertical bars denote standard errors,and lowercase letters denote the mean values among treatments,where different letters indicate significant (P<0.05) mean differences.

3.4.N surplus

The annual N surplus values were calculated in the rotation cycles of 2017–2018 and 2018–2019.The conventional N treatments had the highest N surpluses of 276–387 kg N ha–1yr–1,whereas the optimum N treatments and manure plus synthetic N treatments had significantly lower N surpluses of 35–154 kg N ha–1yr–1.The non-fertilization treatments had negative N surpluses of between–64 to–19 kg N ha–1yr–1.Straw return significantly increased the N surplus,with the exception of the non-fertilization treatments in 2017–2018 (Table 4).Both NH3emissions and EFs increased exponentially with an increase in the N surplus (R2>0.40,P<0.01;Fig.6).

Fig.6 Relationship between N surplus and NH3 emissions (A) and NH3 emission factors (B).＊＊ denotes P<0.01.Open and solid circles denote data for the rotation cycles of 2017–2018 and 2018–2019,respectively.

Table 4 N surplus (kg N ha–1 yr–1) under different nutrient management conditions in summer maize–winter wheat rotation

4.Discussion

4.1.Ammonia emissions in maize and wheat seasons

Maize and wheat planting are the two largest sources of NH3emissions globally,primarily because of the large planting areas of these two crops (Zhanet al.2021;Liuet al.2022),and the wheat–maize double cropping system is one of the most intensively managed high-yield systems in China.This study found that NH3fluxes in the maize season were substantially higher than those in the wheat season,accounting for an average of 83% of the annual NH3emissions (Figs.1–3),which suggests that controlling the NH3emissions in the hot-wet maize season is the key to reducing annual NH3emissions (Zhanget al.2022).The low NH3emissions in the wheat season(NH3EFs of 0.15–0.29%) could be attributed to two main factors: 1) Low temperatures slow down urea hydrolysis,the transformation of NH4+-N to NH3,and NH3diffusion from soil to air,thus limiting NH3emission,even under high N application rates (Fig.2) (Pedersenet al.2021).2)Proper application techniques (even under conventional management by farmers),such as the deep placement of N fertilizer in basal fertilization and immediate irrigation after topdressing,which move the urea to deep soils,prevent the upward diffusion of NH3as deep soil particles retain more NH4+-N (Rochetteet al.2013).Although N fertilizer was deep-placed in the maize season,the high temperature increased NH3concentrations in the soil matrix and NH3diffusion from soil to air,resulting in higher NH3emissions than those from wheat.

Ammonia flux was measured using the dynamic chamber method in this study,this method has been widely used to determine field NH3emissions,and its results are comparable with those of the micrometeorological method (Maet al.2021).Therefore,we compared our results with studies that used the dynamic chamber or micrometeorological method to measure NH3emissions.It was reported that global NH3EFs were 11.13–12.70% and China exhibits higher NH3EFs for maize (14.33%) (Maet al.2021;Zhanet al.2021).Although the soil–climate condition was favorable for NH3emissions during the maize season in the studied region (Fig.1),the NH3EFs were only 3.10–7.79%,much lower than those in previous studies (Fig.4) (Ju and Zhang 2017;Maet al.2021;Zhanet al.2021).These differences could be attributed to the deep placement of N fertilizer,which significantly inhibited NH3emissions(Rochetteet al.2013;Zhanget al.2022).In contrast,surface broadcasting or surface banding of urea at high N fertilizer application rates,which is a common N application method in China (Guet al.2020),has been used in other studies (Juet al.2009;Gaoet al.2014).

4.2.Significant reduction in NH3 emissions using 4Rs

In this study,the cumulative NH3emissions and NH3EFs were significantly reduced upon fully implementing the 4Rs (i.e.,optimum synthetic N and manure plus synthetic treatments) compared to those in the conventional N treatments (Figs.3 and 4).The 4Rs achieved NH3EFs of 3.10–4.70,0.17–0.29,and 1.66–1.81% for maize,wheat,and wheat–maize rotation,respectively,which are substantially lower than the NH3EF of 10.0% for N fertilizers in Tier 1 of the IPCC Guidelines (De Kleinet al.2006).According to recent studies,the estimated global NH3EFs is 11.13–12.70% (corresponding N rates of 120–122 kg N ha–1) for maize and 12.05–13.30%(corresponding N rates of 94–101 kg N ha–1) for wheat(Ladhaet al.2016;Maet al.2021;Zhanet al.2021;IFA 2022),indicating the great NH3mitigation potential of implementing the 4Rs in croplands.Previous studies have generally investigated only one or two aspects of the 4Rs,rather than fully implementing all of them.For instance,Liet al.(2017) evaluated the effectiveness of the urease inhibitor Limus?,under surface broadcast conditions during the maize season;and although Limus?significantly reduced NH3emissions compared to the non-Limus?treatment,the emissions were still high (29 kg N ha–1and accounted for 19% of the applied N fertilizer).Similar results were obtained by Wanget al.(2022) in wheat–maize rotation.

Previous studies have shown that manure substitution significantly reduces NH3emissions because manure enhances the microbial immobilization of inorganic N(Xiaet al.2017b) and increases crop N uptake from the soil organic N poolviabetter synchronization of the crop N demand and soil N supply (Zhang Cet al.2021).Our results indicate that a proper combination of manure and synthetic N can serve as a promising measure for reducing NH3emissions in croplands.Straw return had a minor effect on NH3emissions (Figs.4 and 5),which is in line with the reports of Xuet al.(2022) and Shaet al.(2021).Although straw return supplies additional N,the comparable NH3emissions between the straw return and removal treatments may be attributed to two factors: 1)The high straw C:N ratios (53–94 for wheat and 35–55 for maize) can increase soil microbial N immobilization,thus stimulating inorganic N incorporation into the organic pool(Chenget al.2017).2) Straw return increases the grain yield and crop N uptake (Fig.5;Table 4),thus reducing the substrate for NH3emissions.

4.3.The combination of manure and synthetic N achieves both low NH3 emissions and high crop yield

The best N management strategy should achieve high crop yields,low N losses,and maintain or improve soil fertility (Ju and Zhang 2021).We compared the crop yields,NH3losses,soil fertility (soil organic carbon and total nitrogen),and N surplus under different N and C treatments.The 3Rs (right source,right time,and right place but excessive rate,i.e.,the conventional synthetic N fertilizer treatments) were associated with the highest NH3losses (20 kg N ha–1) and N surplus (335 kg N ha–1),despite being associated with a desirable crop yield of 12.80 Mg ha–1,highlighting the importance of including the right rate in reducing NH3emissions.Above yield is similar to the results of Wuet al.(2014),who reported regional attainable grain yields of wheat–maize rotation in the NCP.The 3Rs also maintained soil fertility,in terms of soil organic carbon and total nitrogen (Hanet al.2017;Huanget al.2017) (Fig.7),suggesting that this strategy is associated with grain production at high environmental costs.

Fig.7 Ammonia losses and crop yields under different N and C managements.The 3Rs denote the right source,right time,and right place but excessive rate (Ncon and SNcon treatments);the 4Rs denote the right source,right time,right place,and right rate(Nopt and SNopt treatments);4Rs+manure (4R Plus) denotes partial manure substitution under the condition of the 4Rs (MNbal and SMNbal treatments);U denotes urea and M denotes cattle manure,EF denotes emission factor (%).

After fully implementing the 4Rs,the N surplus was only 108 kg N ha–1,which is lower than the N surplus benchmark of 160 kg N ha–1for wheat–maize rotation in the NCP proposed by Zhanget al.(2019),resulting in minimal NH3losses (6 kg N ha–1and an NH3EF of 1.72%) without compromising crop yields (12.37 Mg ha–1)and soil fertility (compared with that in conventional N management) (Fig.7).After using 4R plus manure,the NH3losses (7 kg N ha–1and an NH3EF of 1.74%) and N surplus (98 kg N ha–1) were still at very low levels,and as an extra benefit,the grain yield increased by 16–20%,compared with that obtained in the case of the 3Rs and 4Rs (Fig.7),suggesting that the combination of manure and synthetic N can be used to achieve low NH3emissions with even higher crop yields.

5.Conclusion

Conventional chemical N management achieved desirable crop yields but induced high NH3losses.In contrast,the application of N fertilizer by fully implementing the 4Rs effectively reduced the NH3emissions,produced desirable crop yields,and maintained soil fertility.4R plus manure was associated not only with low NH3emissions,similar to the 4Rs,but also with much higher crop yields and soil fertility than the 4Rs.Straw return exerted minor effects on NH3emissions but increased crop yields and improved soil fertility.These findings have important implications for achieving the sustainable goals of food security and environmental protection.

Acknowledgements

This work was supported by the Hainan Key Research and Development Project,China (ZDYF2021XDNY184),the Hainan Provincial Natural Science Foundation of China (422RC597),the National Natural Science Foundation of China (41830751),the Hainan Major Science and Technology Program,China (ZDKJ2021008),and the Hainan University Startup Fund,China(KYQD(ZR)-20098).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on https://doi.org/10.1016/j.jia.2022.12.008