Lin Li,Hua Tian,Mingua Zang,Pingsan Fan,Umair Asraf,Haiong Liu,Xiongfei Cen,Meiyang Duan,Xiangru Tang,Zaiman Wang,Zeng Zang,Senggang Pan,*

a State Key Laboratory for Conservation and Utilization of Subtropical Agro-bioresources,College of Agriculture,South China Agricultural University,Guangzhou 510642,Guangdong,China

b Scientific Observing and Experimental Station of Crop Cultivation in South China,Ministry of Agriculture,Guangzhou 510642,Guangdong,China

c Guangzhou Key Laboratory for Science and Technology of Fragrant Rice,Guangzhou 510642,Guangdong,China

d Key Laboratory of Key Technology for South Agricultural Machine and Equipment,Ministry of Education,Guangzhou 510642,Guangdong,China

e College of Tropical Crops,Hainan University,Haikou 570228,Hainan,China

f Department of Botany,Division of Science and Technology,University of Education,Lahore 54770,Punjab,Pakistan

g Hezhou Academy of Agricultural Sciences,Hezhou 542800,Guangxi,China

h School of Engineering,Jiangxi Agricultural University,Nanchang 330045,Jiangxi,China

Keywords:

A B S T R A C T Deep placement of nitrogen fertilizer is a key strategy for improving nitrogen use efficiency.A two-year field experiment was conducted during the early rice growing seasons(March–July)of 2016 and 2017.The experimental treatments comprised two rice cultivars:Wufengyou 615(WFY 615)and Yuxiangyouzhan(YXYZ),and three N treatments:mechanical deep placement of all fertilizers as basal dose at 10 cm soil depth(one-time deep-placement fertilization,namely OTDP fertilization);manual surface broadcast(the common farmer practice)of 40%N fertilizer at one day before sowing(basal fertilizer)followed by broadcast application of 30% each at tillering and panicle initiation stages;and no fertilizer application at any growth stage as a control.One-time deep-placement fertilization increased grain yield of both rice cultivars by 11.8%–19.6%,total nitrogen accumulation by 10.3%–13.1%,nitrogen grain production efficiency by 29.7%–31.5%,nitrogen harvest index by 27.8%–30.0%,nitrogen agronomic efficiency by 71.3%–77.2%,and nitrogen recovery efficiency by 42.4%–56.7% for both rice cultivars,compared with the multiple-broadcast treatment.One-time deep-placement fertilization reduced CH4-induced global warming potential(GWP)by 20.7%–25.3%,N2O-induced GWP by 7.2%–12.3%,and total GWP by 14.7%–22.9%for both rice cultivars relative to the multiple-broadcast treatment.The activities of glutamine synthetase and nitrate reductase were increased at both panicle-initiation and heading stages in both rice cultivars following one-time deep-placement fertilization treatment.Larger leaf area index at heading stage and more favorable root morphological traits expressed as larger total root length,mean root diameter,and total root volume per hill were also observed.One-time deep-placement fertilization could be an effective strategy for increasing grain yield and nitrogen use efficiency and lowering greenhouse-gas emissions under mechanical direct-seeded cropping systems.

1.Introduction

Rice is a staple food for almost two thirds of Chinese people[1,2].Conventional transplanting is the main method of rice production and includes three steps:nursery cultivation,seedling pulling,and transplanting.The low resource-use efficiency and high labor requirement of this method is no longer suitable for sustainable development of Chinese agricultural systems[3,4].

Manual broadcasting,a common method for direct-seeding rice(DSR),often leads to uneven stand establishment[5].Chinese scientists have invented a technique of mechanized sowing of rice seeds that enforces uniform row spacing of seeds[6].Mechanized direct seeding of rice has become an alternative,cost-effective cropping technique that favors earlier crop establishment,exploiting the increased solar radiation and increasing grain yield[7].

Nitrogen(N)is an essential nutrient element for rice growth.Manual broadcast of N fertilizer is a conventional but inefficient method that leads to high fertilizer application and nutrient loss.In addition to methods invented to improve nitrogen use efficiency(NUE)[8,9],deep placement substantially improve NUE[10,11].Machinery for direct sowing of rice seeds while simultaneously placing N fertilizer in the soil between planting rows is now available[6].

There are few reports describing the effect of deep N fertilizer placement in direct-seeded rice.Liu et al.[12]observed that grain yield of dry direct-seeded rice was identical to grain yield of transplanted and flooded rice.Wang et al.[13]found that grain yield of direct-seeded rice was lower than that of mechanically transplanted rice.Gaihre et al.[14]reported that deep placement of urea increased grain yield of direct-seeded rice by 16%–19% compared to broadcast urea.However,these results are contradictory.

Methane(CH4)and nitrous oxide(N2O)are greenhouse gases(GHGs)that contribute over one-fifth radiative forcing globally[15].Rice paddy fields are large sources of atmospheric CH4and N2O[16].Rice production accounts for about 50% of global cropland GHG emissions[17].N fertilization mitigated CH4emissions from paddy fields[18].But Sun et al.[19]reported that CH4emissions from paddy fields increased with N fertilization.Kim et al.[9]reported a relatively flat quadratic relationship of CH4emissions to N application rate.N fertilization increases methane and nitrous oxide emissions in China[20].Linquist et al.[21]observed an increase of N2O emissions and a decrease of CH4emissions following deep N fertilizer placement in flooded paddy fields.However,it is unclear how efficient nitrogen management practices such as deep N fertilizer placement affect DSR systems.The present study aimed to evaluate the effect of deep placement of N fertilizer in a DSR system on grain yield,nitrogen use efficiency,and greenhouse gas emissions,compared with manual broadcast fertilization.

2.Materials and methods

2.1.Experimental treatments and design

Experiments were performed in the early seasons of 2016 and 2017,on the farm of South China Agricultural University.The regional climatic conditions are subtropical with yearly mean temperatures between 21 and 29 °C.The field soil contained 21.33 g kg-1organic C,1.06 g kg-1total N,1.17 g kg-1total P,21.02 g kg-1total K,75.42 mg kg-1available P,and 102.50 mg kg-1available K.The experimental treatments comprised two rice cultivars and three N treatments.Rice seeds were sown with a mechanical hill-direct-seeding rice machine for all treatments.Three N treatments:mechanical deep placement of all fertilizers as basal dose at 10 cm soil depth(one-time deep-placement fertilization,namely OTDP fertilization);manual surface broadcast(the common farmer practice)of 40% N fertilizer at one day before sowing(basal fertilizer)followed by 30%each at tillering and panicle initiation stages;and no fertilizer application at any growth stage as a control.Field experiments followed a split-plot design with three replications,with the main plot rice cultivar and the subplot N treatment.The area of each plot was 90 m2(15.0×6.0 m).The total N rate of 150 kg N ha-1was applied as commercial compound fertilizer(N:P2O5:K2O=15:15:15)manufactured by YaraMila Fertilizer Company(Guangzhou,China).After rice harvest,all rice straw was removed from the field.

Pre-germinated seeds of two regionally popular rice cultivars,Wufengyou 615(WFY 615)and Yuxiangyouzhan(YXYZ),were sown with the DSR machine at 25×15 cm spacing.About 3–5 seeds were sown in every hill.Paddy fields were puddled and then drained two days before sowing.Rice seeds were broadcast on the soil surface on March 25 and 27 of 2016 and 2017,respectively.All other farming and water management practices followed recommendations of the province agricultural department.No water was applied until the three-leaf stage,when each plot was reflooded to about 3–5 cm depth until plant maturity.

2.2.Yield and its components

Measurement of grain yield and yield components followed Pan et al.[22].At maturity,15 hills from each subplot were harvested for calculation of mean effective panicle number per hill.Eight representative hills were separately sampled to determine yield components.Panicles were hand-threshed and filled spikelets were separated from other spikelets(half-filled and unfilled)by submergence in tap water.Half-filled spikelets were then separated from unfilled spikelets with a blower.Three subsamples,30 g of ripe kernels and 5 g of unripe kernels were taken to count spikelets.The ripe kernels were then oven-dried at 70°C to constant weight for determining grain weight.Percentage of ripe kernels(100×number of ripe kernels/total number of kernels)was calculated.Grain yield was determined by harvesting an area of 5 m2in the center of each subplot(excluding edge plants)and adjusted to a standard moisture content of 14%.

2.3.N use efficiency

Total nitrogen accumulation(TNA),nitrogen agronomic use efficiency(NAE),nitrogen recovery efficiency(NRE),nitrogen harvest index(NHI),and nitrogen grain production efficiency(NGPE)were calculated following Wang et al.[4].At physiological maturity,eight representative hills of plants were separately sampled and separated into leaf blades,stems and sheaths,and grain.The samples were oven-dried at 70 °C to constant weight,ground,and stored for analysis of total N concentration.Dry samples(0.2 g)were taken for digestion using the Kjeldahl method.The digested solution was diluted for analysis of ammonia concentration with an Alliance-Futura NP analyser(Alliance Instruments,France),and the N content was calculated.

TNA(kg ha-1)was calculated as total plant weight(kg ha-1)at harvest multiplied by N concentration,NHI was calculated as the ratio of total nitrogen uptake in grain to total nitrogen uptake in aboveground plants in each fertilized plot,NGPE was calculated as the ratio of grain yield to total nitrogen uptake in aboveground plants in each fertilized plot,NRE was calculated as the ratio of the increase in plant nitrogen accumulation at harvest to the N application rate;and NAE was calculated as the increase in grain yield per unit of N applied.

2.4.Determination of glutamine synthetase(GS)and nitrate reductase(NR)activities

The uppermost fully expanded leaves at mid-tillering(MT)and panicle initiation(PI),and the flag leaves at heading(HS),and maturity stage(MS)were randomly sampled from each plot.GS activity was determined following Liu et al.[23].Fresh leaves(1 g)were homogenized in 3.0 mL of Tris-HCl(0.05 mol L-1,pH 8.0)in a precooled mortar in an ice bath,followed by extraction with MgSO4(2 mmol L-1),DTT(2 mmol L-1)and sucrose(0.4 mol L-1).The homogenate was centrifuged at 15,000 r min-1at 4 °C for 20 min.An aliquot of the crude GS enzyme solution(0.7 mL)and 0.7 mL of 40 mmol L-1ATP were added to 1.6 mLof reaction mixture B containing imidazole-HCl buffer(0.1 mol L-1,pH 7.4),sodium hydrogen glutamate(20 mmol L-1),MgSO4(80 mmol L-1),cysteine(20 mmol L-1),EDTA(2 mmol L-1),and hydroxylamine hydrochloride(80 mol L-1).For the control,reaction mixture A without hydroxylamine hydrochloride was used instead of reaction mixture B.The mixture was incubated at 25 °C for 15 min.Approximately 1 mL of chromogenic agent(TCA(0.2 mol L-1),FeCl3(0.37 mol L-1),HCl(0.6 mol L-1)was added to terminate the reaction.After centrifugation at 5000 r min-1for 10 min,the absorbance of the supernatant was measured at 540 nm.

Fig.1.Effects of mechanical deep placement of nitrogen fertilizer on GS activity in flag leaves in direct-seeded rice during 2016 and 2017.MT,mid-tillering stage;PI,panicle initiation stage;HS,heading stage;MS,maturity stage.

NR activity was measured following Liu et al.[23].Fresh leaves(0.5 g)were homogenized on ice in 4.0 mL of extraction solution containing phosphate buffer(25 mmol L-1,pH 7.5,a mixture of K2HPO4and KH2PO4),cysteine(5 mmol L-1)and EDTA-Na2(5 mmol L-1).The homogenate was centrifuged at 10,000 r min-1at 4 °C for 15 min,and the supernatant was collected.An aliquot(0.4 mL)of NR enzyme extract was added to 1.6 mL of reaction reagent(1.4 mL of 0.1 mol L-1KNO3-1phosphate buffer and 0.2 mL of 2.0 mg mL-1NADH),and the mixture was incubated at 25°C for 30 min.For the control,0.2 mL phosphate buffer was used instead of 0.2 mL NADH.The reaction was stopped by addition of 1.0 mL of 1% 4-aminobenzene sulfonic acid and 1.0 mL of 0.2%1-naphthylamine and incubated at room temperature for 15 min;the samples were centrifuged at 4000 r min-1for 10 min and the absorbance of the supernatant was measured at 540 nm.

2.5.Measurement of leaf area index(LAI)

LAI was measured following He et al.[1].Briefly,eight hills were randomly sampled from each plot at MT,PI,HS,and MS.All plants were washed to remove soil.Aboveground plant parts were separated into leaves,leaf sheaths,and panicles.Total green leaf blades were collected for measurement of leaf area with an area meter(Li-Cor model 3100,Li-Cor Inc.,Lincoln,USA)and leaf area per m2(LAI)was calculated.

2.6.Root sampling and measurement at heading stage

Root sampling and measurements followed Pan et al.[24].Eight rice seedlings in each plot were taken out with soil one week after transplanting.Eight PVC cylinders with diameter of 20 cm(30 cm high)were inserted into the soil to a depth of 25 cm,and the slurry in the cylinders was removed.A mesh bag with 200 screen mesh(20 cm diameter)was inserted into each cylinder and the slurry was returned to the mesh bag.Finally,one rice seedling from each hill was planted in the bag.The mesh bag ensured the free passage of water and nutrients and prevented the root system from penetrating it.The buried cylinders were removed from the soil.At heading stage,the eight mesh bags containing rice seedlings in each plot were removed and all roots were rinsed with clean tap water.Three cleaned whole roots of rice were taken to the lab for measurement of morphological characters(total root length,mean root diameter,root superficial area,and total root volume)using a root analysis instrument named WinRhizo-LA1600(Regent Instruments Inc.,Quebec,Canada)and the remainder were stored at-80 °C for enzyme activity determination.

2.7.Measurements of CH4 and N2O flux

The static chamber method was used to measure CH4and N2O fluxes as described by Tao et al.[3].CH4and N2O flux were determined by the closed-chamber technique with two different-sized chambers(0.60 m length×0.50 m width×0.80 m height and0.60 m length×0.50 m width×1.30 m height).All of the chambers were made from polymethyl methacrylate with the same lengths and widths.The base of each chamber was permanently installed in the field during the rice growing season.Electric fans were installed on the tops of the chambers to ensure complete gas mixing before each gas sample was taken.The gases in the chamber were drawn off with a syringe and immediately transferred into a 25-mL glass vacuum container at 0,10,20,and 30 min after chamber closure.Samples were collected in the morning(9:00–11:00 AM)every Tuesday from April 20 to July 6 in both years.CH4and N2O concentrations were determined in a gas chromatograph(Shimadzu GC-14B;Kyoto,Japan)equipped with an electron capture detector for N2O and a flame ionization detector for CH4estimation.N2(flow rate of 330 mL min-1),H2(flow rate of 30 mL min-1),and air(flow rate,400 mL min-1)were used as the carrier,fuel,and supporting gas,respectively.The temperatures of the column,injector,and detector were set at 55,100,and 200 °C,respectively.Fluxes were determined from the slope of the mixing ratio change in four samples.

Fig.2.Effects of mechanical deep placement of nitrogen fertilizer on NR activity in direct-seeded rice in 2016 and 2017.MT,mid-tillering stage;PI,panicle initiation stage;HS,heading stage;MS,maturity stage.

Fig.3.Effects of mechanical deep placement of nitrogen fertilizer on leaf area index in direct-seeded rice in 2016 and 2017.MT,mid-tillering stage;PI,panicle initiation stage;HS,heading stage;MS,maturity stage.

Fig.5.Seasonal variation in CH4 emission fluxes from rice paddy soil in 2016 and 2017.

Fig.6.Seasonal variation in N2O emission fluxes from rice paddy soil in 2016 and 2017.

2.8.Total global warming potential(GWP)of CH4 and N2O

GWP was determined following Bayer et al.[25].Briefly,the radiative forcing potential of CH4and N2O are 28 and 265,respectively[15].

2.9.Data analysis

A two-way ANOVA(with rice cultivar and nitrogen fertilization method as factors)and an independent-samplest-testwithin each nitrogen fertilization method treatment were performed in both experiments.All analyses were performed with the statistical software(SAS Institute 2003).Differences between treatments were considered significant atP＜0.05.

3.Results

3.1.Grain yield and yield components

Grain yield and yield components differed with N management in WFY 615 and YXYZ during 2016 and 2017.OTDP fertilization produced the highest grain yield among all treatments for WFY 615 and YXYZ,and differed in grain yield from the multiple surface-broadcasting treatment.Mean grain yields of both rice cultivars under OTDP fertilization were 7.93–8.29 t ha-1,values 11.81%-19.63%higher than under the multiple surface broadcasting treatment.OTDP fertilization produced the highest number of productive panicles ha-1and spikelets per panicle,which was 276.45×104–277.26×104and 161.64–171.95,respectively.However,the control treatment resulted in the maximum grain-filling percentage in both years,and the lowest productive panicles and spikelets per panicle led to the lowest grain yield(Table 1).WFY 615 produced more grain yield than YXYZ in 2016,but there was no difference(P＞0.05)in 2017.On average,OTDP fertilization significantly increased grain yield by 15.50% relative to multiple surface broadcasting treatment for both rice cultivars in both years.There were significant interactions between cultivar and N treatment in productive tillers ha-1,spikelets per panicle,1000-grain weight,and grain yield.

Table 1Effects of mechanical deep placement of nitrogen fertilizer on grain yield and its components in hill direct-seeded rice in two years.

Table 2Effects of mechanical deep placement of nitrogen fertilizer on nitrogen use efficiency in hill direct-seeded rice in two years.

Table 3Nitrogen management practices affected global warming potential(GWP)in hill direct-seeded rice in two years.

3.2.Nitrogen use efficiency

OTDP fertilization treatment significantly affected TNA,NGPE,NHI,NAE,and NRE in both rice cultivars during 2016 and 2017(Table 2).In both years,significant differences in TNA were found between OTDP fertilization and multiple surface broadcasting.The TNA of the two rice cultivars for OTDP fertilization treatment was 150.6–159.1 kg ha-1,values 10.3%–13.1% higher than those under multiple surface broadcasting treatment.Maximum NGPE was observed in CK,followed by OTDP fertilization treatment,and the lowest NGPE was found for multiple surface broadcasting treatment.The NGPE for OTDP fertilization treatment was markedly larger than that for multiple surface broadcasting treatment,whereas OTDP fertilization and the control treatments were statistically similar(P＞0.05)with respect to NGPE in both cultivars.OTDP fertilization treatment led to the largest NAE among all treatments,a value markedly greater than that for the other treatments.The NHI for OTDP fertilization treatment was larger than that for multiple surface broadcasting treatment.The NAE for WFY615 and YXYZ ranged from 10.15 to 20.45 kg kg-1and 8.55 to 16.52 kg kg-1during 2016 and 2017.Similarly,OTDP fertilization treatment showed the maximum NRE,while multiple surface broadcasting treatment showed the lowest values for NRE among all treatments(Table 2).The NRE of WFY615 was higher than that of YXYZ in both years.Overall,OTDP fertilization increased TNA by 10.2%–13.2%,NGPE by 29.7%–31.5%,NHI by 27.8%–30.0%,NAE by 71.3%–77.2%,and NRE by 42.4%–56.4%for the two rice cultivars,compared with multiple surface broadcasting,in both years.Large differences in TNA,NGPE,and NRE between the two rice cultivars were also found.Marked differences in TNA,NGPE,NAE,and NRE were recorded between N treatments.There were significant interactions between N treatment and rice cultivar in TNA,NGPE,NAE,and NRE(Table 2).

3.3.GS and NR activities and LAI

Fertilization method affected GS activity in flag leaves at MT,PI,and HS.During the PI,HS,and MS stages,the highest GS activities of WFY 615 and YXYZ in both years were observed under OTDP fertilization.The PI and MS stages differed significantly in GS activity,whereas the GS activity in WFY 615 was higher than that of YXYZ in 2016(Fig.1).Marginal differences in NR activity in flag leaves of both rice cultivars were observed at MT.During the PI and HS stages,the highest NR activities in YXYZ during both years were found under OTDP fertilization compared to multiple surface broadcasting.A similar trend was recorded for WFY 615 in 2016,but no obvious difference in NR activity was found between OTDP fertilization and multiple surface broadcasting in 2017(Fig.2).

Some differences in LAI were noted among all treatments.At MT and PI,the LAI values for multiple surface broadcasting treatment were significantly higher than those for OTDP fertilization treatment;however,larger LAI was observed at HS and MS in OTDP fertilization than in multiple surface broadcasting treatment.The results suggested that OTDP fertilization treatment could sustain a longer growth period than multiple surface broadcasting treatment(Fig.3).

3.4.Root morphological traits at heading stage

Fig.4 shows marked differences in root morphological traits:total root length,average root diameter,root superficial area,and total root volume,at heading stage.OTDP fertilization treatment produced the largest root length per hill among all treatments for both rice cultivars.Total root lengths of the two rice cultivars under OTDP fertilization treatment were 35.06 and 30.15 m per hill,values respectively 29.09%and 25.96%larger than those under multiple surface broadcasting treatment.The mean root diameters of WFY 615 and YXYZ for OTDP fertilization treatment were 0.41 and 0.37 mm,values respectively 9.00% and 13.24% larger than those under multiple surface broadcasting treatment.There was significantly larger total root volume per hill for both rice cultivars under OTDP fertilization than under multiple surface broadcasting treatment.

3.5.Total global warming potential of CH4 and N2O

OTDP fertilization treatment significantly mitigated CH4and N2O fluxes compared to multiple surface broadcasting treatment(Figs.5 and 6).CH4emissions showed two peak fluxes,at the PI and GF stage.The CH4fluxes for WFY 615 and YXYZ ranged from 5.02 to 62.31 mg m-2h-1and from 6.47 to 64.52 mg m-2h-1,respectively.Only one peak in the N2O flux was observed at the end of MT.The N2O fluxes for WFY 615 and YXYZ ranged from 0.03 to 1.23 mg m-2h-1and from 0.03 to 1.57 mg m-2h-1,respectively.In comparison with multiple surface broadcasting treatment,CH4and N2O emissions were substantially decreased by 22.13% and 24.44%,respectively,for under OTDP fertilization treatment.

N fertilization markedly increased CH4-induced GWP,N2Oinduced GWP,and total GWP of CH4and N2O(Table 3).Significant differences were also found in CH4-induced GWP and total GWP among all treatments.CH4-induced GWP and GWP were highest for multiple surface broadcasting treatment,followed by OTDP fertilization treatment,and the lowest values were for the control treatment.CH4-induced GWP and GWP for WFY 615 from OTDP fertilization were respectively 7.81 and 9.30 t ha-1,values 25.31%and 22.30% lower than for multiple surface broadcasting treatment.CH4-induced GWP and GWP for YXYZ from OTDP fertilization were respectively 8.75 and 11.45 t ha-1,values 20.67% and 14.71% lower than for multiple surface broadcasting treatment.Significant decrease in N2O-induced GWP was noted for OTDP fertilization compared to multiple surface broadcasting treatment.OTDP fertilization significantly decreased GWP by 14.71%–22.90%for both rice cultivars in comparison with multiple surface broadcasting.There were also significant differences in CH4-induced GWP and total GWP between WFY 615 and YXYZ.Interactions between cultivar and nitrogen treatment were found for CH4-induced GWP,N2O-induced GWP,and total GWP(Table 3).

4.Discussion

Effective nitrogen fertilizer management practices can improve grain yield and nitrogen use efficiency[4,8,9].Bandaogo et al.[26]observed that N deep placement significantly increased rice grain yields in the wet season.However,Adviento-Borbe et al.[27]found that the effect of deep placement fertilization on rice grain yield was not significant compared to that of urea broadcast on the soil surface in continuously flooded rice systems.In the present study,OTDP fertilization treatment markedly increased the grain yield of direct-seeding rice in comparison to multiple surface broadcasting treatment.This increase was attributed mainly to an increase in yield components:productive panicles per m2and spikelets per panicle.It is well known[5,28]that a larger leaf area supplies more carbohydrates to roots for developing root functions,while active roots can supply more nutrients and water to aboveground organs of rice to increase biomass.OTDP fertilization increased rice root growth(Fig.4),increasing nutrient acquisition(Table 2),and leading to larger green leaf area at heading stage(Fig.3).The higher GS(Fig.1)and NR activities(Fig.2)in flag leaves following OTDP fertilization were found.

In a previous study[29],deep fertilization application reduced the loss of N fertilizer,resulting in increased NO3--N and NH4+-N concentrations in the rice root zone[10].Huda et al.[30]reported that urea-briquette deep placement markedly improved nitrogen recovery efficiency because of lower amounts of floodwater NH4+-N,compared to broadcasting of prilled urea.We found that OTDP fertilization treatment increased total N accumulation,N recovery,and N agronomic efficiencies compared to multiple surface broadcasting treatment.A stronger root system expressed as larger total root length and total root volume was observed for OTDP fertilization treatment(Fig.4),increasing nutrient absorption from the soil.The higher NAE and NRE of WFY 615 than for YXYZ in 2016 and 2017 which may be due to the stronger root system of hybrid than of inbred rice types,leading to higher absorption of nutrients under identical N treatments[5].

Generally,CH4and N2O emissions from paddy fields are the results of multi-factor interactions between rice plants and soil microbes and agronomic measures.Wang et al.[31]reported that both alternative wetting and drying regime and furrow irrigation markedly reduced CH4emissions because of increased soil redox potential,root and shoot biomass,root oxidation activity,and harvest index.In the present study,OTDP fertilization treatment significantly reduced CH4emission compared to multiple surface broadcasting treatment,owing to an extensive root system(Fig.4).The reason may be that OTDP fertilization treatment promoted rice growth in the form of larger total root length and total root volume.The resulting increase in O2transported into the soil would lead to higher CH4oxidation[32].

Usually,N fertilization affects N2O emissions from paddy fields.Chu et al.[33]found that deep fertilizer application significantly increased the total N2O emissions relative to surface application.However,Rychel et al.[34]reported that deep N fertilizer placement mitigated N2O emissions in a Swedish field trial with cereals.In the present study,the fertilization method showed a large effect on N2O emissions.OTDP fertilization significantly reduced the N2O emissions relative to multiple surface broadcasting.It is evident[35]that production of N2O occurs mainly near the soil surface because of the higher microbial nitrification and denitrification other than in deeper soil layers.Urea deep placement at the reduced zone retains N in NH4+form.Slow diffusion of NH4+-N from a reduced zone to the surface of the soil or in floodwater was observed[36].Thus,OTDP fertilization reduces the supply of inorganic N substrates in the most biologically active zone.Because rice roots tend to proliferate in a fertilization layer,deep fertilizer placement can promote rice root growth(Fig.4),and increase N uptake.The lower N substrate availability for OTDP fertilization could also explain the observed 24.44%reduction in N2O emissions in comparison with the multiple surface broadcasting.

Moreover,application of N fertilizer increased the total GWP of CH4and N2O.OTDP fertilization mitigated the GWP in comparison with multiple surface broadcasting treatment for both rice cultivars.The N2O-induced GWP accounted for only＜17.18% of total GWP among all treatments(Table 3).The CH4-induced GWP was the main source in the rice paddy field,a finding in agreement with Li et al.[37],who reported that CH4was the main greenhouse gas in paddy fields.CH4-induced GWP,and total GWP of CH4and N2O for WFY 615 were significantly higher than those for YXYZ.WFY 615 was recently bred as a hybrid rice type with high grain yield(Table 1),in agreement with the finding of Jiang et al.[32,38]that new rice cultivars had larger panicle sizes,higher yields,and lower methane emissions.

5.Conclusions

One-time deep-placement fertilization treatment increased grain yield of WFY615 and YXYZ owing to increases in productive tillers per unit area and spikelets per panicle.Larger LAI at heading stage and stronger root morphological traits such as larger root length,average root diameter,and total root volume per hill were also observed for one-time deep-placement fertilization.This treatment increased NUE and the activities of NR and GS of flag leaves at panicle initiation and heading stages.The lowest total global warming potential of CH4and N2O was recorded for onetime deep-placement fertilization treatment.These results suggest that one-time deep-placement fertilization may be an effective technology for increasing rice yield and NUE and reducing greenhouse gas emissions in mechanical direct-seeded cropping systems.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The research was supported by the Key-Area Research and Development Program of Guangdong Province(2019B020221003),National Natural Science Foundation of Guangdong Province(008175187004),and National Natural Science Foundation of China(31471442).

CRediT authorship contribution statement

Shenggang Pan and Xiangru Tanginitiated and designed the research;Lin Li,Hua Tian,Umair Ashraf,Minghua Zhang,Haidong Liu,Xiongfei Chen,Zaiman Wang,Meiyang Duan,and Zheng Zhangperformed the experiments;Shenggang Pan and Pingshan Fananalyzed the data and wrote the manuscript;Umair Ashraf and Shenggang Panrevised and edited the manuscript.