HAN Yu-ling ,GUO Dong ,MA Wei ,GE Jun-zhu ,Ll Xiang-ling ,Ali Noor MEHMOODZHAO MingZHOU Bao-yuan

1 Institute of Crop Sciences,Chinese Academy of Agricultural Sciences/Key Laboratory of Crop Physiology and Ecology,Ministry of Agriculture and Rural Affairs,Beijing 100081,P.R.China

2 College of Agriculture,Northeast Agricultural University,Harbin 150030,P.R.China

3 College of Agronomy &Resource and Environment,Tianjin Agricultural University,Tianjin 300384,P.R.China

4 College of Agronomy and Biotechnology,Hebei Normal University of Science &Technology,Qinhuangdao 066004,P.R.China

Abstract Inappropriate tillage practices and nitrogen (N) management have become seriously limitations for maize (Zea mays L.)yield and N use efficiency (NUE) in the North China Plain (NCP).In the current study,we examined the effects of strip deep rotary tillage (ST) combined with controlled-release (CR) urea on maize yield and NUE,and determined the physiological factors involved in yield formation and N accumulation during a 2-year field experiment.Compared with conventional rotary tillage (RT) and no-tillage (NT),ST increased the soil water content and soil mineral N content (Nmin) in the 20–40 cm soil layer due to reduction by 10.5 and 13.7% in the soil bulk density in the 0–40 cm soil layer,respectively.Compared with the values obtained by common urea (CU) fertilization,CR increased the Nmin in the 0–40 cm soil layers by 12.4 and 10.3% at the silking and maturity stages,respectively.As a result,root length and total N accumulation were enhanced under ST and CR urea,which promoted greater leaf area and dry matter (particularly at post-silking),eventually increasing the 1 000-kernel weight of maize.Thus,ST increased the maize yield by 8.3 and 11.0% compared with RT and NT,respectively,whereas CR urea increased maize yield by 8.9% above the values obtained under CU.Because of greater grain yield and N accumulation,ST combined with CR urea improved the NUE substantially.These results show that ST coupled with CR urea is an effective practice to further increase maize yield and NUE by improving soil properties and N supply,so it should be considered for sustainable maize production in the NCP (and other similar areas worldwide).

Keywords:maize,strip deep rotary tillage,controlled-release urea,nitrogen accumulation,grain yield

1.lntroduction

Maize (ZeamaysL.) is one of the most important cereal crops worldwide and has an important role in expanding grain production capacity.The North China Plain (NCP)is one of China’s largest agricultural areas,where more than 35% of the nation’s maize is produced.During the last few decades,the grain yield of maize in the NCP has been improved greatly in response to mineral nitrogen (N)fertilization,mechanization,and other agronomic practices(NBSC 2020).However,inappropriate mechanized tillage practices and N fertilization seriously limit maize yield,decrease nitrogen use efficiency (NUE) (Zhaoet al.2006;Liuet al.2010),and result in groundwater and air pollution(Heet al.2009;Juet al.2009;Guoet al.2010).

To save time and labor,successive years of rotary tillage(RT) using heavy machinery prior to the sowing of wheat,and direct stubble sowing with zero tillage for maize,have recently been adopted as the main farming practices on the NCP.However,long-term no-tillage (NT) treatment increases soil bulk density (Jinet al.2007;Kanet al.2020),and continuous RT leads to the formation of a hard plough pan at a depth of 15 cm (Hassan and Gregory 1999;Wanget al.2015).Penetration resistance increases when soil bulk density is high and a plough pan is present (Whalleyet al.1995;Qinet al.2018),and water distribution,nutrient distribution,root development,and root extension are restricted throughout the soil profile (Govaertset al.2006;Hartmannet al.2012;Bianet al.2016;Fioriniet al.2018).These current agronomic practices reduce the absorption of water and N from the deep layers (Guanet al.2015;Wanget al.2015),and eventually increase N loss during the early crop growth stages,leading to inadequate N supply during the late growth stages (Sunet al.2017).

Across most of the NCP,farmers with inadequate knowledge and limited access to appropriate technologies have applied large amounts of chemical N fertilizer to increase crop yields (Cuiet al.2010);they routinely apply almost 60% of the annual N fertilizer dose once prior to sowing maize,or early in the growing season (Menget al.2012),which exceeds the current growth requirements of maize.Excess N is likely to be lost through leaching under irrigation or high rainfall (Zhaoet al.2009),particularly in fields with high soil bulk density.These N losses lead to environmental damage (Liuet al.2010;Zhanget al.2010).Furthermore,excess application of N during early growth stages accelerates leaf senescence in maize because insufficient N is available during the late growth stages (Dinget al.2005).

Great efforts are being made to develop effective tillage and N fertilization practices,to simultaneously improve maize yield and NUE.Many different approaches,such as soil tillage,nitrogen fertilization,other agricultural management practices,and the interactions among them,have been developed recently to increase crop yield or NUE (Piaoet al.2016;Zhaiet al.2017;Zhouet al.2017a;Qiet al.2020).Deep tillage practices such as subsoiling,which significantly improve the physical,biological,and hydraulic properties of soil,are effective for improving crop yield and nutrient and water use efficiency (Bottaet al.2006;Jabroet al.2011;Muet al.2016;Shiet al.2016;Wu Fet al.2021).In a new method of subsoiling tillage,i.e.,strip deep vertical rotary tilling (ST),the maize planting rows are tilled to a maximum depth of 40 cm and a maximum width of 15 cm using a strip rotary device to break up the plough pan and promote root growth and nutrient distribution into deeper soil layers,which eventually improves nutrient uptake and production(Wanget al.2015;Sunet al.2017).Moreover,controlledrelease (CR) N fertilization is among the most effective approaches for increasing crop yield and minimizing N losses (Genget al.2015,2016;Liet al.2017;Wu Qet al.2021).This is because the slow release of nutrients through the use of CR N fertilizers enables uniform fertilizer application to improve NUE,thereby reducing N loss to the environment and maintaining an adequate N supply during the late crop growth stages (Shavivet al.2001;Kiranet al.2010;Gaoet al.2015).There are clear and predictable interactions between tillage and N fertilization (Díaz-Zorita 2000;Zhanget al.2013;Zhouet al.2019).Huet al.(2011) reported that the CR urea fertilization with subsoiling increases maize yield and water use efficiency in a semi-humid region of China.

Combining ST with slow-release N fertilization would further improve NUE and maize yield simultaneously(Zhouet al.2016,2017a).However,the physiological determinants of N accumulation and yield formation affected by ST combined with CR urea are poorly understood.In this study,we aimed to evaluate the effects of tillage and N fertilization procedures on (i) soil bulk density,soil water content,mineral N content,and root length,all of which affect N accumulation,(ii) leaf area development and dry matter (DM) accumulation in terms of how they affect yield formation,and (iii) maize yield and NUE.The results of these investigations should provide data useful for determining the physiological processes associated with maize yield and NUE that are affected by ST combined with CR urea,and improve NUE by improving both tillage practice and nitrogen management,eventually providing guidance for sustainable maize production in the NCP and similar agricultural regions worldwide.

2.Materials and methods

2.1.Experimental site

A 6-year field experiment was carried out during the 2013–2018 summer maize growing seasons at Xinxiang Experimental Station (35°11′30′′N,113°48′E),Chinese Academy of Agricultural Sciences,which is located in Henan Province.Xinxiang is a typical region of the NCP,where the planting area of winter wheat/summer maize cropping system accounts for more than 80%of the arable fields.The region has a temperate and continental monsoonal climate,in which the annual mean air temperature is 14.3°C,the ＞10°C accumulated temperature is 4 653°C,sunshine hours are 2 319 h,and precipitation is 571 mm.Here,we report only the experiments conducted in 2017 and 2018.The daily meteorological data during the maize growing seasons in these two study years are shown in Fig.1.The soil of the region was sandy loam,with 12.6 g kg-1organic matter content,0.65 g kg-1total N,16.2 mg kg-1available P,110.0 mg kg-1available K,and a pH of 8.2 in the 0–20 cm depth layer prior to maize planting in 2013.

Fig.1 Daily precipitation and the maximum and minimum air temperatures during the maize growing seasons of 2017 and 2018.

2.2.Experimental design

We used a two-factor split-plot experimental design with three replicates.The main plots were subjected to three tillage practices (NT,RT,and ST).Under NT,we did not till throughout the whole growth period of summer maize;so,the only soil disturbances occurred during sowing and fertilizer applications.We used a rotary tiller (1GKN-250,Yungangxuangengjixie Co.,Ltd.,Jiangsu,China)to plough to a 15-cm depth in the RT plots.The rotary blades were spaced 75 cm apart.The machine was equipped with a set of finishing disks behind the main blades to break up the large soil clods.In the ST plots,we applied vertical RT in the maize planting rows prior to sowing.The soil was cut to a maximum depth of 40 cm and width of 15 cm using a strip deep rotary machine(Hehuinong Agricultural Science and Technology Co.,Ltd.,Zhangjiakou,China).The device had four vertical rotary blades with 60-cm spacing.We did not till between adjacent rows of maize.During the wheat season before maize,the rotary tillage was conducted to a 15-cm depth when planting wheat in each plot.The subplots consisted of three fertilization treatments:CK (no N applied,control),CU (common urea) and CR urea.Under CU,we applied a total of 225 kg ha–1N as CU (46.4% N content)split into two equal applications,one at planting and one at the maize six-leaf stage (V6).Under CR urea,we applied a total of 225 kg ha–1N of CR urea at planting in the form of a resin-coated CR urea with 43.47% N content produced by Shandong Kingenta Ecological Engineering Co.,Ltd.(Shandong Province,China).The longevity of CR urea in 25°C water is 3 mon for maize according to the growth period.In each experiment and year,we applied 75 kg P2O5ha-1as Ca(H2PO4)2and 75 kg K2O ha-1as KCl in all of the plots at the time of maize planting.Each plot (4.8 m×35 m) comprised eight maize lines with an average spacing of 60 cm.Plots were separated from one another with 20-cm-high ridges and 50-cm-wide buffer zones to minimize inter-plot effects.

Zhengdan 958,one of the most widely grown maize cultivars in the NCP (NBSC 2016),was selected in this study.The maize planting dates were June 12,2017 and June 14,2018.Plants were harvested on October 4,2017 and October 7,2018.The final maize plant density was 60 000 ha-1with 60-cm row spacing.No obvious incidences of weeds,pests,disease,or water stress were observed in any of the plots.

2.3.Sampling and measurements

Dry matter (DM) accumulationThree plants in each plot were sampled randomly at the six-leaf stage (V6),12-leaf stage (V12),silking stage (R1),milking stage (R3) and physiological maturity (R6),which were identified using a standardized maize developmental staging system(Ritchieet al.1993).DM was measured by weighing the sample after it was dried to a constant moisture content.The dry matter accumulation rate (DMR) was obtained using the following logistic equation (Zhouet al.2017b):

whereais the total DM,bis the initial DM,cis a growth rate parameter,dis a derivation operation,eis a natural constant,xis the number of days,andyis the DM at each stage.

Leaf area index (LAl)Leaf areas of harvested plants were determined at the developmental stages V6,V12,R1,R3 and R6.We measured the length and maximum width of each green leaf,and calculated leaf area as follows (McKee 1964):

N accumulationThe sampled plants were milled into a powder after drying to constant moisture content.Subsamples were digestedviawet oxidation and subjected to micro-Kjeldahl distillation and titration to measure the N concentration in the maize plants (Bremner and Mulvaney 1982).N contents were calculated using the N concentration and the DM weight of each organ.The total N accumulation of the plant was given as the sum of all organ N contents (Zhouet al.2017b).

Soil bulk density (SBD)SBD was determined at the maize silking stage (R1) in 2017 and 2018 using the cutting ring method (Zhaiet al.2017).Samples were collected with a steel cylinder (6 cm in diameter and 30 cm high;total volume,847.8 cm3) which represented the 0–40 cm soil layer at 10-cm increments.Collections were performed in triplicate in each plot.SBD was determined as the dry weight of soil divided by its volume(Blake and Hartge 1986).

Soil water content (SWC)SWC was measured at the maize R1 stage in 2017 and 2018 using the ovendrying method (Zhaiet al.2017).Soil samples at 10-cm increments to a depth of 40 cm were collected with a soil auger (4.5 cm in diameter).Collections were performed in triplicate in each plot.SWC (g g–1) was calculated as the difference between sample soil fresh weight (FW)and soil dry weight (DW),calculated as follows (Maet al.2015;Zhouet al.2017b):

Soil mineral N content (Nmin)Nminvalues in the 0–40 cm soil layer at 10-cm increments were collected at the maize R1 and R6 stages in 2017 and 2018.Soil samples were obtained at equidistant points between pairs of plants.KCl (2 mol L-1) was used to extract the soil samples to determine ammonium N and nitrate N concentrations using a continuous flow auto-analyzer.

Root lengthRoot samples at 10-cm increments to a depth of 40 cm were collected after stalk sampling at the R1 stage in 2018 to calculate root length using the threedimensional (3D) spatially distributed monolith protocol(Wanget al.2015;Zhouet al.2019).Each soil profile had dimensions of 50 cm×50 cm×40 cm,and one maize plant was located at the center.Roots in each soil block were collectedviasubmersion in water followed by shaking in running tap water over an 853-μm sieve.Roots collected in this manner were washed until free of soil and scanned.Root lengths were calculated with WinRHIZO Software(Regent Instruments,Quebec,ON,Canada) based on the digital images thus produced.

Grain yieldWe manually harvested a 48-m2crop area in each maize plot at physiological maturity.The four center rows were counted to calculate the ear number per unit area.A total of 500 kernels were measured three times from each plot,and the result was converted to 1 000-kernel weight.The kernel number was measured as the average number on 20 ears for each plot.Maize yield was determined assuming 14% kernel water content(Zhouet al.2019).

N use efficiencyN partial factor productivity (PFPN),agronomic N efficiency (AEN),and the recovery efficiency of N (REN) were calculated as follows (Dobermann 2005):

whereYoandUoare grain yield (kg ha-1) and N accumulation in plants at maturity (kg ha-1),respectively,under CK;YN,GN,andUNare the yield (kg ha-1),N accumulation in kernels,and N accumulation in plants at maturity (kg ha-1),respectively;andFNis the N application rate (kg ha-1) under the N treatments.

2.4.Statistical analysis

A two-way analysis of variance using SPSS version 20.0 Software (SPSS Inc.,Chicago,IL,USA) was conducted to detect significant differences of maize grain yield and NUE among treatments with year,tillage type and fertilization treatment as fixed effects.The DM,DM accumulation rate,LAI,and maize N accumulation were subjected to repeated-measures analysis,with treatment and stage at sampling (repeated measurement) as fixed effects.Root length per plant,Nmin,and water content were subjected to repeated-measures analysis with treatment and soil layer depth (repeated measurement) as fixed effects.Tests for the normality of residuals were determined using quantile-quantile plots and the homogeneity of variance was conducted using Levene’s test.Replication and interactions with replication were treated as random effects.Means were compared using Fisher’s LSD test at the 5% level of probability.

3.Results

3.1.Grain yield and yield components

Grain yield was significantly affected by tillage,N treatment,and the interaction between year and N treatment (Table 1).The grain yield under ST was 9.1 Mg ha–1across years and N treatments,and this yield was 8.3 and 11.0% higher than those under RT and NT,respectively.We found no obvious differences in yield between RT and NT,regardless of year or N treatment.Across years and tillage practices,CR urea produced the highest grain yield among N treatments,whereas CK produced the lowest.The grain yield under CR urea was 9.8 Mg ha–1across years and tillage practices,and this yield was 8.9 and 42.0% higher than those under CU and CK,respectively.

Ear number per hectare and kernel number per ear varied significantly by N treatment,and the interaction between year and N treatment;the 1 000-kernel weight varied significantly by tillage,N fertilization treatment,the interaction between year and N treatment,and that between tillage and N treatment (Table 1).Ear number and kernel number were not significantly different among tillage practices.The 1 000-kernel weight under ST was 5.7 and 6.2% higher across years and N treatments than under RT and NT,respectively.Ear number and kernel number under CR and CU were higher than under CK,but these parameters were not significantly different between CR and CU.The 1 000-kernel weight under CR was 6.0 and 14.0% higher across years and tillage practices than under CU and CK,respectively.

Table 1 Grain yield and yield components of summer maize by tillage and N treatments in 2017 and 2018

3.2.DM and DM accumulation rate

Maize DM differed by tillage practice,N fertilizer treatment,growth stage,and their interactions (Fig.2).We found no obvious differences in maize DM among the three tillage practices at the V6 stage regardless of year or N fertilization treatment.However,DM under ST was significantly higher than those under RT and NT at the V6–R1 and post-silking stages,but there was no obvious difference in DM between RT and NT.DM accumulation under ST was 7.9 and 10.6% higher at the R1 stage than under RT and NT,respectively,and 10.0 and 13.3% higher at the post-silking stage.Across years and N fertilization treatments,total DM accumulation at maturity (R6) under ST was 8.9 and 11.9% higher than under RT and NT,respectively.Across years and tillage practices,maize under CR and CU accumulated more DM than under CK at all developmental stages.DM accumulation under CR,in the V12–R1 and post-silking stages,exceeded that under CU.DM accumulation under CR at the R1 and post-silking stages was 11.0 and 20.1%higher,respectively,than under CU.Overall,the total DM accumulation at maturity (R6) under CR was 15.4%higher than under CU.

Fig.2 Dry matter (DM) and dry matter accumulation rate (DMR) of maize by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE(n=3).Different lower-case letters adjacent to the error bars denote significant pairwise differences between means for a given date after sowing (P＜0.05;LSD test).

The DM accumulation rate was affected by tillage practice,N fertilization treatment,growth stage,and their interactions (Fig.2).ST had significantly higher DM accumulation rates at the V6–R1 and post-silking stages than the NT and RT treatments,regardless of year or N fertilization treatment,but DM accumulation rates under RT and NT were not obviously different.The DM accumulation rates at stages V6–R1 under CR were 11.1 and 16.9% higher than under RT and NT,respectively,across years and N fertilization treatments.The DM accumulation rate in the post-silking stage under CR was 6.8 and 7.1% higher than under RT and NT,respectively,across years and N fertilization treatments.Across years and tillage practices,the DM accumulation rates under CR at the V12–R1 and post-silking stages were the highest among N treatments;the accumulation rates under CK were the lowest among treatments.The DM accumulation rates at stages V12–R1 under CR urea were 14.7 and 41.4% higher than accumulations under CU and CK,respectively.The DM accumulation rate post-silking under CR was 16.2 and 36.9% higher than under CU and CK,respectively.

3.3.Leaf area index

The maize LAI varied by tillage practice,N fertilization treatment,growth stage,and their interactions (Fig.3).We found no obvious difference in LAI at the V6 stage among the three tillage practices,regardless of year or N fertilization treatment.However,under ST,the LAI was significantly different from those under RT and NT at the V12,R1,R3,and R6 stages,but LAI values were not obviously different between RT and NT.The maximum LAI value at the R1 stage under ST was 10.2 and 14.7%higher than under RT and NT,respectively,across years and N fertilization treatments.The LAI at maturity (R6)under ST was 19.3 and 20.6% higher than under RT and NT,respectively.Across years and tillage practices,the LAI at V6–R6 stages under CR was the highest among the N treatments;the LAI under CK was the lowest.The maximum LAI at the R1 stage under CR urea was 14.8 and 42.0% higher than under CU and CK,respectively.The LAI at maturity (R6) under CR was 23.2 and 57.9%higher than under CU and CK,respectively.

Fig.3 Maize leaf area index by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE (n=3).Different lower-case letters adjacent to the error bars denote significant pairwise differences between means for a given date after sowing (P＜0.05;LSD test).

3.4.N accumulation

N accumulation varied by tillage practice,N fertilization treatment,growth stage,and their interactions (Fig.4).Plants under ST had significantly higher N accumulation rates at each stage than under the RT and NT treatments.We found no obvious difference in N accumulation between the RT and NT treatments.N accumulation under ST was 6.4 and 8.3% higher at the R1 stage than under RT and NT,respectively.At the post-silking stage,the accumulation rates under ST were 10.8 and 11.5%higher,respectively.The total N accumulation at maturity(R6) under ST was 8.3 and 9.6% higher than under RT and NT,respectively,across years and N fertilization treatments.Across years and tillage practices,maize under CR and CU accumulated greater amounts of N at all stages than plants under CK.The N accumulation rates under CR at the V6–R1 and post-silking stages exceeded those under CU.N accumulation rates at the R1 and post-silking stages under CR were 11.8 and 16.2% higher than those under CU,respectively.The total N accumulation at maturity (R6) under CR urea was 13.5% higher than accumulation under CU.

Fig.4 Nitrogen accumulation of maize plants by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE (n=3).Different lower-case letters adjacent to the error bars identify significant pairwise differences between means for a given date after sowing (P＜0.05;LSD test).

3.5.Root length

Root lengths in the 0–40 cm layer were determined at stage R1 in 2018.Values varied by tillage practice,N fertilization treatment,soil layer depth,and their interactions(Fig.5).We found no obvious differences in root lengths in the 0–20 cm layer among the tillage practices.However,across N treatments,the root length in the 20–40 cm layer under ST was 10.1 and 10.6% greater than under RT and NT,respectively.There was no obvious difference in the 20–40 cm layer root lengths between the RT and NT treatments.Across N treatments,the total root length in the 0–40 cm layer under ST was 5.0 and 5.4% higher than those under RT and NT,respectively.

Fig.5 Maize root length per plant at the silking stage within the 0–40 cm soil layers by treatment in 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE(n=3).Different lower-case letters adjacent to the error bars identify significant pairwise differences between means for a given tillage treatment (NT,RT or ST) (P＜0.05;LSD test).

Across tillage practices,root lengths in each of the soil layers from 0 to 40 cm depth under CR and CU exceeded those under CK.The root length in the 0–20 cm soil layer under CR exceeded the value under CU by 13.6%.The root length in the 20–40 cm soil layer under CR exceeded those under CU and ST by 12.2%.Under RT and NT,root lengths were not obviously different between the CU and CR treatments.The total root length in the 0–40 cm soil layer under CR was 11.9 and 38.4% greater than under CU and CK,respectively.

3.6.Soil mineral N content

Soil Nminvalues at the maize developmental stages R1 and R6 in 2017 and 2018 varied by tillage practice,N fertilization treatment,soil layer depth,and their interactions(Figs.6 and 7).We observed no obvious differences in the Nminvalues among tillage practices in the 0–20 cm layer at the R1 and R6 stages.However,the Nminvalue in the 20–40 cm layer under ST was 12.0 and 13.6% higher at the R1 stage,and 11.6 and 12.4% higher at the R6 stage,than under the RT and NT,respectively,across years and N treatments.We found no obvious differences in Nminin the 20–40 cm soil layer between RT and NT at the R1 and R6 stages.The total Nminin the 0–40 cm soil layer under ST was 6.1 and 5.6% higher at the R1 stage,and 6.1 and 6.2%higher at the R6 stage,than under RT and NT treatments,respectively,across years and N treatments.

Across years and tillage practices,the Nminvalues within each soil layer from 0 to 40 cm depths under CR and CU exceeded the values under CK.The Nminvalues in the 0–20 cm layer under CR at R1 and R6 stages were 16.8 and 10.1% higher,respectively,than under CU.Overall,the total Nminvalues in the 0–40 cm layer under CR at the R1 and R6 stages were 12.4 and 10.3%higher,respectively,than under CU.Under ST,CR urea increased the Nminvalues in the 20–40 cm layer at the R1 and R6 stages by 13.6 and 11.1% above the values obtained under CU,respectively.These differences were not observed under RT or NT.

Fig.6 Soil mineral N content (Nmin,i.e.,ammonium-N+nitrate-N) within the 0–40 cm soil depth layers at the maize silking stage by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE (n=3).Different lower-case letters adjacent to the error bars denote significant pairwise differences between means within the figure panels (P＜0.05;LSD test).

Fig.7 Soil mineral N content (Nmin,i.e.,ammonium-N+nitrate-N) within the 0–40 cm soil depth layers at the maize maturity stage by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE (n=3).Different lower-case letters adjacent to the error bars denote significant pairwise differences between means within the figure panels (P＜0.05;LSD test).

3.7.Soil bulk density and soil water content

We determined SBD in the 0–40 cm layers at the maize R1 stage in 2017 and 2018.Values varied by tillage practice,soil layer depth,and their interaction (Fig.8).ST reduced SBD in the 0–20 cm layer by 19.8 and 8.0%below the levels in RT and NT,respectively,across years and N fertilization treatments.ST also decreased the SBD in the 20–40 cm layer by 8.1 and 9.2% below the levels in RT and NT,respectively.We found no obvious differences in SBD in the 0–40 cm layer among N treatments regardless of year or tillage practice.

Fig.8 Soil bulk density in the 0–40 cm soil layers at the maize silking stage by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE(n=3).Different lower-case letters adjacent to the error bars denote significant pairwise differences between means within the figure panels (P＜0.05;LSD test).

The SWC in the 0–40 cm layers at the maize R1 stage in 2017 and 2018 varied by tillage practice,soil layer depth,and their interaction (Fig.9).ST and RT reduced the SWC in the 0–20 cm layer by 16.6 and 13.9%,respectively,below the level under NT across years and N treatments.We found no obvious difference in the SWC in this layer between ST and RT.

Fig.9 Soil water content within the 0–40 cm soil layers at the maize silking stage by treatment in 2017 and 2018.NT,no-tillage;RT,rotary tillage;ST,strip deep rotary tillage.CK,no N (control);CU,common urea;CR,controlled-release urea.Values are mean±SE (n=3).Different lower-case letters adjacent to the error bars denote significant pairwise differences between means within the figure panels (P＜0.05;LSD test).

ST increased the SWC of the 20–40 cm layer by 23.8 and 27.2% above the levels in NT and RT,respectively.We found no obvious difference in the water content in this soil layer between NT and RT.Overall,the total SWC in the 0–40 cm layer under ST was 14.2% higher than under RT,regardless of year or N treatment.We found no difference in water content between ST and NT.N fertilization treatment had no obvious effect on the SWC in the 0–40 cm layer.

3.8.Nitrogen use efficiency

ThePFPNandAENvaried by tillage,N fertilization treatment,and their interaction,whereasRENdiffered by tillage practice and N fertilization treatment (Table 2).Treatment ST had significantly largerPFPN,AEN,andRENvalues than RT and NT,regardless of year or N treatment.We found no obvious differences in these parameters between RT and NT.ThePFPN,AEN,andRENvalues under ST were 8.9,19.9,and 15.1% greater than those under RT,respectively,across years and N treatments,and 11.6,20.9 and 16.1% higher than those under NT,respectively,across years and N treatments.Across years and tillage practices,PFPN,AEN,andRENvalues under CR were 9.3,51.7,and 75.9% higher,respectively,than those under CU.

Table 2 N partial factor productivity (PFPN),agronomic N efficiency (AEN),and recovery efficiency of N (REN) of summer maize in 2017 and 2018 by tillage and N treatments

4.Discussion

Tillage practice and N management,and their interaction,are important parameters which influence maize productivity (Díaz-Zorita 2000;Menget al.2012;Miritiet al.2012;Guanet al.2014;Zhouet al.2017a;Xuet al.2019).However,inappropriate mechanized tillage practices and N fertilization seriously limit maize yield,and they decrease NUE by increasing SBD and N loss(Zhaoet al.2006;Liuet al.2010;Zhaiet al.2017).Previous studies have shown that both deep tillage (Wanget al.2015) and CR N fertilization (Liet al.2017) can improve maize grain yield and NUE concurrently.We tested the effects of ST combined with CR urea on maize productivity during the 2017–2018 period.Our results demonstrate that ST decreased the SBD in the 0–40 cm soil layer to a level 7.1% below those that occur when a hard plough pan with high SBD develops after long term application of NT or RT practices (Hassan and Gregory 1999;Kanet al.2020).Decreased SBD in the 0–40 cm soil layer may increase soil porosity,air permeability,and hydraulic conductivity (Guanet al.2015;Zhaiet al.2017),effects that will in turn promote root growth and nutrient distribution into deeper soil layers (Varsaet al.1997;Guanet al.2014;Wanget al.2015;Shaoet al.2016).Thus,we found that the SWC,root length,and Nminvalues in the 20–40 cm soil layer at both the R1 and R6 stages under ST were significantly improved in comparison with the values obtained with other tillage practices,regardless of N fertilization treatment.These effects increased total N accumulation in maize plants.Furthermore,the slow release of nutrients that occurs under CR urea fertilization ensures a more uniform application of N,thus ensuring adequate N during the late growth stages of maize (Shaoet al.2013;Zhaoet al.2013).We found that CR urea fertilization significantly increased the Nminvalue of the 0–20 cm soil depth layer at both the silking and maturity stages above those in the other Nfertilization treatments,regardless of tillage,and this eventually increased N accumulation in the maize plants.CR urea fertilization also increased the Nminvalue of the 20–40 cm soil layer above those obtained in the other N fertilization treatments,but only under ST,indicating that the combination of strip deep rotary tilling and CR urea fertilization further improved soil Nmincontent.The increased root length and soil Nmincould significantly improve N accumulation in plants (Wanget al.2015;Zhouet al.2019).Thus,we found that maize under the ST and CR urea fertilization treatment accumulated greater N during the pre-and post-silking stages compared with the other tillage and N fertilization treatments.

Greater N accumulation in plants results in a higher LAI and greater DM accumulation (Hirelet al.2007;Yanet al.2014).Increased N accumulation during the early growth stages increases leaf area (Zhouet al.2017b),thereby increasing radiation interception and carbon accumulation(Echarteet al.2008;Ciampitti and Vyn 2011).Moreover,adequate post-silking N accumulation increases canopy longevity,thereby delaying leaf senescence and decreasing the photosynthetic activity as the season progresses (Rajcan and Tollenaar 1999;Echarteet al.2008;Zhouet al.2017b),which promotes enhanced DM accumulation after silking (Ninget al.2013;Yanet al.2014).Thus,both the LAI and DM accumulation rate during the V12–R1 and post-silking stages obtained with the ST and CR urea fertilization treatment were substantially higher than the values obtained with other tillage procedures and N fertilization protocols.As a result,increased DM accumulation (particularly postsilking) was observed during both years under ST and CR urea fertilization compared with other tillage and N fertilization treatments.

Many studies have demonstrated that a greater capacity of DM accumulation (particularly post-silking)contributes significantly to maize yield (Mollet al.1994;Ninget al.2013;Sunet al.2017).The improved capacity of plants to supply assimilates during post-silking by enhancing the N supply (Echarteet al.2008;Zhouet al.2017b),as is provided by our ST/CR urea fertilization treatment combination,resulted in an increase of kernel weight at maturity,and an increase in maize yield.Our results indicate that ST increased maize grain yield by 8.3 and 11.0% compared to the NT and RT practices,respectively.CR urea fertilization increased maize grain yield by 8.9 and 42.0% compared to CU fertilization and the no N control,respectively.The combination of ST and CR urea fertilization produced a higher maize grain yield than those obtained from any of the other treatments or treatment interactions,which was consistent with the previously published results obtained in 2013 and 2014(Zhouet al.2016,2017a),as well as with those of other investigations (Huet al.2011).The increased grain yield obtained with the ST/CR treatment was largely attributable to the high 1 000-kernel weight produced by the ST and CR combination.Similar results (Zhaoet al.2013) were also found in other studies,that due to higher N uptake after anthesis of maize under CR fertilizer treatment,maize grain yield increased by 9.69 to 14.15%compared with the treatment using a common compound fertilizer due to an improvement in the 1 000-kernel weight.Increases in grain yield and N accumulation in maize were achieved by applying the ST/CR urea fertilization treatment,with improvements inPFPN,AEN,andRENcompared with the other tillage and N fertilization protocols.These results are consistent with those of an experiment conducted in the same plots using the same design in 2013 and 2014 (Zhouet al.2016,2017a).

Thus,ST led to the formation of a highly appropriate soil environment which allowed the roots to develop and take up N when combined with an adequate N supply provided by CR urea fertilization.This procedure substantially improved both maize production and NUE.Therefore,ST combined with CR urea can be considered for sustainable agricultural development in the NCP,particularly in areas which are experiencing severe N overuse.However,further investigation of the economic benefits and environmental sustainability of our agronomic practice would be highly desirable.

5.Conclusion

Combining ST with CR urea fertilization was an effective approach for improving the yield and NUE of maize simultaneously,and this approach can be used to promote the sustainability of maize production in the NCP.ST increased the SWC,root length,and soil Nmincontent in the 20–40 cm soil layers by decreasing soil bulk density across the 0–40 cm soil layers.CR urea fertilization improved the Nmincontent in the 0–40 cm soil depth layer after the maize silking stage,which ensured an adequate post-silking N supply for the crop,thereby increasing N accumulation in the plants.In turn,greater N accumulation promoted the increase of LAI and DM accumulation,which eventually improved maize production and NUE substantially.

Acknowledgements

This research was funded by the National Natural Science Foundation of China (32071957),the Key National Research and Development Program of China(2018YFD0300504),the Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences (2060302-2),and the China Agriculture Research System of MOF and MARA (CARS-02).

Declaration of competing interest

The authors declare that they have no conflict of interests.