Integrated agronomic practice increases maize grain yield and nitrogen use efficiency under various soil fertility conditions

Boyun Zhou, Xuefng Sun, Dn Wng, Zisong Ding, Congfeng Li, Wei M,*,Ming Zho,*

aInstitute of Crop Sciences,Chinese Academy of Agricultural Sciences, Key Laboratory of Crop Physiology& Ecology, Ministry of Agriculture,Beijing 100081,China

bCollege of Plant Science and Technology, Huazhong Agricultural University, Key Laboratory of Crop Physiology, Ecology and Cultivation(Middle Reaches of the Yangtze River)of Ministry of Agriculture,Wuhan 430070,Hubei, China

Keywords:Summer maize Integrated agronomic practice Soil fertility Grain yield Nitrogen use efficiency

A B S T R A C T Crop yield potential can be increased through the use of appropriate agronomic practices.Integrated agronomic practice(IAP)is an effective way to increase maize(Zea mays L.)grain yield and nitrogen use efficiency (NUE); however, the physiological processes associated with gains in yield potential obtained from IAP,particularly the different under various soil fertility conditions, remain poorly understood. An IAP strategy including optimal planting density, split fertilizer application, and subsoiling tillage was evaluated over two growing seasons to determine whether the effects of IAP on maize yield and NUE differ under different levels of soil fertility. Compared to farmers' practices (FP), IAP increased maize grain yield in 2013 and 2014 by 25%and 28%,respectively,in low soil fertility(LSF)fields and by 36% and 37%, respectively, in high soil fertility (HSF) fields. The large yield gap was attributed mainly to greater dry matter (DM) and N accumulation with IAP than with FP owing to increased leaf area index (LAI) and DM accumulation rate, which were promoted by greater soil mineral N content (Nmin) and root length. Post-silking DM and N accumulation were also greater with IAP than with FP under HSF conditions, accounting for 60%and 43%,respectively,of total biomass and N accumulation;however,no significant differences were found for post-silking DM and N accumulation between IAP and FP under LSF conditions.Thus,the increase in grain yield with IAP was greater under HSF than under LSF.Because of greater grain yield and N uptake,IAP significantly increased N partial factor productivity, agronomic N efficiency, N recovery efficiency, and physiological efficiency of Abbreviations:HSF,high soil fertility field;LSF,low soil fertility field;CK,control without nitrogen; FP,conventional farmers'practice;IAP,integrated agronomic practice;DM,dry matter;LAI,leaf area index;N,nitrogen;Nmin,mineral N content(ammonium N and nitrate N); NUE, nitrogen use efficiency; AEN, agronomic N efficiency; PFPN, N partial factor productivity; REN, recovery efficiency of N; PEN,physiological efficiency of applied N;V6, six-leaf stage; V12,12-leaf stage; R1, silking stage; R3, milk stage; R6,physiological maturity applied N compared to FP, particularly in the HSF fields. These results indicate that considerable further increases in yield and NUE can be obtained by increasing effective soil N content and maize root length to promote post-silking N and DM accumulation in maize planted at high plant density, especially in fields with low soil fertility.

1. Introduction

Maize(Zea mays L.),is one of the most important cereal crops in efforts to expand global grain production. In recent decades, maize yield has been greatly improved by increased mineral fertilization and mechanization and changes in planting density [1,2]. However, a 40% yield gap exists between maize farm yields and potential yield under nonstress conditions, suggesting a potential yield increase of 1.8-2.6 Mg ha-1[3]. In addition, harmful agronomic practices such as excessive fertilizer application have resulted in low fertilizer use efficiency and environmental damage [4,5].Thus, a challenge is to further improve the grain yield of maize while simultaneously increasing fertilizer use efficiency.

In the North China Plain, summer maize is planted immediately after winter wheat each year in a doublecropping system. There are many problems associated with the conventional cropping practices of smallholder farming,owing to lack of appropriate knowledge and technologies among farmers. For example, excessive N (exceeding the wheat-maize cropping system requirements of 200 to 300 kg N ha-1) fertilizers are applied to maximize crop yields,but fertilization techniques lag behind[6-9].Large amounts of N (60% of total N application) are often applied before maize sowing or during early growth stages (six-leaf stage, V6)[10,11],and only about 10%of the N is used efficiently[12,13].The remaining N is susceptible to loss through ammonia volatilization (120 kg N ha-1), denitrification (16 kg N ha-1),and leaching (136 kg N ha-1), particularly during irrigation and periods of high rainfall[14-17].These processes may lead to insufficient N remaining in the soil during later growth stages, resulting in accelerated leaf senescence [18-20] and environmental damage [21-24]. Increasing the planting density to reduce the rate of N application per plant is an effective way to increase NUE and maize yield. Record high maize yields have typically been obtained at high planting density.Owing to excellent canopy structure and maximized light interception [1,25]. In the North China Plain, grain yield greater than 15 Mg ha-1was obtained when maize hybrid planting densities were in the range of 80,000-120,000 plants ha-1[1,26]. However, the planting density of summer maize hybrids in farmers'fields in this region is considerably lower,below 60,000 plants ha-1. In addition, long-term minimal or no-tillage during the maize growing season increases the soil bulk density [27-29], while continuous conventional rotary tillage before wheat sowing causes the formation of a hard plow pan at a depth of approximately 15 cm [30]. Soil with high bulk density and a plow pan hinders N distribution,root growth, and root extension [31,32]; reduces the absorption of nutrients in the deep soil layers [33]; and eventually reduces yield and NUE. For these reasons, efforts should be made to improve cropping practices to further increase maize yield and NUE simultaneously.

Recently, many approaches that focus on fertilizer management, plant density, soil tillage, and other agricultural management practices or their interactions have been developed to increase maize grain yield and N use efficiency [34-37]. These approaches have demonstrated that greater integration of agronomic management factors such as planting density,nutrients,and water can increase crop yields[38-41].For example,in high-yielding maize experiments,a combination of optimized agronomic practices, including optimal planting density, balanced fertilizer use, and planting methods produced grain yields greater than 15 Mg ha-1in the North China Plain [40]. However, these yields were typically obtained in fields with high soil fertility, whereas more than two-thirds of China's farmland consists of medium- to low-yielding fields with excessive permeability and low nutrient retention capacity[42].The effect of IAP on maize yield and NUE under various soil fertility conditions, and the physiological processes that contribute to yield,remain poorly understood. We hypothesized that the effect of IAP incorporating integration of split N application fertilizer, subsoiling tillage, and optimal planting density on maize growth and grain yield would differ between high and low soil fertility conditions. We evaluated the effects of IAP on maize grain yield, DM production, and root growth; recorded the changes in soil Nmin, N accumulation in plants, and NUE caused by IAP; and investigated whether the physiological processes affected by IAP differ between high and low soil fertility conditions.

2. Materials and methods

2.1. Site description

Field experiments were conducted during the maize growing season (June to October) in 2013 and 2014 at the Xinxiang Experimental Station, Chinese Academy of Agricultural Sciences (35°11′30″N, 113°48′E), located in Xinxiang county, Henan province. Xinxiang county is typical of the North China Plain in being agriculturally intensive with more than 80% of the agricultural fields planted with a winter wheat (Triticum aestivum L.) and summer maize cropping system.This region is in a temperate zone,with a continental monsoon climate and annual mean temperature of 14 °C,cumulative temperature above 10 °C of 4647 °C, total of 2324 sunshine hours,and precipitation of 573 mm.Fig.1 and Table S1 show the daily mean temperature and rainfall during the2013 and 2014 maize growing seasons.

Fig.1-Daily precipitation and mean temperature during the 2013 and 2014 maize growing seasons.

2.2. Soil properties

HSF and LSF were selected for this study according to different soil properties and representation of high-yielding fields with more than 12 Mg ha-1grain yield and medium-to low-yielding fields with less than 9 Mg ha-1grain yield. The soil type, bulk density, organic matter content, total N, readily available phosphorus (P) and readily available potassium (K) in the 0-50 cm depth of the soil profile are described in Table 1.

2.3. Experimental design and field management

Experiments were conducted in both HSF and LSF fields and were laid out in a randomized complete block design with three replicates. Both types of fields received the same three management treatments: CK, FP, and IAP. The FP treatment followed local farmers'conventional cropping practices with a planting density of 60,000 plants ha-1and 300 kg ha-1of total applied N. N fertilizer was applied as urea by hand as a broadcast and side-dress application over the soil surface with irrigation or precipitation before planting and at V6. No tillage was conducted in the fields after 2010. The only disturbance to the soil occurred during planting and fertilizer application.The IAP treatment tested an optimal combination of a planting density of 84,000 plants ha-1with 300 kg ha-1total applied N. N fertilizer was split-applied as urea with irrigation or precipitation before planting and at the V6, V12,R1 stages. Before 2013, long-term minimal or no-tillage was performed during the maize growing season. After 2013,subsoiling tillage was performed in the maize rows before sowing to a depth of 30 cm and a width of 10 cm using a stripe deep loosening machine (Hehuinong Machine Co. Ltd., Beijing, China). The machine had four vertical rotary shovels with alternating wide and narrow spacing of 40 and 80 cm,consistent with the maize row spacing. The CK treatment followed the conventional farmers' practice but with no N fertilizer application. For all three treatments, crop residues after wheat harvest were flattened and left on the soil surface.Irrigation was applied by conventional flood irrigation during dry weather conditions or at fertilizer application. Details of the cropping system and N application for the three treatments are shown in Table 2. Applications of 150 kg P2O5ha-1as Ca(H2PO4)2and 150 kg K2O ha-1as KCl were made at maize planting for all of the plots according to the amounts of determination of available P and K in the soil and maize requirements.

The maize hybrid cultivar Zhengdan 958(Henan Academy of Agricultural Sciences) was selected for this study. It is characterized by high yields, high quality, multiple disease resistance, early maturity, and extensive adaptability and is one of the most widely grown maize cultivars in the North China Plain[43].Maize was planted on June 12,2013 and June 14,2014,with harvest on October 4,2013 and October 7, 2014.Seeds were sown by hand after a 3-cm-deep trench was made with a hand hoe for proper seed placement with alternating wide and narrow spacings of 40 and 80 cm in each plot. The plot size was 124.8 m2(4.8 m × 26.0 m),with a distance of 1 mbetween each block. Each plot consisted of eight rows of maize.Water,weeds,pests,and diseases were well controlled during the maize growing seasons in 2013 and 2014.

Table 1-Soil properties before maize planting for high soil fertility and low soil fertility fields.

Table 2-Cropping system and fertilizer application details for different management practices in 2013 and 2014.

2.4. Soil analysis

Soil samples from the 0-50 cm soil layer collected before maize planting and soil tillage in 2013 were air-dried, sieved,and then used to measure organic matter [44], total N [45],extractable Olsen-P [46], ammonium acetate extractable K[47],and pH(soil:water = 1:2.5).

Soil bulk density was measured at R1 in 2013 using the cutting-ring method and was calculated as follows: soil bulk density(g cm-3) = soil dry weight(g)/cutting-ring volume(cm3).Measurements were made at five soil depths(0-10,10-20,20-30,30-40,and 40-50 cm),and repeated three times in each plot.

Nminvalues of 0-10,10-20,20-30,30-40,and 40-50 cm soil layers were determined at R1 in 2014.Soil samples were taken at the center between two rows after 24 h of irrigation for both conventional fertilization and drip fertigation treatments.The soil samples were extracted with 2 mol L-1KCl to analyze ammonium N and nitrate N by continuous flow analysis(TRAACS 2000,Bran and Luebbe,Norderstedt,Germany).

2.5. Crop sampling and measurements

To identify the stage of crop development, the standardized maize developmental staging system was used [48], and the date was recorded when more than 50%of the maize plants in each plot reached V6,V12,R1,R3,and R6.

Green leaf areas of the sampled plants were measured at V6, V12, R1, R3, and R6. For each leaf, the length and maximum width were recorded and the leaf area was computed based on the following formula[49]:

LAI was calculated as follows:

Plant samples were collected to determine DM content at V6, V12, R1, R3, and R6. Three adjacent plants (at least 1 m from the plot edge and 0.5 m from previous sampling sites)in each row were sampled randomly from each plot. The sampled plants were dried at 105 °C for 30 min and then at 70 °C to constant moisture content before being weighed.

Root samples from 0-10,10-20,20-30,30-40,and 40-50 cm soil layers were taken after stalk sampling at R1 in 2014 according to the three-dimensional (3D) spatially distributed monolith scheme commonly used in other study [30]. These roots were washed free of soil and scanned into images to calculate root length using the software WinRHIZO (Regent Instruments,Quebec,Canada).

At physiological maturity, 34 m2of the crop area was harvested manually from the four center rows in each plot.The ears were counted at harvest from the four center rows in each plot to determine the number of ears per ha. The 1000-kernel weight was calculated from the average of three random samples of 500 kernels. The kernel number was recorded as the mean kernel number of 10 ears from each replication. Grain yield was calculated at 14% moisture content.

Subsamples of plants were taken to determine N content,using the Kjeldahl procedure. N partial factor productivity(PFPN), agronomic N efficiency (AEN), recovery efficiency of N(REN), and physiological efficiency of applied N (PEN) were calculated as follows[50]:

where,Y0and U0are grain yield(kg ha-1)and N accumulation in plants at maturity(kg ha-1),respectively,for CK;YN,GN,and UNare grain yield(kg ha-1),N accumulation in kernels,and N accumulation in plants at maturity(kg ha-1), respectively,for N application treatments; and FNis the N rate (kg ha-1) in N application treatments.

2.6. Statistical analysis

Each experiment was subjected to ANOVA using SPSS 16.0(SPSS Institute, Inc., Chicago, IL). The soil fertility effect,cropping practice effect, and the interaction of soil fertility and cropping practice were tested using the full error term.Means were compared using the least significant difference(LSD)test at the 0.05 probability level.

3.Results

3.1. Grain yield and yield components

Grain yields differed significantly between soil fertility levels,treatments, and their interaction for both years (Table 3).Grain yield across the two years was 23%-37%greater for all of the treatments in HSF fields than for the corresponding treatments in LSF fields. The highest yield regardless of year and soil fertility condition was produced with IAP,followed by FP,while CK produced the lowest yield.In the LSF fields,grain yields with IAP were 9.0 and 9.5 Mg ha-1in 2013 and 2014,respectively, 25% and 28% greater than those produced with FP. In the HSF fields, grain yields with IAP were 12.1 and 12.5 Mg ha-1in 2013 and 2014, respectively, 36% and 37%greater than those produced with FP.The grain yield produced with FP corresponded to the average production level of local farmers'fields in Henan province.

Ear number per hectare, kernel number per ear, and 1000-kernel weight varied significantly between soil fertility conditions, treatments, or their interaction (Table 3). Ear number and kernel number were greater with all of the treatments in the HSF fields across the two years than in the LSF fields, while no significant differences were found in 1000-kernel weight between HSF and LSF fields (except for CK). Among the cropping practices, IAP consistently produced the greatest number of ears per hectare regardless of year and soil fertility, with a rank order for ears per hectare of IAP ＞FP ＞CK similar to that of grain yield. In the LSF fields, ear numbers produced with IAP were 51% and 52% greater in 2013 and 2014, respectively, than those with FP, while those were 47% and 54% greater than those with FP in the HSF fields.

3.2. DM, DM accumulation rate, and LAI

DM production was significantly affected by soil fertility,cropping practices, or their interaction (Fig. 2). Among the three treatments,IAP produced the greatest DM accumulation regardless of the year or soil fertility level with a ranking order of IAP ＞FP ＞CK similar to that of grain yield.In the LSF fields,DM accumulation values at R1 with IAP in 2013 and 2014 were 36% and 30% greater, respectively, than those produced with FP.At R6,the total DM values with IAP in 2013 and 2014 were 36%and 31%greater,respectively,than with FP.With IAP,DM accumulation post-silking (R1 to R6) accounted for 54% and 54% of total DM accumulation in 2013 and 2014, respectively,and no significant differences for those were found between IAP and FP. In the HSF fields, DM accumulation values at R1 with IAP in 2013 and 2014 were 18% and 13% greater,respectively, than with FP. At R6, the total DM accumulation values with IAP in 2013 and 2014 were 35% and 26% greater,respectively, than with FP. With IAP, the amounts of postsilking DM accumulation with IAP accounted for 60%and 59%of total DM accumulation in 2013 and 2014, respectively,which were greater than with FP. Post-silking DM accumulation under HSF conditions with IAP increased by 44%and 26%in 2013 and 2014, respectively, relative to post-silking DM accumulation under LSF conditions with IAP. No significant differences were found in DM accumulation pre-silking(sowing to R1)between LSF and HSF with IAP.

The DM accumulation rate was affected by soil fertility,cropping practice, and their interaction (Fig. 2). In the LSF fields,the DM accumulation rates with IAP from sowing to R1 in 2013 and 2014 were 64%and 50%greater,respectively,than those with FP. From R1 to R6, the DM accumulation rate with IAP in 2013 and 2014 were 35% and 29% greater, respectively,than those with FP. In the HSF fields, the DM accumulation rates with IAP from sowing to R1 in 2013 and 2014 were 25%and 27% greater,respectively, than those with FP. From R1 to R6,the DM accumulation rates with IAP in 2013 and 2014 were47% and 49% greater, respectively, than those with FP.Compared to LSF conditions with IAP, the post-silking DM accumulation rate under HSF conditions with IAP increased by 35% and 25% in 2013 and 2014, respectively, whereas no significant differences were found in the pre-silking DM accumulation rate between LSF and HSF with IAP.

Table 3-Grain yield and yield components for maize in response to various treatments in 2013 and 2014.

Fig.2- Dry matter accumulation in summer maize for various treatments in 2013 and 2014.HSF,high soil fertility field; LSF,low soil fertility field.CK,control without nitrogen;FP,conventional farmers' practice;IAP,an integrated agronomic practice.V6,6-leaf stage;V12,12-leaf stage;R1, silking stage;R3, milk stage;R6,physiological maturity.

The LAI was affected by soil fertility,cropping practice,and their interaction(Fig.3).In the LSF fields,the LAI values at R1 with IAP in 2013 and 2014 were 44% and 41% greater,respectively, than those with FP. At maturity, the LAI values with IAP in 2013 and 2014 were 63% and 60% greater,respectively, than those with FP. In the HSF fields, the LAI values at R1 with IAP in 2013 and 2014 were 40% and 39%greater,respectively,than those with FP.At R6,the LAI values with IAP in 2013 and 2014 were 78% and 71% greater,respectively, than those with FP.Compared to LSF conditions with IAP, the average post-silking LAI values under HSF conditions with IAP in 2013 and 2014 increased by 16% and 12%, respectively, whereas no significant differences were found in the average pre-silking LAI between LSF and HSF with IAP.

3.3. N accumulation and partitioning between pre- and postsilking

N accumulation was affected by soil fertility, cropping practice, and their interaction (Fig. 4). Among the three treatments, the greatest N accumulation over the entire growing season was with IAP regardless of the year or soil fertility level with a rank order of IAP ＞FP ＞CK similar to that for DM. In the LSF fields, N accumulation amounts at R1with IAP in 2013 and 2014 were 35% and 32% greater, respectively,than with FP. At R6, the total N accumulations with IAP in 2013 and 2014 were respectively 36% and 35% greater than with FP.With IAP,post-silking N accumulation accounted for 33% and 34% of total N accumulation in 2013 and 2014,respectively, and no significant difference in N accumulation was found between IAP and FP. In the HSF fields, N accumulations at R1 with IAP in 2013 and 2014 were respectively 17% and 20% greater than with FP. At R6, the total N accumulations with IAP in 2013 and 2014 were respectively 28% and 29% greater than with FP. With IAP,post-silking N accumulation was greater than that with FP and accounted for 43% and 43% of total N accumulation in 2013 and 2014,respectively.

With IAP, post-silking N accumulation increased by 59%and 48%in 2013 and 2014,respectively,under HSF conditions relative to LSF conditions, whereas no significant differences were observed in pre-silking N accumulation between LSF and HSF with IAP.

3.4. Soil bulk density, Nmin content,and root length

Soil bulk density, Nmincontent, and root length were determined at R1 in 2014, and were affected by soil fertility,cropping practice, and their interaction. In both the LSF and HSF fields, IAP decreased the soil bulk density within the 0-20 cm soil layer, whereas no significant differences were observed in the 20-50 cm layer among all of the treatments(Fig.5).In the LSF fields with IAP,soil bulk densities within the 0-10 and 10-20 cm layers were 1.43 and 1.55 g cm-3,reflecting decreases of 5% and 8%, respectively, compared to the FP treatment. In the HSF fields with IAP, soil bulk densities within the 0-10 and 10-20 cm layers were 1.33 and 1.45 g cm--3, reflecting decreases of 6% and 8%, respectively, compared to the FP treatment. Compared to IAP under LSF conditions,the soil bulk densities within the 0-10 and 10-20 cm layers with IAP under HSF conditions decreased by 8% and 7%,respectively.

Fig.3-Maize leaf area index for various treatments in 2013 and 2014.HSF,high soil fertility field;LSF,low soil fertility field.CK,control without nitrogen; FP,conventional farmers'practice;IAP,an integrated agronomic practice.V6,6-leaf stage;V12,12-leaf stage;R1,silking stage;R3,milk stage;R6, physiological maturity.

Among the treatments, IAP resulted in the highest soil Nmincontent throughout all soil layers(0-50 cm)regardless of soil fertility (Fig. 6). In the LSF fields with IAP, the soil Nmincontents within the 0-10, 10-20, 20-30, 30-40, and 40-50 cm layers increases of 12%,11%,24%,32%,and 33%,respectively,compared to the FP treatment. In the HSF fields with IAP,the soil Nmincontents within the 0-10, 10-20, 20-30, 30-40, and 40-50 cm layers increases of 13%, 14%, 29%, 38%, and 40%,respectively, relative to FP. With IAP, the soil Nmincontents within the 0-10, 10-20, 20-30, 30-40, and 40-50 cm layers increased by 42%,29%,25%,23%,and 23%,respectively,under HSF conditions relative to LSF conditions.

IAP increased root length density within the 20-50 cm soil layer in both the LSF and HSF fields, whereas no significant differences were observed in the 0-20 cm layer among all of the treatments (Fig. 7). In the LSF fields with IAP, root length densities within the 20-30, 30-40, and 40-50 cm layers increases of 48%, 59%, and 63%, respectively, relative to FP.In the HSF fields with IAP,root length densities within the 20-30,30-40,and 40-50 cm layers increases of 44%,53%,and 51%,respectively, relative to FP. With IAP, root length density within the 10-20,20-30,30-40,and 40-50 cm layers increased by 12%,20%,29%,and 28%,respectively,under HSF conditions relative to LSF conditions.

3.5. NUE

The NUE indices PFPN, AEN, and RENwere affected by crop management practice,soil fertility levels,and their interaction(Table 4).IAP produced significantly higher PFPN,AEN,REN,and PENvalues than those obtained with FP across years and soil fertility levels.In the LSF fields,PFPN,AEN,REN,and PENvalues obtained with IAP averaged across the two years were 31.0,14.0,0.39, and 36.2 kg kg-1, which were 29%, 99%, 56%, and 26%greater,respectively,than those with FP.In the HSF fields,PFPN,AEN,REN,and PENvalues with IAP averaged across the two years were 41.5, 18.8, 0.50, and 37.0 kg kg-1, which were 38%, 155%,68%, and 54% greater, respectively, than those with FP. With IAP,PFPN,AEN,and RENvalues in HSF averaged across the two years increased by 34%,34%,and 27%,respectively,under HSF conditions relative to LSF conditions.

Fig.4- Nitrogen accumulation in maize plants for various treatments in 2013 and 2014.HSF,high soil fertility field;LSF,low soil fertility field.CK,control without nitrogen;FP,conventional farmers'practice;IAP,an integrated agronomic practice.V6,6-leaf stage;V12,12-leaf stage;R1,silking stage;R3, milk stage;R6,physiological maturity.

4. Discussion

Fig.5-Soil bulk density in 0-50 cm soil layers at silking for various treatments in 2014.HSF,high soil fertility field;LSF,low soil fertility field;CK,control without nitrogen;FP,conventional farmers'practice;IAP,an integrated agronomic practice.Letters around the error bars indicate significant differences at the 0.05 probability level.

A more integrated approach involving planting density,nutrients, water, and other agronomic management factors could increase crop yield and fertilizer use efficiency [38-41].In this study, we found that the use of IAP increased maize grain yield in both LSF and HSF field,relative to FP.In addition,grain yield with IAP under HSF conditions was greater than that with IAP under LSF condition. The large yield gaps between IAP and FP agronomic management practices and between HSF and LSF conditions suggest considerable potential for increasing summer maize yield, particularly in LSF fields. This view is confirmed by previous studies [40,41], in which the use of IAP increased the grain yields of rice (Oryza sativa L.) and summer maize. The increased grain yield with IAP resulted mainly from greater ear number per hectare regardless of soil fertility and year, and the ear number per hectare with IAP was greater under HSF conditions than under LSF conditions.

Fig.6- Soil Nmin content in 0-50 cm soil layers at silking for various treatments in 2014.HSF,high soil fertility field;LSF, low soil fertility field.CK,control without nitrogen;FP,conventional farmers' practice; IAP,an integrated agronomic practice.Letters around the error bars indicate significant differences at the 0.05 probability level.

Several researchers have reported [19,51] that maize grain yield is associated mainly with dry matter accumulation.Our results indicate that greater DM accumulation both before and after silking could contribute to the yield increase with IAP compared to FP under both soil fertility conditions. In addition, greater post-silking DM accumulation with IAP was observed under HSF (60% of total DM accumulation) than under LSF (54% of total DM accumulation), whereas no significant differences were found in pre-silking (sowing to R1) DM accumulation between LSF and HSF with IAP. This finding indicates that providing sufficient post-silking assimilates is essential to support high sink potential and consequently obtain higher total biomass and grain yield. This inference is consistent with findings of other studies[19,35,36,51] concluding that increased post-silking DM is the key to further increases in yield.In our study,superior source ensured that the high sink potential was used effectively under IAP.The maximum and average post-silking LAI values with IAP were all significantly greater than those for the other treatments,owing to higher planting density and sufficient N supply. Higher planting density increased the LAI to improve maize population light interception and assimilate availability[52-54],while sufficient N supply sustained greater canopy longevity to increase solar radiation interception and carbon accumulation after silking [51,55-57]. Thus, the greater average LAI and DM accumulation post-silking observed with IAP under HSF conditions was due to continued postsilking N accumulation relative to that under LSF conditions,whereas no significant differences were observed during presilking growth.

Fig.7-Root length density in 0-50 cm soil layers at silking for various treatments in 2014.HSF,high soil fertility field;LSF,low soil fertility field.CK,control without nitrogen;FP,conventional farmers' practice; IAP,an integrated agronomic practice.Letters around the error bars indicate significant differences at the 0.05 probability level.

Table 4-Nitrogen use efficiency of summer maize for various treatments in 2013 and 2014.

N uptake by plants was significantly affected by the growth and spatial distribution of roots [30,58,59]. IAP increased the root length density in the 20-50 cm layers and the soil Nminin the 0-50 cm layers at silking compared to FP,resulting in greater total N accumulation with IAP than with FP across the two years and both soil fertility levels. Subsoiling tillage in the IAP treatment decreased the soil bulk density in the 0-20 cm layer to promote root extension and N distribution into deeper soil layers,consequently increasing N uptake by the plants[30,36].In addition,although with IAP the planting density was higher and the N application rate was the same as with FP, the postsilking N accumulation was greater with IAP than with FP under both soil fertility conditions. Split application of N, which can provide adequate N to plants after silking, was the main contributor to the increased post-silking N accumulation[11,15,60]. However, IAP under HSF conditions resulted in greater N accumulation in plants, particularly for post-silking N accumulation (43% of total N accumulation), compared to that under LSF over the two years. This finding could be attributed mainly to excessive permeability and low water and nutrient retention capacities of the less fertile soil in the LSF fields [42,61,62], which resulted in lower soil Nmincontent within the 0-50 cm layers and lower root length density within the 10-50 cm layers under LSF compared to HSF.Because of the increased N uptake with IAP compared to FP,the use of IAP also improved the PFPN, AEN, REN, and PENindices in both the HSF and LSF fields,while PFPN,AEN,and RENvalues with IAP under HSF conditions were greater than those under LSF conditions across the two years.

These results indicate that considerable potential for increasing the yield and NUE of summer maize exists,particularly in LSF fields, and could be exploited by further optimizing the IAP, particularly N fertilizer management to increase post-silking N and DM accumulation. Other aspects of maize production under IAP,such as the economic benefits and environmental sustainability of IAP, should also be examined.

5. Conclusions

IAP is an effective way to increase maize grain yield and NUE simultaneously. In field experiments over two growing seasons, both high yield and NUE were achieved for maize in both LSF and HSF fields using IAP incorporating increased planting density, subsoiling tillage, and split N application in contrast to FP. The large yield gap was attributable mainly to greater root length and Nmincontent with IAP, resulting in increased N accumulation that promoted greater LAI and DM accumulation than with FP. Greater post-silking DM and N accumulation due to increased root length and Nminwere also major contributors to greater grain yield and NUE with IAP under HSF than with IAP under LSF.Owing to the greater grain yield and N uptake,IAP significantly increased NUE compared to FP,particularly in HSF fields.

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2018.12.005.

Conflict of interest

Authors declare that there are no conflicts of interest.

Acknowledgments

This work was supported by the Key National Research and Development Program of China (2016YFD0300207,2016YFD0300103), and the China Agriculture Research System(CRRS-02).