Zhaoxin Liu, Fang Gao, Yan Liu, Jianqun Yang, Xiaoyu Zhen, Xinxin Li, Ying Li, Jihao Zhao,Jinrong Li, Bichang Qian, Dongqing Yang*, Xiangdong Li*

State Key Laboratory of Crop Biology,Shandong Agricultural University,Tai'an 271018,Shandong,China

College of Agronomy,Shandong Agricultural University,Tai'an 271018,Shandong,China

Keywords:Nitrogen management Wheat-peanut relay intercropping system Crop yield Nitrogen recovery efficiency Apparent N loss

ABSTRACT Agronomically optimizing the timing and rates of nitrogen(N)fertilizer application can increase crop yield and decrease N loss to the environment.Wheat(Triticum aestivum L.)-peanut(Arachis hypogaea L.)relay intercropping systems are a mainstay of economic and food security in China.We performed a field experiment to investigate the effects of N fertilizer on N recovery efficiency,crop yield,and N loss rate in wheat-peanut relay intercropping systems in the Huang-Huai-Hai Plain,China during 2015-2017.The N was applied on the day before sowing,the jointing stage(G30)or the booting stage(G40)of winter wheat,and the anthesis stage(R1)of peanut in the following percentage splits:50-50-0-0(N1),35-35-0-30(N2),and 35-0-35-30(N3),using 300 kg N ha-1,with 0 kg N ha-1(N0)as control.15N-labeled(20.14 atom%)urea was used to trace the fate of N in microplots.The yields of wheat and peanut increased by 12.4%and 15.4%under the N2 and N3 treatments,relative to those under the N1 treatment.The15N recovery efficiencies(15NRE)were 64.9%and 58.1%for treatments N2 and N3,significantly greater than that for the N1 treatment(45.3%).The potential N loss rates for the treatments N2 and N3 were 23.7%and 7.0%,significantly lower than that for treatment N1(30.1%).Withholding N supply until the booting stage(N3)did not reduce the wheat grain yield;however,it increased the N content derived from15N-labeled urea in peanuts,promoted the distribution of15N to pods,and ultimately increased pod yields in comparison with those obtained by topdressing N at jointing stage(N2).In comparison with N2,the N uptake and N recovery efficiency(NRE)of N3 was increased by 12.0%and 24.1%,respectively,while the apparent N loss decreased by 16.7%.In conclusion,applying N fertilizer with three splits and delaying topdressing fertilization until G40 of winter wheat increased total grain yields and NRE and reduced N loss.This practice could be an environment-friendly N management strategy for wheat-peanut relay intercropping systems in China.

1.Introduction

Intercropping involves growing two or more crops in the same space at the same time[1].In comparison with monoculture cropping,intercropping can increase crop yields through more effective use of water,nutrients,and solar energy,along with greater land-use efficiency[2].Intercropping with cereals and legumes is implemented in various countries worldwide and is now receiving increased attention as more sustainable agriculture is being developed[3,4].Intercropping improves soil fertility,reduces weeds and insect pests,and increases yields[5].

The North China Plain and the Huang-Huai-Hai Plain are major producers of wheat(Triticum aestivum L.),accounting for approximately 68%of total production in the zone[6].This region also accounts for 60%of the total area sown and the output of peanut(Arachis hypogaea L.),a pillar of a stable vegetable oil market[7].Strategies for increasing the production of both peanut and wheat in the same production year should be developed.The wheat-peanut relay intercropping system represents a crop management strategy that can increase economic efficiency in the region,particularly in Henan and Shandong provinces.Peanut is a legume that employs a symbiotic relationship with Bradyrhizobia bacteria to fix nitrogen into a plant-usable form[8].Cereal/legume mixtures can reduce competition for nitrogen(N).The legume not only improves soil fertility through nitrogen fixation,but also transfers the fixed N to the companion species,promoting the growth and development of the cereal crop[9,10].In this system,wheat is sown in October and harvested in early June the following year,whereas peanut is sown in late May and harvested in September,usually 15-20 days before the sowing of wheat.Wheat and peanut thus coexist in the field for a short time,as the peanut emerges and the wheat matures.By the time the wheat is harvested,the peanut seedlings are already established and continue their growth cycle.In comparison with summer sowing of peanut,the wheat-peanut relay intercropping system allows a longer growth period,and the associated increase in radiation capture over the season tends to increase yields[11].This strategy could also reduce the effects of tomato spotted wilt virus(TSWV)infection,as damage caused by thrips to peanuts is usually reduced if a previous crop or crop residue remains on the soil surface[12].

N fertilizer use contributes to increased yields of major food crops and has supported the rapid expansion of the world population[13].N fertilizer application is expected to increase to produce enough food to feed the world population,which is expected to be 9.3 billion by 2050[14].However,increasing N rates is unlikely to be effective in increasing yields,as N use efficiency declines at high nitrogen levels[15].In addition,imperfect fertilizer management results in inconsistent and inappropriate application of fertilizers in agricultural production,with consequent environmental risks[16].Under the wheat-peanut relay intercropping system,fertilizer requirements of crops are normally estimated based on the response of individual crops to direct application of fertilizers,while the effects of preceding crops and the residual effects of the fertilizer on succeeding crops are largely ignored.Asa result,relatively high,and often uneconomic,doses of fertilizers are recommended to maintain yield.Excessive use of N fertilizers and inappropriate application methods reduce N use efficiency[17],causing environmental problems such as surface and subsurface water contamination[18],the emission of N2O to the atmosphere,a potent greenhouse gas[19],acidification of major croplands[20],and degradation of soil quality[16].It is desirable to seek effective ways to increase food production while minimizing N loss and its consequent environmental damage.

Better management and appropriate use of N fertilizers are a convenient and effective way to improve crop productivity and N use efficiency[15],with minimum environmental risks,and it has been achieved through wheat-maize rotation[21]and rice-oilseed rotation[22]systems.Peng et al.[21]also showed that excess application of N fertilizers can be avoided by considering site-specific conditions that affect residual soil N,such as crop management practices and rotation systems.Split application of N,where a portion is applied before seeding and a portion is applied at later growth stages,contributes to N concentration in the grain,total N accumulation at harvest,or nitrogen use efficiency[23].The combined use of organic and inorganic fertilizers can balance plant nutrient requirements,optimizing soil physical and chemical properties and increasing the availability of soil nutrients[24].

N fertilizer application rates can be adjusted based on the actual crop need by accounting for the amount of N retained by the previous crop.In the wheat-peanut relay intercropping system,N applied to wheat not only increases the wheat yield,but also exerts a residual effect on peanut(the succeeding crop).As peanut can obtain N from rhizobia,all the N can be applied to the wheat to ensure the yield of wheat,when the annual N application is not enough for two crops.N fertilizers can be applied to both crops(wheat and peanut,three times a year)when the total annual N application is sufficient(over 150 kg ha-1);the yearly optimum N ratio of wheat to peanut was 1:0.40-0.55[25,26].Under the same N application rate,to obtain higher yields of intercropped peanut,the optimal amount of fertilizer for wheat was 60%-80%of the total fertilizer applied;topdressing was performed until the jointing to booting stages of wheat and the rest of the fertilizer was applied to peanut before flowering[27].

Previous studies of the wheat-peanut relay intercropping system have focused mainly on yield advantages.However,N use efficiency,the distribution of N among plant parts,and N loss in the crops in this system remain unclear.We hypothesized that split applications of N fertilizer to wheat and peanut at different growth stages would increase grain yield and reduce fertilizer loss.The aim of this study was to quantify the effects of timing and split of N fertilizer on yield,N use efficiency,and apparent N loss to environment in the wheat-peanut relay intercropping system.

2.Materials and methods

2.1.Experimental area

Experiments were performed over two years during 2015-2017 at the State Key Laboratory of Crop Biology and the experimental farm of Shandong Agricultural University,Tai'an,Shandong province,China(36°09′N,117°09′E;128 m elevation).The area has a temperate continental monsoon climate.The mean rainfall amounts during the wheat growth periods were 171.1 mm in 2015-2016 and 202.1 mm in 2016-2017,and those during the peanut growth periods were 470.6 mm and 427.9 mm,respectively(Fig.1).The soil type was sandy loam(Cambisol)[28],and soil pH was 8.25.Soil samples obtained from the plow layer(0-20 cm)before the experiment contained 10.2 g kg-1of organic matter,with total amounts of N 0.9 g kg-1,of rapidly available phosphorus(P)50.3 mg kg-1,and of rapidly available potassium(K)85.4 mg kg-1.

2.2.Experimental design

2.2.1.Field experiment

The experiment was arranged in a randomized block design with three replications.Plot size was 2.5 m×2.5 m,with a concrete wall embedded 2 m into the soil between each plot.Fertilizer treatments consisted of four different splits of N,with a total N application of 300 kg ha-1.The fertilizer was applied on the day before sowing,at G30 or G40 of winter wheat,and at R1 of peanut in the following percentage splits:50-50-0-0(N1),35-35-0-30(N2),and 35-0-35-30(N3),with 0 kg N ha-1(N0)as control.A common compound fertilizer(CCF)manufactured by the Shandong Agricultural University Fertilizer Technology Co.,Ltd.(Feicheng,Shandong province,China)was used in our experiment.The contents of N,P2O5,and K2O in CCF were 20%,15%,and 10%,respectively.The amount of applied CCF was 1500 kg ha-1(converted into pure form,300 kg ha-1of N,225 kg ha-1of P2O5,and 225 kg ha-1of K2O)part of which was manually distributed over the soil surface prior to sowing and then plowed into the soil at a depth of 20 cm as a basal dressing.For top-dressed N,ditching and fertilizing were manually performed at the G30,G40,and R1 stages.The winter wheat cultivar Jimai 22 was grown in the plots in nine rows(0.30 m between rows),and the peanut cultivar 606 was sown manually between the rows of winter wheat at two seeds per hill with 22 cm between hills.Seeding and harvest times are shown in Table 1.Plant densities were kept uniform at 225 and 15 seeds m-2for wheat and peanut,respectively.The water applied at each irrigation was about 60 mm.The same irrigation applications were used for each crop(three for wheat and two for peanut).Disease,weeds,and pests were well controlled in each treatment.

2.2.2.Microplot experiment

Fig.1-Monthly rainfall and mean temperature during the crop growing seasons at the experiment site in 2015-2017.

Table 1-Timing of each operation for wheat and peanut in the 2015-2017 cropping seasons.

To monitor the fate of15N-labeled fertilizer,three microplots(0.4 m×0.5 m×0.5 m;2016-2017)without top and bottom were established within the fertilized plots,with the length perpendicular to the rows at 3-5 cm higher than ground level.15N-labeled urea(46.6%N,20.14 atom%,provided by the Shanghai Chem-Industry Institute)was applied to the microplots in the same doses as applied to the corresponding field trials.The application rates of all other fertilizers(P2O5and K2O)and the application times were the same as those in the field.The number of seeds was calculated according to design in the field and seeds were sown artificially in two rows parallel to the rows in the field.Detailed stages and fractions of15N-labeled fertilizer application are shown in Table 2.

2.3.Plant and soil sampling and measurements

For wheat,plant material in each microplot was sampled close to the ground at maturity,and the plants were then counted for dry matter calculation.Plants were separated into four organ categories:leaves,sheaths+stems,glumes+ear rachises,and grain.For peanut,ten15N-labeled plants were sampled from each microplot and separated into leaf,stem,root,and pod.Plant samples were killed at 105 °C for 30 min and dried to constant weight at 80 °C.Aliquots of each dry sample were ground(to＜0.15 mm)in a ball mill for N content determination and isotopic analysis.Soil samples from 0-20 and 20-40 cm depth were collected using a 4-cm-diameter soil auger at 20-cm intervals at plant maturity.Three sites were selected and then mixed into one sample for each microplot.The soil sample was air-dried and ground to pass a 0.15-mm sieve.15N enrichment was determined in the soil and plant samples using an elemental analyzer(Elementar vario MICRO cube;Elementar Analysensysteme,Hanau,Germany)and a mass spectrograph(Isoprime 100;Isoprime,Cheadle,UK).Total N accumulation was calculated as the product of N concentration and dry weight.The natural15N atom%,which averaged 0.3657 at this site,was subtracted from the atom%measured in each crop fraction;15N uptake was calculated as the product of total N accumulation and the proportion of15N(from labeled urea)in total N.

For wheat harvesting,2.0 m sections were cut from two rows in each plot in both years.Kernel number per spikes was counted in 50 selected spikes.Samples were threshed using a Pint-size Seeding Threshing Machine(Zhengzhou ZiKai Machinery Co.Ltd.,Zhengzhou,China).Grain was air-dried,weighed,and standardized to 12%moisture content.Three samples were weighed to determine the average 1000-kernel weight for each plot.

During peanut harvest,2.5 m of two rows was demarcated in each plot,and all the peanut plants in the quadrat were dug out for yield measurement;10 representative plants were sampled from each quadrat to record numbers of pods per plant.All pods were collected from the peanut plants and air-dried,weighed,and adjusted to a standard 6%water content.Pods were shelled to obtain kernel yield,kernels per kilogram,and shelling percentage.

2.4.Calculations

N uptake,N apparent recovery efficiency(ARE),and N harvest index(NHI)were calculated following[29,30]:

The percentages of N derived from basal N,topdressing N,and residual N in soil were calculated following Chen et al.[31].

2.5.Statistical analyses

All data were analyzed using LSD tests with the DPS v 7.05 statistical software package.Differences between treatments were tested by the least significant difference test at the 0.05 probability level.Plots were drawn with SigmaPlot 10.0 program.The patterns for the data from two years were consistent,and the data in 2017 were used in the present study.

3.Results

3.1.Crop yield and yield components

In both growing seasons,N treatment significantly(P＜0.001)affected the grain yield of wheat(Table 3).Grain yields with all N fertilizer treatments were significantly increased by 36.9%-47.8%and 36.0%-41.8%,respectively,compared with those in N0.However,in comparison with N1,grain yields of N2 and N3 did not decrease despite a 30%reduction in fertilizer use.The highest yield,which averaged 8283.8 kg ha-1,was observed in N3 in both years.There was no significant difference between the grain yields in N2 and N3.Yield components also differed among treatments.The effects of N treatment and year on 1000-kernel weight were significant(P＜0.001).Spike number,kernel number per spike,and 1000-kernel weight in all N fertilizer treatments were higher than those in N0,especially kernel number per spike and 1000-kernel weight.In 2016,kernel number per spike and 1000-kernel weight were increased by 4.5%and 3.7%,respectively,in N2,and by 11.2%and 2.3%,respectively,in N3,compared to those in N1.In 2017,the 1000-kernel weights in N2 and N3 increased by 3.2%and 3.7%,respectively,compared to those in N1.In addition,the interaction of N treatment×year had a significant effect on the grain yield(P＜0.001).

Table 3-Wheat yield and yield components with different N split applications.

Table 4-Peanut yield and yield components with different N split applications.

Table 5-Analysis of variance of N uptake,N harvest index,and N apparent recovery efficiency of wheat and peanut as affected by N treatment and year.

The effects of N treatment and year on pod yield and kernel yield in peanut were significant(Table4).The application of N fertilizer significantly increased the pod and kernel yields of peanut,compared to those yield components in N0.Pod yields in N2 and N3 were higher than those in N1 by 18.1%and 25.3%in 2016,and by 34.7%and 42.2%in 2017.The effects of N treatment and year were significant for pods per kg and kernels per kg,whereas only the effect of N treatment was significant for pods per plant and shelling percentage(Table 4).There was no significant difference in pods per kg and kernels per kg in N2 and N3,but pod yields in N3 were 6.1%and 5.6%higher than those in N2 in 2016 and 2017,respectively.Pod and kernel yield in 2017 were lower than those in 2016.These results show that splitting N application could significantly increase peanut yield and that G40 is the optimum stage for topdressing the fertilizer.

3.2.N uptake,NHI,and ARE for wheat and peanut

Both N treatment and year showed significant effects on total N uptake of wheat and peanut,but neither crop showed a significant interaction of N treatment×year(Table 5).The total N uptakes of wheat and peanut in all N-fertilized treatments were significantly increased by 32.7%-35.1%and 56.1%-121.9%,compared to those in N0 over the two years.For wheat,total N varied from 178.2 to 240.6 kg ha-1in 2016 andfrom 195.1 to 263.8 kg ha-1in 2017.However,there were no significant differences among the N1,N2,and N3 treatments in 2016 and between the N1 and N3 treatments in 2017.For peanut,total N uptake ranged from 83.8 to 181.0 kg ha-1in 2016 and from 76.0 to 173.6 kg ha-1in 2017,and there were significant differences among all the treatments(Table 6).Averaged for the N treatment,the total N uptake was 9.5%higher in wheat and 6.8%lower in peanut in 2017 than in 2016.

Table 6-Effects of N split applications on N uptake,N harvest index and apparent recovery efficiency of wheat and peanut.

Fig.2-Regressions of N apparent recovery efficiency(ARE)on N uptake for wheat(left)and peanut(right)with three replicates.**Significant at the 0.01 probability level.

Different levels of N treatment had a significant effect on the NHI of both wheat and peanut,whereas only the effect of year was significant for peanut.However,neither crop showed a significant interaction of N treatment×year(Table 5).NHI showed the same pattern as total N uptake.For wheat,NHI varied from 67.7%to 73.1%in 2016 and from 65.8%to 72.3%in 2017;there were no significant differences between the N2 and N3 treatments in 2016 or between the N1 and N2 treatments in 2017.For peanut,total N uptake by aerial plant parts ranged from 65.5%to 75.3%in 2016 and from 67.2%to 76.1%in 2017,and there were significant differences among all the treatments(Table 6).

Significant(P＜0.001)effects of year and N treatment on N apparent recovery efficiency(ARE)of wheat were observed,whereas only the effect of N treatment on the ARE of peanut was significant(Table 5).Positive correlations were observed between ARE and total N uptake(Fig.2),both in wheat and in peanut.ARE ranged from 18.6%to 32.3%for wheat and 14.8%to 32.5%for peanut in all treatments.The ARE increased as the time of N fertilizer application was postponed.Compared to the ARE in N1,those in N2 and N3 were significantly increased by 56.6%and 68.0%for wheat and by 75.8%and 117.8%for peanut in the two years.Furthermore,there were significant differences among all treatments.The ARE obtained in 2017 was 7.6%higher in wheat than that in 2016,whereas no significant difference was observed in ARE in peanut(Table 6).

3.3.Proportions of crop N derived from fertilizer(Ndff)and15N distribution in wheat and peanut

Approximately 34.6%-39.8%and 35.4%-61.9%of plant N uptake were derived from15N-labeled fertilizers in wheat and peanut,respectively(Table 7).Compared to N1,the Ndff in wheat under N2 and N3 was decreased by 8.0 and 5.9 kg ha-1(8.6%and 6.3%decrease);however,there was no significant difference between the N2 and N3 treatments.In contrast,the Ndff in peanut under N2 and N3 was increased by 46.6 and 60.8 kg ha-1(109.1%and 142.4%increases).The annual total Ndff values of wheat and peanut for treatments N2 and N3 were significantly higher thanthat for treatment N1,by28.4%and 43.3%.Among all treatments in wheat,Ndff from topdressing15N-labeled fertilizer was higher than that from basal15N-labeled fertilizer.In peanut,in the N2 and N3 treatments,N uptake from the second topdressing N(topdressing at R1)was approximately twice that from the first topdressing N(topdressing at G30 and G40).The topdressing N(topdressing at G40)in N3 treatments was significantly higher than that of the topdressing N(topdressing at G30)in N2 treatments,although the same15N application rate was used at both stages.

Table 7-Wheat and peanut plant N derived from15N-labeled urea(Ndff)during 2016-2017.

Fig.3-Effects of N split applications(basal and topdressing)on distribution of15N among wheat and peanut organs during 2016-2017.The different letters on bars refer to significant differences at the 0.05 probability level.B denotes basal N;G30,G40,and R1denote topdressing N at the beginning of the jointing stage of wheat,the booting stage of wheat,and the anthesis stage of peanut,respectively.

The distribution of15N among the different organs was ordered as follows:grain＞stem＞spike axis and glumes＞leaves in wheat(Fig.3-A),and pod＞stem＞leaves＞root in peanut(Fig.3-B).In wheat and peanut,63.4%-75.1%and 76.0%-81.7%,of total15N uptake by the plant was partitioned to wheat grain and peanut pod.Among all treatments in wheat,the proportion of15N partitioned to grain from topdressing N(topdressing at G30 and G40)was 72.3%-75.1%higher than that partitioned to grain from basal N.For peanut,the proportion of15N partitioned to pod from the second topdressing N(topdressing at R1)was significantly higher than that partitioned from the first topdressing N(topdressing at G30 and G40).The opposite trend was observed for other nutritive organs,indicating that basal N increased early N accumulation in straw,whereas topdressing N increased late N accumulation in grain.Further,applying the fertilizer at G40 significantly increased N accumulation in the pod,compared to application at G30.NHI was strongly and positively(P＜0.01)associated with15N distribution in the wheat grain and peanut pod.However,NHI showed a negative association with15N distribution in the stem(Fig.4).

3.4.Fate of15N-urea in the wheat-peanut cropping system

For all treatments,the mean fertilizer-15N recovery efficiency(15NRE)and residual and apparent loss rates were 56.1%,23.1%,and 20.3%,respectively,for the annual relay intercropping system(Table 8).The15NRE values of wheat for the treatments N2 and N3 were 40.5%and 41.6%,significantly higher than that for the N1 treatment(31.1%).However,there was no significant difference between the N2 and N3 treatments.The15NRE for the treatments N1,N2,and N3 was 14.2%,29.8%,and 35.8%in peanut,and 45.3%,58.1%,and 64.9%,respectively,in the annual relay intercropping system,with significant differences among all treatments.The annual apparent loss rates were lower for N2 and N3(23.7%and 7.0%),in comparison with N1(30.1%).However,no differences were observed among N1 and N2 for residual N in the 0-40-cm soil layer in the annual relay intercropping system.

Fig.4-Regressions of N harvest index(NHI)on15N distribution in stem and grain of wheat and peanut with three replicates.**Significance at the 0.01 probability level.

The15NRE of topdressing N at G30 and G40 was significantly higher than that of basal N,for all treatments in wheat.In peanut,at the same N rate,the15NRE of topdressing N at G40(32.7%)in N3 was higher than that of topdressing at G30(22.6%)in N2.Topdressing N at R1 was significantly higher than that of topdressing N at G30 in N2 and topdressing N at G40 in N3.The opposite trend was observed for apparent loss rates(Table 8).The15NRE and residual and apparent loss rates were 58.1%,18.2%,and 23.7%,respectively,in N2,and 64.9%,28.1%,and 7.0%,respectively,in N3.The apparent loss rates in N3 were the lowest among all treatments,because of the high15NRE and residual N in the soil.Thus,N3 was the optimal treatment for the wheat-peanut relay intercropping system.

4.Discussion

4.1.Wheat and peanut yield

Dividing N fertilizer application into basal and topdressing applications can increase grain yields and NRE[32].In our experiment,under the wheat-peanut relay intercropping system,N fertilizer was applied to both crops as three treatments(base and topdressing to wheat and topdressing to peanut,N2 and N3).This practice significantly increased wheat and peanutyields,compared to a treatment in which the total annual N fertilizer was applied to wheat in two treatments(basal and topdressing to wheat,N1).These results are consistent with previous studies by Shi et al.[33],who found that higher grain yields and increased NRE in winter wheat when N fertilizer application was divided into an appropriate ratio of basal and top dressed N.Wang et al.[34]and Zhang et al.[35]also reported that wheat and peanut yields were affected by the proportion of N provided to wheat and peanut and fertilization time under the wheat-peanut planting system.

Table 8-Fate of15N-urea in the wheat and peanut cropping system during 2016-2017.

Sustainable crop production depends on the continuous renewal of soil fertility through a balance between N demand and supply in cropping systems.Nitrogen(N)is the most essential of all fertilizers for crop growth,productivity,and grain quality[36].Efficient nitrogen fertilizer management is essential for achieving economic yield and enhancing N use efficiency[37].In the present study,at the same N fertilizer rates,the peanut pod yield in N3(topdressing until G40)was higher by 5.9%than that in N2(topdressing atG30)(Table 4);further,N3did not reduce wheat yield(Table 3).As a result,N3 resulted in a higher total crop yield of wheat and peanut than N2.This finding is consistent with those of Li et al.[27],who found that delaying the topdressing fertilization time of wheat could boost crop productivity in the wheat-peanut relay intercropping system.Thus,the nutrients released under the N3 treatment met the requirements ofthe crops,in turn promoting the transport of dry matter from vegetative organs to grain and pod during yield formation and thereby increasing yield.

The finding that grain yields in both crops varied when the N application treatment was varied in the two years demonstrates the significance of N for increasing crop productivity,in agreement with previous reports[32,38]on wheat and peanut.Differences in yield were mainly the consequent of different N management strategies associated with soil fertility and the N uptake ratio by the aerial plant parts(Tables 6,7).This finding is consistent with those of Shi et al.[39]and Zhang et al.[35],who reported that peanut yield components were affected by the N rate and the ratio of base fertilizer to topdressing,and that crop yields are usually more dependent on fertilizer levels and fertilization time.Previous study[40]has also shown that the N absorbed by peanut is derived not only from N fertilizer and soil,but also from symbiotic N2fixation in the root nodules.A close correlation between N uptake and crop yield has also been documented[41].

4.2.Effects of splitting N applications on N derived from15N-labeled urea(Ndff)and distribution in organs

Providing the crop with labeled N at different times enables us to monitor the different sources of plant N uptake and the dynamic change of N within the plant.Our study showed that the amount of15N derived from basal N was lower than that from topdressing15N in wheat.It indicates that although basal N is essential for crop vegetative growth at the early growth stage,timely topdressing at the middle or later growth stages is important for obtaining high grain yields.This finding is consistent with previous study by Shi et al.[32].Our results also revealed an increase in peanut plant uptake of fertilizerderived N with the delay of fertilization time,and significant differences in Ndff among all treatments.15N from the topdressing N at G40(N3)was higher than that from the topdressing N at G30(N2),despite the application of the same amount of labeled N in the N2 and N3 treatments(Table 7).These results are similar to those of Li et al.[27],who found that providing topdressing fertilizer until the flag leaf stage of wheat could increase the portions of peanut N derived from the fertilizer in the wheat-peanut relay intercropping system.This result also demonstrates the importance of appropriate timing of N for promoting the absorption of fertilizer N and achieving higher N recovery efficiency(Table 8).

N application time affected the distribution of15N in wheat and peanut.Andersson et al.[42]found that a greater proportion of N uptake was allocated to developing grain in spring wheat,and similar results have been reported by Subedi and Ma[43]in maize.In our study,of the total15N taken up by each plant,63.4%-75.1%was partitioned to the grain,followed by the stem,spike axis and glumes,and leaves in wheat,and topdressing-derived15N was more prevalent in grain than basal-derived15N.These results are similar to those of Shi et al.[32],who found that the quantity of15N derived from topdressing N in grain was higher than that in straw,whereas15N derived from basal N in grain was lower than that in straw.In peanut 76.0%-81.7%of total15N uptake by the plant was partitioned to the pod,followed by stem,leaves,and root.Splitting N application and postponing fertilization time increased the proportion of15N to total15N uptake in wheat grain and peanut pod,whereas the opposite trend was observed in straw(Fig.3).This finding is consistent with previous study by Yang et al.[44],who found that there was more cumulative N in grain from the topdressing15N than from the basal-15N in maize-wheat-maize rotations.Zhao et al.[45]showed that a greater proportion of N uptake was allocated to the developing grain when supplementary N was provided during the late growth stage and that withholding N supply until the flag-leaf stage increased15N in oat kernels.NHI is an important index of nitrogen utilization efficiency that is closely related to harvest organ yield[46].In the present study,NHI increased with N fertilizer postponing and the time of split applications,indicating that more15N of the total15N uptake by the plant was partitioned to wheat grain and peanut pod(Fig.4).

4.3.N application timing and the fate of applied nitrogen

Better N management strategies are needed to optimize crop growth while protecting the environment.Normally,N fertilization will enhance grain yield and increase profits.However,overuse of N fertilizer can result in low N use efficiency,causing further negative environmental effects[15,20].Proper timing of N application and adequate N supply are critical for meeting plant needs and increasing N uptake and overall NRE[47].In the present study,the NRE in treatments N2(58.1%annual)and N3(64.9%annual)were significantly higher than that in treatment N1(45.3%annual).Topdressing N at G40 resulted in a higher NRE than that at G30,being 50.8%and 37.8%in wheat and 32.7%and 22.6%in peanut despite the same amount of topdressing.In the relay intercropping system,topdressing N at a later growth stage does not reduce the NRE of wheat,but improves NRE of peanut,The observation that the annual NRE in treatment N3(64.9%)was higher than that in treatment N2(58.1%),indicate that the timing of N application strongly affected NRE in peanut,in agreement with the results of Wang et al.[34].

Apparent N losses are closely related to the application technique and N rate[48].Luce et al.[49]noted the importance of adjusting N fertilizer rates based on preceding crops to minimize the potential for N loss in canola and wheat cropping systems.In the present study,the full amount of N fertilizer was applied to wheat in N1,whereas in treatments N2 and N3 only 70%of the total N fertilizer was applied to wheat and 30%of it was applied to peanut.The apparent N loss rates in N2 and N3 were 23.7%and 7.0%,respectively,significantly lower than that in N1(30.1%,Table 8).Interestingly,in both wheat and peanut,lower apparent N loss was observed with topdressing N than with basal N,indicating that delaying N application until the later growth stage or applying N with multiple split applications is a promising strategy for improving the synchrony between soil N availability and crop N demand.This is consistent with results of Shi et al.[32]with winter wheat under different N splits.

Topdressing N at G40 produced higher NRE in peanut than topdressing N at G30,whereas the opposite trend was observed for the apparent N loss rate(Table 8).The explanation may be that the fertilizer applied on wheat during G40 provided nutrients not only for wheat,but also for the early growth stage of peanut during the overlapping growth periods of wheat and peanut.Thus,it effectively served as a basal fertilizer for peanut[50],enhancing the synchronization between N supply and crop demand and reducing the risk of soil N loss.

5.Conclusions

Economic management of fertilizer application is essential not only for improving crop productivity,but also for maximizing the N use efficiency of crops,ensuring the environmental sustainability of China's agricultural production.Different timings and splits of fertilizer application affected grain yield,total N uptake,NHI,and ARE.Split N applications increased plant N uptake derived from fertilizer by 9.5%-14.5%compared with N1.Three splits in N application(base and topdressing to wheat and topdressing to peanut)and withholding N supply until G40 did not affect wheat grain yield,relative to topdressing N at G30.However,the peanut pod yield and the NRE of both wheat and peanut were significantly increased,while apparent N loss in to the environment decreased.In terms of yield,NRE,and apparent loss rates,the N3 treatment was the most beneficial for wheat-peanut relay intercropping in the study region.The findings of the present study can be used to develop optimal N management strategies for wheat-peanut relay intercropping systems in the Huang-Huai-Hai Plain,China.

Acknowledgments

This research was supported by the National Key Technology R&DP rogram of China(2014BAD11B04-2),the National Natural Science Foundation of China(30840056,31171496)and Shandong Modern Agricultural Technology and Industry System(SDAIT-04-01).