Modified split application of nitrogen with biochar improved grain yield and nitrogen use efficiency in rainfed maize grown in Vertisols of India

Conventionally, non-judicious and blanket fertilizer nitrogen (N) used in rainfed maize lead to higher N losses, low N use efficiency (NUEs) and poor yields due to substandard agronomic management practices. To avoid such N losses, fertilizer additions are synchronized with plant uptake requirements. In this context, agronomic based management focused on optimizing N rates and biochar application is essential for improved NUEs and crop productivity. Keeping this in view, a field experiment was conducted during 2014, 2015 and 2016 in rainfed maize (Zea mays L.) grown in Vertisols of India. In this study, twelve treatments that comprised of N omission plot (N0), skipping of basal rate, multi-split topdressing at varying time as broadcast and band placement, soil test crop response (STCR) based NPK with target yield 6.0 t ha-1 in maize and biochar application (10 t ha−1) were investigated. The experiment was conducted following a Randomized Complete Block Design (RCBD) set up with three replications. Pooled analysis of three years data revealed that the application of N rates (120 kg Nha−1) in 2 equal splits (60 kg Nha−1) at knee high (V8) and tasseling (VT) stages with skipped basal N rate, achieved higher maize grain yield (5.29 t ha−1) ascribed to the greater growth parameters, yield components and N uptake compared to the recommended practices. Biochar application (10 t ha−1) as soil amendments along with multi top dressed N (120 kg N ha−1) into 3 splits also increased the grain yield. Delayed N application at V8 and VT growth stages, resulted in higher N uptake, agronomy efficiency (AE), partial factor productivity (PFP), physiology efficiency (PE) and recovery efficiency (RE). Biochar along with N fertilizer also improved the soil organic carbon (5.47g kg−1), ammonium-N (2.40 mg kg−1) and nitrate-N (0.52 mg kg−1) concentration in soil (P<0.05) as compared to non-biochar treatments. Application of biochar along with chemical fertilizer (120 kg Nha−1) significantly increased the concentration of ammonium (2.40 mg kg−1) and nitrate (0.52 mg kg−1) in soil (P<0.05) as compared to non-biochar treatments. The perfect positive linear relationship illustrated that the grain yield of rainfed maize was highly dependent (R2=0.99 at p<0.0001) on N availability, as indicated by the fitted regression line of maize grain yield on N uptake. On the other hand, factor analysis revealed, the one to one positive function relationship of biomass with N uptake at V8 and VT growth stages. Principal Component Regression (PCR) analysis showed that PC1 acted as a major predictor variable for total dry matter yield (TDMY) and dominated by LAI and N uptake. Consequently, these results expressed that the agronomic management based multi-top dressed N application and biochar application to achieve higher yield and greater NUEs in rainfed maize is strongly linked with N application into splits.

multi top dressed N (120 kg N ha -1 ) into 3 splits also increased the grain yield. Delayed N application at

27
V8 and VT growth stages, resulted in higher N uptake, agronomy efficiency (AE), partial factor 28 productivity (PFP), physiology efficiency (PE) and recovery efficiency (RE). Biochar along with N 29 fertilizer also improved the soil organic carbon (5.47g kg -1 ), ammonium-N (2.40 mg kg -1 ) and nitrate-N 30 (0.52 mg kg -1 ) concentration in soil (P<0.05) as compared to non-biochar treatments. Application of 31 biochar along with chemical fertilizer (120 kg Nha -1 ) significantly increased the concentration 32 of ammonium (2.40 mg kg -1 ) and nitrate (0.52 mg kg -1 ) in soil (P<0.05) as compared to non-biochar 33 treatments. The perfect positive linear relationship illustrated that the grain yield of rainfed maize was 34 highly dependent (R 2 =0.99 at p<0.0001) on N availability, as indicated by the fitted regression line of 35 maize grain yield on N uptake. On the other hand, factor analysis revealed, the one to one positive 36 function relationship of biomass with N uptake at V8 and VT growth stages. Principal Component

37
Regression (PCR) analysis showed that PC1 acted as a major predictor variable for total dry matter yield 38 (TDMY) and dominated by LAI and N uptake. Consequently, these results expressed that the agronomic 39 management based multi-top dressed N application and biochar application to achieve higher yield and Introduction 52 Globally maize (Zea mays L.) is an important food crop with highest productivity compared to any other 53 food crops [1]. Since, maize is a heavy nutrient exhaustive crop [2] and being C4 plant type, requires a 54 regulated and assured supply of nutrient particularly nitrogen (N) fertilizer throughout its growing period 55 [3]. Nitrogen fertilization assured centerstage for maize production [4] but N fertilizer is a costly input 56 and it is very complicated to manage because its utilization is largely dependent on agronomic, genetic, 57 biological, soil and climate factors [5]. Nitrogen fertilization and time of its application is most crucial 58 for higher productivity of maize [6]. However, non-judicious and excess use of chemical N fertilizer to 59 cropland may have negative impact on air, water, soil and biodiversity and also generate green house 60 gases [7]. Besides, conventional practice of N application (large portion) through surface broadcasting 61 just after maize planting instead of split application have led to the decrease in the crop production and 62 low N use efficiency in India [8] because the N is applied to the soil is vulnerable to losses from the soil-

63
plant system passing through volatilization, de-nitrification, leaching, run-off causing serious threat to 64 environmental quality [9]. Thus, full supply/application of N at the time of planting has resulted in poor 76 temperature and rainfall were also strong relationship with crop yield and N uptake [20]. Skipping the 77 basal rate of N and sifting it at crop establishment stage enhanced the grain yield and agronomic use 78 efficiency at 130 kg N ha -1 application in rice [21]. Maize produced higher growth and yield component 79 with multi-split N application at later growth stage and also had a positive impact on maize yields [22].

80
With the application of N at V6 stage, plant exhibited N deficiency resulting in decreased cell division 81 and cell elongation, which decreased leaf length and delayed time for leaf expansion and consequently 82 decreased grain yield [23]. In general, it was reported that the benefit of late split N applications is 83 mainly depended on quantity and application pattern and there were no evidence of decline in yield level 84 with delayed N application in maize until V11 [24].

85
Most of previous studies on multi-split N application for improving crop yield and N use efficiencies in 86 rainfed maize are limited and also N recommendations for maize are less accurate than desired. The heavy clayey in texture (24.5% sand, 23.5% silt and 52.0% clay) and slightly alkaline in reaction 98 (pH=7.8). The soil was low in organic carbon (4.5 g kg -1 ), low in available KMnO4-N (206.6 kg ha -1 ), 99 but high in available Olsen' P (50.0 kg ha -1 ) and high in NH4OAc-K status (621.0 kg ha -1 ) ( Crop culture and management 159 The maize was planted during kharif or wet/rainy season on 10 July in 2014, 19 July in 2015 and 26 June 160 in 2016 as rainfed crop. Before planting of the maize, field was leveled by tractor drawn laser leveler 161 subsequently; two cross ploughings with disc plough and one ploughing with cultivator followed by 162 planking were done for land leveling. Maize seed (Pro-agro 4212-95-110 day's duration) was treated 163 with imidacloprid @ 2g kg -1 seed, to avoid the insect attack and the seed (20 kg ha -1 ) was planted at uniformly at 12% moisture of three years data. Stover/ straw yield was obtained as difference between 192 total dry matter yield and grain yield. The values were finally expressed in terms of t ha -1 .

Plant analysis for N accumulation and N use efficiency (NUEs)
194 Dry matter of whole plant, grain and stover samples of maize from each treatment were ground into 195 powder by a grinding machine and 0.5 g dried powder were taken for digestion using Kjeldhal method.

196
AR grade concentrated sulfuric acid was used for digestion of dried samples at temperature between 360 197 and 410 ºC. The rate of digestion was accelerated by using copper sulphate as catalyst, and anhydrous 198 sodium sulphate to raise the boiling temperature of sulfuric acid (H 2 SO 4 ). After completion of digestion, 199 the samples were cooled and diluted concentrated alkali (40% NaOH) was added to H 2 SO 4 digest for 200 distillation. The distilled ammonia was quantitatively absorbed in boric acid and titrated against standard 201 acid (0.1 N H 2 SO 4 ) and the N content was calculated. Accordingly, N uptake pattern at different growth 202 stages by maize was calculated by multiplying nutrient concentration with respective dry matter, 203 accumulations/absorption. The agronomic efficiency (AE), partial factor productivity (PFP),

204
physiological efficiency (PE) and recovery efficiency (RE) were calculated by the following formulas: 208 PFP = GY/FN (7) 209 where GY0 and GY represent the grain yield in the N0 plot and fertilized N plots, respectively; and FN is 210 the quantity of N fertilizer applied in N fertilized plot; N0 up and N up are the total nitrogen uptake in 211 above ground biomass in the N0 plot and fertilized N plots, respectively.

213
Composite soil samples were collected from each plot using a core sampler from 0 to 15 cm depth after

228
Linear bivariate regression analysis was done between maize grain yield (as a dependent variable) and 229 total N uptake (as an independent variable) using PROC REG. Factor analysis was conducted using 230 PROC FACTOR to understand the factors that influence nitrogen uptake as well as to assess patterns or 231 correlations among biomass and N accumulation variables. Factor analysis was conducted for two

Biomass and nitrogen (N) accumulation
286 Combined analysis of the three years data indicated that the dry matter (g plant -1 ) significantly varied 287 with different treatments, different growth stages viz., at the knee high (V8), tasseling (VT), silking (R1) 288 and physiological maturity (PM) (Fig 5). The differences among the treatments were significant at 289 different growth stages in all the three years, and highest biomass was observed with STCR based N 290 application when compared to the remaining treatments. Among the different time and rates, the highest 291 biomass accumulation was observed when, basal N rate was skipped and total amount of N was applied 292 into two equal split (T5) at V8 and VT stages. However, the highest biomass production was recorded at 293 PM stage. Application of biochar (10 t ha -1 ) with full amount of N fertilizer (T3) into 3 unequal splits 294 were also significantly improved the biomass accumulation at V8, VT, R1 and PM stages of maize. As a 295 similar fashion of N accumulation in dry matter, the N accumulation was the highest at PM followed by 296 the R1, VT, and the lowest was in the V8 of maize growth stages (Figs 6 and 7). The highest N 297 accumulation in biomass was recorded with the STCR based fertilization (T10). However, among the 298 varying rates and time of application, the highest N accumulation in biomass was observed with the 299 application of total amount of N (120 kg N ha -1 ) applied into 2 equal splits at V8 and VT growth stages 300 while basal N rate was omitted (T5) as compared to other treatments. Remaining treatments were 301 comparable and showed non-significant differences at all growth stage in respect to N accumulation in 302 whole plant biomass during the experimentation.

303
Maize yields and yield components 304 Pooled results indicated that the spliting of N rates had significant effect on yield components like cob 305 length, cob girth, 100-kernel weight and number of grains cobs -1 during all three years (

315
Biochar application also improved the yield components as compared to no application of biochar. Lower 316 rate of N (90 Nha -1 ) had no significant effect on yield components when it was applied into either 2 or 3 317 equal splits. Among the different yield components, the differences in the 100-grain weight were not 318 significant among the applied treatments. The highest number of grains cob -1 for the T5 and T11 was 319 mainly due to the cob length and cob girth from 17.8 to 19.2 cm which was greater than the other 320 treatments. Grain yield of maize varied in all three years with the applied treatments and yield reduction 321 was noticed with N control (N0) and absolute control (N0P0K0) treatments relative to all other treatments 322 (Table 6).   (Table 6).

334
Furthermore, the grain yield was also significantly increased with the application of biochar (10 t ha -1 ) 335 along with recommended dose of N into 3 splits as compared to no biochar application in all three years.

336
STCR based NPK application recorded the highest total dry matter yield (TDMY) during all three years.

337
Among the split application of N, the highest total biomass yield was obtained with the application of N application, grain yield (5.29 t ha -1 ) and TDMY (11.62 t ha -1 ) were significantly highest in the treatment 344 where basal rate of N was skipped and total N (120 kg Nha -1 ) was applied in 2 equal splits (60 kg Nha -1 ) 345 at V8 and VT, respectively (T5). This treatment (T5) increased grain yield and TDMY by 13.5% and 346 11.2% over recommended doses/rates, respectively. However, application of biochar along the 347 recommended rate of N (120 kg Nha -1 ) was also significantly increased the grain yield and total biomass 348 yield as compared to other treatments. Furthermore, the differences in grain yield and total biomass yield 349 between the T5, T10 and T11 were not statistically significant.

Nitrogen uptake and N use efficiencies 351
The total amount of N uptake in maize (grain+ straw) differed with the treatments. Combined result of N 352 uptake was the highest with the STCR, and the lowest was in the N control (N0). However, among the 353 split applications of N rates, the higher amount of N uptake was observed with the application of N rate 354 (120 kg Nha -1 ) in 2 equal splits at V8 and VT with omission of basal N (T5) followed by the application 355 of biochar (10 t ha -1 ) with recommended split as broadcast (T11). Evidently, grain yield of maize was 356 highly dependent (R 2 =0.99 at p<0.0001) on N availability, as indicated by the fitted regression line of 357 maize grain yield on N uptake (Fig 8.). The perfect positive linear relationship further illustrated that the 358 change in grain yield is significantly associated with the changes in N uptake. Hence, the fact that, 359 maximum N uptake was possible with the timely N application to the crop for higher productivity. Band

375
As a similar trend of AE, the application of N rate (120 kg Nha -1 ) in 2 equal splits at V8 and VT noticed 376 the significant maximum PFP (49.7 kg kg -1 N) as compared to all treatments except T4 and T12.

377
Furthermore, among the all treatments, STCR based N rate always exhibited the lowest PFP (31.9 kg kg -1 378 N). Band placement of N (90 kg Nha -1 ) along with biochar (10 t ha -1 ) also significantly increased the 379 PFP in maize as compared to broadcast application (Table 7). Whereas, the difference in PE between 380 different treatments were not significant, however, the application of N (120 kg N ha -1 ) into 2 or 3 split 381 application significantly increased the PE as compared to N application equivalent to 90 kg N ha -1 . The

382
T5 recorded the highest RE (48.0%) in maize, which was significantly higher than other treatments, 383 however application of biochar along with recommended split application of N (T11) was found non-

Principal component regression (PCR) relationship to total dry matter yield (TDMY)
415 Principal component analysis (PCA) was used to identify the factors influencing total dry matter of maize 416 yield and has been used as the preliminary step in the development of a prediction model for TDMY as 417 the Pearson correlation analysis revealed the result that the independent variables were significantly 418 correlated to each other (Table 8) i.e. the multicollinearity was present among the independent variables.

426
To resolve the multicollinearity problem, principal component regression model was fitted for total dry 427 matter yield (TDMY) with respect to the independent variables viz., plant height (PH); LAI, DMA, cob 428 length (CobL), cob girth (CobG), 100-grain weight, grains cob -1 (grainC), N content in dry matter 429 (NDM), N uptake in dry matter (NUDM). In the analysis, 3 principal components were selected because 430 these components explained the cumulative variation of 99.22% in independent variables data set (Fig   431  10). PCA as illustrated from the PC loading coefficient of independent variables revealed that first PC 432 (PC1) contributed 92.98 % explained variation (Fig 10) and dominated by grains cob -1 , LAI and NUDM 433 (Table 9).

Soil Chemical properties 449
There was an improvement in the soil chemical properties under the different agronomic management 450 practices at the end of the study (Table 11). Application of N along with biochar significantly increased 451 the soil organic carbon (SOC), concentration of ammonium (NH 4 + -N) and nitrate (NO 3 --N) in soil as 452 compared to recommended agronomic practices. A perusal of data showed that the highest SOC (5.47 g 453 kg -1 ) was recorded with application of N along with biochar at 10 t ha -1 (T11), while lowest SOC was 454 noticed under the control treatment.

459
Further results showed that, there were no significant differences among the different split N application 460 in relation to SOC at the end of crop season however; application of biochar improved the SOC (5.47g 461 kg -1 ) level as compared non-biochar treatments. Application of biochar along with chemical fertilizer 462 (120 kg Nha -1 ) significantly increased the concentration of ammonium (2.40 mg kg -1 ) and nitrate (0.52 463 mg kg -1 ) in soil (P<0.05) as compared to non-biochar treatments. Since, application of biochar causes 464 amelioration effects, which results into improvement soil fertility. Furthermore, results showed that the 465 application of N rate (120 kg Nha -1 ) in 2 equal splits at V8 and VT growth stages also recorded higher values of NH 4 + -N (2.29 mg kg -1 ) and NO 3 --N (0.44 mg kg -1 ) concentration as compared recommended noticed that N availability in biomass indices might be able to explain the variation in maize yield and N 572 uptake. Moreover, the study showed the positive and significant linear association of DMA (biomass)

573
with N uptake at V8 and VT growth stages of maize. In general the relationship between biomass and N 574 uptake turned out to be more robust to late split N-applications at V8 and VT growth stages over the 575 years in rainfed maize.

Principal component regression (PCR) relationship to total dry matter yield (TDMY)
577 The total dry matter yield (TDMY) was greatly influenced by the growth and yield attributes along with

Soil Chemical properties 591
The improving soil fertility is critical for enhancing the maize yield and NUE and N fertilization into 592 different splits along with biochar might be beneficial soil amendment for higher side of grain yield and 593 NUE. SOC status changed with different splits of N and biochar treated plots over that of the initial 594 status on surface soil (0-15 cm soil depth). The higher concentration of SOC with the N fertilizer (120 kg 595 Nha -1 ) into 3 splits along with biochar (10 t ha -1 ) treated plots was a results of increased the higher root 596 biomass and plant residues [38]. Additions of biochar along with N also increase microbial activity in the 597 soil which may increase the soil organic carbon content in soil [63]. Hence it is proved that application of 598 stable organic matter like biochar along the mineral fertilizer, not only increases soil fertility but also 599 protects environment by its multifarious functions. Application of N rate (120 kg Nha -1 ) in 2 equal splits 600 at V8 and VT growth stages also recorded higher values of NH 4 + -N (2.29 mg kg -1 ) and NO 3 --N (0.44 mg India. The recommended N management practices recorded lower yield, N uptake and NUEs compared to the agronomic based multi-split N application with skipping of basal N rates. The delayed N