The field experiment was conducted at the Pengyang Experimental station (35°51¢N, 106°48¢E) on the Loess Plateau in Pengyang country of Ningxia Hui Autonomous Region, China. The climate is a typical semi-arid with an average yearly air temperature of 8.1 oC and average annual sunshine of 2518.2 h. The soil had a total N of 0.74 g kg-1, 100.0 mg kg-1 available potassium, 7.39 g kg-1 organic carbon content, 12.62 g kg-1 soil organic matter, 4.3 mg kg-1 adequate phosphorus, and pH 8.8 in the 0-40 cm soil layer, respectively. In the maize growing seasons, the precipitation in 2018, 2019, and 2020 was 567.6, 650.9, and 491.0 mm, respectively, and the average temperature was 17.9, 17.3, and 17.3 oC, respectively Fig. 1.
Experimental design and crop management
The experiment started in 2018, and it was a design that used a randomized block design with fives treatments and three replications. Each plot area is 6 m ´ 7 m (42 m2) and with a 1.2 wide isolation belt away from the others. The five fertilization rates (N0, N1, N2, N3, N4) were 0, 120, 175, 230, and 285 kg N ha-1, respectively.
The high-yield spring maize variety (Dafeng 30) was planted at 82500 ha-1 (wide (60 cm) and narrow (20.2 cm) row spacing) using a hand-planted at a depth of 3-5cm. Maize was sowed on April 25, 24, and 30, 2018-2020, respectively, and harvested on September 29, 29, and October 8. Urea (46% N) was applied to provide N, and diammonium phosphate (46% P and 18% N) provided phosphorus. The specific steps and details of these field management were described in Zhang et al. (2021a) and Zhang et al. (2021b).
Sampling and analysis
Soil NO3--N concentration and N balance
Soil samples were collected to analyze soil NO3--N content (AA3, SEAL Company, Germany) from 0-100 cm soil depths at 20 cm increments at about 0 and 160 days after sowing (DAS) during maize growing season. Soil samples were taken from the ridge, the furrow, and the section connecting ridges and furrows using a hand-held iron soil drill (54 mm diameter) and mixed thoroughly at the same soil depth. After removing impurities such as maize roots and straw, the soil samples were sieved and transported to the laboratory, and placed in a 4℃ refrigerator. Soil NO3--N was extracted using KCl (1 mol L-1) solution. And, the samples were analyzed by the microflow AutoAnalyzer3 to determine NO3--N concentration.
Calculated the N residue (kg ha-1) in each soil layer as follows (Yang et al., 2015):
NR=Cn BD d 0.1 (1)
Cn is the soil N concentration (mg kg-1), BD is the soil bulk density (g cm-3), d is the depth of the soil layer (cm), and 0.1 is a conversion factor.
The soil N balance was calculated as follows (Huang et al., 2020):
N soil +N fertilizer – N uptake –N residue =N loss （kg N ha-1） (2)
N soil is the residual NO3-N in 0-100 cm soil profile before sowing, N fertilizer is the amount of N fertilizer applied, N uptake is the plant N uptake, and N residue is the residual NO3-N in 0-100 cm soil profile at harvest. In this study, N loss was quantitatively comparable to N leaching.
Total available N was calculated as follows (Ding et al., 2020; Huang et al., 2020):
TNS= N soil +N fertilizer (3)
TNS is the total N supply.
Dry matter accumulation and N concentration
Five maize plants were randomly selected from each plot at 35(V1), 50(V3), 60(V6), 80(V12), 100(R1), 130(R3), and 160(R6) days after sowing by hand for determining the dry matter accumulation. Maize plants were first dried at 105°C for 0.5 h and continued to be dried at 75°C for 24 h. The N content of dry matter accumulation was analyzed using a Kjeldahl Autoanalyzer (Kjeltec 8400, FOSS, Denmark).
At the physiological maturity of maize, continuous and unbroken four-row plants (area 7.2 m2) with a length of 3.0 m in the middle row were hand-harvested from each plot to determine the grain yield. The harvest index (HI) was calculated as the ratio of grain yield to biomass yield. The N harvest index (NHI) was determined by the ratio of grain N uptake to biomass N uptake.
Establishment and evaluation of NC
The critical N dilution curve and N dilution boundary curves were determined and developed by Justes et al. (1994) and Plénet et al. (2000):
Where Nc(%) is the N concentration in plant expressed in g 100 g-1 of aboveground biomass weight, DM is the weight of aboveground biomass (Mg ha-1), a is the plant N concentration when the aboveground biomass is 1 Mg ha-1, and b is the characteristic of decreasing nitrogen concentration during maize growth.
Curve validation was performed using the root mean square error (RMSE) and standardized root mean square error (NRMSE) to evaluate the curve fit. The curve stability was measured with reference to the criteria proposed by Jamieson et al. (1991): NRMSE < 10%, excellent curve stability; 10% < NRMSE < 20%, good curve stability; 20% < NRMSE < 30%, fair curve stability; NRMSE > 30%, poor curve stability.
Critical N uptake and nitrogen nutrition index
Nuptc (kg ha-1), the plant N uptake corresponding to the maximum accumulated aboveground biomass of the crop at the critical N concentration state, was the critical N uptake.
Where Nuptc is the critical N uptake (g kg-1).
Substitution of (4) into (5)：
According to the N nutrient index (NNI) curve described by Lemaire et al. (2008),
Where Na is the N concentration value (g kg-1), and Nc is the critical N concentration value. If NNI < 1 indicated an N deficiency, NNI = 1 indicated just the right amount of N, and NNI > 1 indicated an excess of N.
In addition, the integrated NNI value (NNIi) was calculated to describe the crop N status of treatment throughout the whole growth period. An integrated NNI value was calculated as (Lemaire et al., 2008):
NNIi=1/D NNIs ds (8)
Where D is the total number of days in the whole growth stages, NNIs is the NNI value under each sampling day, and ds is the interval between two sampling dates.
Nitrogen use efficiency
Calculated N use efficiency indexes as follows (Caviglia et al., 2014):
NuptE = (N uptake)/ TNS (9)
NutEGY = GY/ (N uptake) (10)
NutEBY = BY/ (N uptake) (11)
NUEGY = GY/ TNS (12)
NUEBY = BY/ TNS (13)
NuptE (kg kg-1) is N uptake efficiency, NutE (kg kg-1) is N use efficiency, N uptake is plant N uptake, GY is grain yield, BY is biomass yield, and TNS is total N supply.
Results were calculated using Microsoft Excel 2013, plotted with Origin 2021 software, and statistically analyzed using SPSS 26.0 software. Mean comparisons were made at P < 0.05 using Fisher's LSD (Least Significant Difference) test. Used 2018-2019 data to build the curve and used 2020 data for verification. This study evaluated linear, quadratic, and linear-plateau models and reported best-fitting curves (Cerrato and Blackmer, 1990).