Optimized nitrogen rate, plant density, and irrigation level reduced ammonia emission and nitrate leaching on maize farmland in the oasis area of China

Experimental location

The experiment was conducted at Wuwei, Gansu Agricultural University Experiment field in 2018 and 2019, from April to September, on 37°96′N, 102°64′E altitude, and 1,776 m above sea level. The experiment station is situated in the eastern part of the Hexi corridor of Northwestern China with a temperate and arid climate. Soils in this area are classified as Aridisol; a type of calcareous desert soil (Shao et al., 2021). Average annual sunshine duration, air temperature, precipitation, and potential evaporation in the location are 2,945 h, 7.2 °C, 150 mm, and 2,400 mm, respectively. Daily precipitation and air temperature were obtained for the experiment in both years by the climate data sub-station of Wuwei, 100 m away from the experiment field (Fig. 1). Soil ammonium nitrogen (NH₄⁺–N), soil nitrate-nitrogen (NO₃–N), soil total nitrogen (TN), and soil bulk density by soil depth-layers before the start of the experiment in April of 2018 and 2019 are shown in (Table 1). The experimental field was monocropped with maize the previous year before this experiment was conducted.

Experimental layout

The experiment was a 3 × 3 × 2 factorial experiment in a randomized complete block design (RCBD) with three replicates. Each treatment had a plot measuring 7 × 5 m and was separated by 80 cm. A 50 cm ridge was built between plots to reduce the potential of water movement between plots. The treatment comprises three N rates of urea (46–0–0 of N–P₂O₅–K₂O) (N₀ = 0 kg N ha⁻¹, N₁ = 270 kg N ha⁻¹, and N₂ = 360 kg N ha⁻¹), three maize plant densities (D₁ = 75,000 plants/ha⁻¹, D₂ = 97,500 plants/ha⁻¹, and D₃ = 120,000 plants/ha⁻¹), and spaced at 40 × 33 cm, 40 × 26 cm, and 40 × 21 cm to get the required plant densities respectively, and combined with two irrigation levels (W₁ = 5,250 m³/hm² and W₂ = 4,740 m³/hm²). Drip irrigation lines with valves were connected to water meters to regulate and measure the amount of water irrigated.

Field management practices

A moldboard plough was used to till soil to a depth of 20 cm in April of both years, followed by a rotary tiller to loosen soil lumps. Nitrogen application was applied in triplicates, with 20% of the total N broadcasted and incorporated to 20 cm soil depth using rotary tiller prior to maize seeding, and the remaining split into two portions. It was applied by deep placement into the soil at the nine-leaf collar stage and 15 days after flowering. A white plastic film of (120 cm wide and 0.01 mm thickness) was mulched on the plots before maize seeding was done each year. Maize cultivar (Xianyu 335) was sown on 17th and 19th April 2018 and 2019 using a hand-held pressure-inject planter. Irrigation was done at 90, 120, and 90 mm at the nine, fourteen leaf, and grain-filling stages, respectively, using drip irrigation lines. Water meters were connected to the drip lines to measure the amount of water irrigated. Weeds were controlled by hand throughout the crop growing period in both years. No chemical application for weed and pest control and no diseases and pests were recorded in the experiment.

Ammonia gas sampling and analysis

The measurement of ammonia emission was done as described in Wang et al. (2005), Akbari et al. (2020). The polyvinyl chloride rigid measuring container is a plastic tube with an interior diameter and height measuring 15 and 12 cm respectively. Two-disc foams 2 cm of 2 cm thickness and 16 cm diameter were uniformly submerged in 20 ml solution of glycerol phosphate (40 ml of glycerol + 50 ml of phosphoric acid, and the volume was adjusted to 1,000 ml). The two layers of disc foams were positioned in the firm polyvinyl chloride rigid tube so that the lower layer of the disc foam is 5 cm from the ground and the upper layer of the disc foam is at the level with the top of the tube. Ammonia emission was believed to start on the day of fertilization, sowing, and during the crop growth period, and after crop harvest. Therefore measurements started immediately after sowing. The gas was extracted by putting the disc foam into a 150 ml sealed bag and adding 100 ml of Sodium chloride and shook for an hour. The samples were filtered, and 5 ml of the filtrate was transferred to a 50 ml volumetric flask, 5 ml of Sodium chloride was added to the 50 ml volumetric flask containing the sample filtrate. Phenol liquor, Sodium hydroxide aqueous of 5 ml and 1 ml of masking agent was added and allowed to stand for an hour before distilled water was filled to the volumetric flask mark. The samples were analyzed using a mass spectrophotometer. The machine was switched on for 20 min in advance to warm and adjusted to 625.0 nm. Daily (mg day⁻¹) and cumulative NH₃–N losses (kg N ha⁻¹) were calculated to express the N loss in relation to the N applied.

Soil sampling for nitrate and analysis

Soil samples were collected from 0–30, 30–60, 60–90, 90–120, and 120–150 cm soil layers using a soil sampling auger (internal diameter of 4.0 cm) from each plot immediately preceding maize sowing, during and after maize harvest in each year. The interval and days for the sampling were done before planting and at 30, 60, 90, 120, and 150 days after planting. The samples were air-dried using the oven drying method (O’Kelly, 2004) and ground into fine powder and sieved through (<2 mm) wire mesh for the chemical analysis. Soil TN was determined by the standard Semi-micro–Kjeldahl method (Bao, 1999). Soil NH₄⁺–N and NO₃–N were determined by the spectrophotometry method with a Discrete Auto Analyzer (SMARTChem 450, Beijing, China), (NATESC, 2006). The leached and accumulated N (LAN) was calculated using the following equation (Dai et al., 2016): $$\mathrm{L}\mathrm{A}\mathrm{N}=\mathrm{T}\mathrm{i}\times \mathrm{D}\mathrm{i}\times \mathrm{C}\mathrm{i}\text{/}10$$where Ti and Di represents the thickness and bulk density of the soil layer at (30 cm), (g cm⁻³), respectively, Ci represents soil NO₃–N leached or accumulated (mg kg⁻¹) of the corresponding layer, and 10 represents the conversion coefficient.

Statistical Analyses

Analysis of variance and effect was conducted using the general linear model function in Genstat statistical software (version 12., Chicago, IL, USA) to determine whether N application rate, plant density, and irrigation levels influence, NH₃–N emission and NO₃–N leaching and their interactive effects. Differences among means were tested with Fisher’s protected least significant difference at p ≤ 0.05.

Soil ammonium nitrogen, nitrate-nitrogen, total nitrogen, bulk density, precipitation, and temperature before the start of the experiment in 2018 and 2019

The soil at the experiment site is coarse sandy clay and classified as aridisol, a desert soil with calcareous particles (Table 1) (Shao et al., 2021). The lowest NH₄⁺–N was recorded in 2018 (11.945 mg kg⁻¹) while the highest NH₄⁺–N was recorded in 2019 (16.077 mg kg⁻¹) at 60–90 soil depth. The lowest NO₃–N was recorded at 120–150 cm soil depth (3.232 mg kg⁻¹) in 2018 while the highest NO₃–N (5.567 mg kg⁻¹) was recorded at 60–90 cm soil depth in 2019. At 120 and 150 cm soil depths in both years, there was no much variation in NH₄⁺–N and NO₃–N concentration but differed to other layers. This variation in NH₄⁺–N and NO₃–N in the experimental soil may be partly attributed to the varying precipitation and air temperature (Fig. 1).

Influence of nitrogen fertilization, plant density, and irrigation regime on nitrate accumulation and leaching

The effect of nitrogen fertilizer rate, plant density and irrigation level on NO₃–N leaching and accumulation at the various sampling depths and days after planting over the 2 years are shown in Figs. 2A, 2B, 2C, 2D & 2E and Tables 2, 3, 4, and 5. It was found that the factors had marked individual and interactive effects on leaching and the accumulation of nitrate in the soil. Irrigation level significantly (p < 0.05) affected the leaching and accumulation of nitrate in the soil. Nitrate accumulation relatively increased with a decrease in irrigation level for the first three soils depths but increased correspondingly at 90–120 cm and 120–150 cm especially at 30 days through to 120 days after planting in 2018. In 2019 however, the amount of nitrate accumulated in the soil was generally higher at 4,740 m³/hm² (W₂) except for 60 days after planting where the higher (W₁) irrigation level resulted in more leaching of nitrate (Figs. 2A, 2B, 2C, 2D & 2E). With regard to N rate main effect, considerable (p < 0.05) variation occurred in relation to leaching and accumulation of the nitrate. Accumulation increased as higher amount of N fertilizer was applied with exception of 90 days after planting in 2019 where the reverse occurred at the first four soil depths (Figs. 2A, 2B, 2C, 2D, & 2E). A significant effect was also recorded in relation to plant density, with a general increase in nitrate accumulation at the lower plant densities as compared to the highest density (D₃) (Figs. 2A, 2B, 2C, 2D, & 2E) in both years. In terms of interaction effects, all two-factor and three-factor combinations generally were significant (p < 0.05) and affected leaching and accumulation regardless of the soil depth, duration after planting and year of cultivation. However, no linear or specific trends were observed to that effect. Nonetheless, the lowest (0.35 mg kg⁻¹) leaching and accumulation occurred at the combination of 270 kg N ha⁻¹, 120,000 plants/ha⁻¹ and 4,740 m³/hm² at 90–120 cm soil depth at 150 days after planting in 2018 while the highest (9.94 mg kg⁻¹) occurred at the combination of 0 kg N ha⁻¹, 75,000 plants/ha⁻¹ and 4,740 m³/hm² at 90–120 cm soil depth at 60 days after planting in 2019.

It was also observed that accumulation of nitrate generally decreased as the soil depth increased to a depth of 60–90 cm where it increased thereafter for planting duration of 30 days to 120 days (Figs. 2A & 2D). Conversely, nitrate accumulation in the soil generally declined as soil depth advanced at 150 days after planting (Fig. 2E). Nonetheless, more accumulation was shown to have occurred at the extreme soil depths at certain durations after planting. For example, higher nitrate accumulation occurred at 0–30 cm for 30 days, 90 days, and 150 days after planting while higher values were recorded at a soil depth of 120–150 cm for 60 days and 120 days after planting in 2018. No such trends were however recorded for soil depth for the various factor interactions. It was further observed that leaching in 2019 was more pronounced than in 2018.

Ammonia emission on maize farmland as affected by nitrogen fertilization, plant density and irrigation, and their interactions

The influence of N rate, plant density and irrigation level on ammonia emission is presented in Fig. 3 and Tables 6 and 7. It was found that ammonia emission was significantly (p < 0.05) affected by irrigation level at 60 days after planting in 2018 and at 30 days, 90 days, and 120 days after planting in 2019 with a general decrease as irrigation level increased (Fig. 3). Nitrogen rate did not have any significant influence on ammonia emission in 2018 but markedly affected emission at (p < 0.05) at 90 days and 120 days after planting in 2019. The quantity of ammonia emitted averagely increased as more fertilizer was applied especially at 90 days and 150 days in 2018 and at 30 days, 90 days and 150 days after planting in 2019 (Fig. 3). Plant density alone did not have any significant effect (p > 0.05) on the amount of ammonia emitted irrespective of the duration after planting and the year of cultivation even though D₁ (75,000 plants/ha⁻¹) relatively recorded the highest emission of ammonia as compared to the higher plant densities (Fig. 3).

Interactively, the combination of N rate and irrigation level affected emission (p < 0.05) at 120 days and 90 days after planting in 2018 and 2019, respectively (Table 6). Regardless, there was an overall decrease in emission as more N rate was applied at the higher irrigation level (W₁) while an increase was rather observed as N rate increased and irrigation level lowered. The combination of irrigation level and plant density as well as that of N rate and plant density however did not have any substantial influence (p > 0.05) on ammonia emission irrespective of the days after planting. In terms of three-factor influence, marked differences only existed at 150 days after planting in 2018 and at 90 days after planting in 2019 (Table 7). At 150 days after planting in 2018, the lowest (0.033 kg N ha⁻¹) emission occurred when zero N rate was applied to the highest plant density irrigated at 5,250 m³/hm² (W₁ × N₀ × D₃) while the highest (0.090 kg N ha⁻¹) was recorded by W₁ × N₀ × D₁. For the 90 days after planting in 2019, the lowest (0.010 kg N ha⁻¹) emission was recorded at W₁ × N₁ × D₁ and the highest (0.91 kg N ha⁻¹) was recorded at W₂ × N₂ × D₁. Nonetheless, the overall lowest (0.001 kg N ha⁻¹) emission however occurred on W₁ × N₀ × D₃ at 60 days after planting in 2018 while the highest (0.248 kg N ha⁻¹) was recorded on W₁ × N₀ × D₁ at 60 days after planting in 2019. It could also be observed that emission relatively increased with an increase in duration after planting except at 60 days in 2018 and 90 days in 2019. It could further be observed that ammonia emission in 2019 was relatively higher than that recorded in 2018.