E ﬀ ects of Irrigation Water Salinity on Soil Properties, N 2 O Emission and Yield of Spring Maize under Mulched Drip Irrigation

: Brackish water has been widely used to irrigate crops to compensate for insu ﬃ cient freshwater water supply for agricultural use. The goal of this research was to determine an e ﬃ cient brackish water use method to increase irrigation e ﬃ ciency and reduce N 2 O emission. To this end, we conducted a ﬁeld experiment with four salinity levels of irrigation water (1.1, 2.0, 3.5, and 5.0 g · L − 1 with drip irrigation) at Hetao Irrigation District (Inner Mongolia, China) in 2017 and 2018. The results show that irrigation with 3.5–5.0 g · L − 1 water salinity increased the soil salinity compared with irrigation using 1.1–2.0 g · L − 1 water salinity. The soil water content with 5.0 g · L − 1 brackish water irrigation was signiﬁcantly higher than with 1.1–3.5 g · L − 1 water salinity due to the e ﬀ ect of salinity on crop water uptake. The overall soil pH increased with the increase in irrigation water salinity. Saturated soil hydraulic conductivity decreased with the increase in irrigation water salinity. These results indicate that brackish water irrigation aggravates the degree of soil salinization and alkalization. The soil N 2 O cumulative ﬂux resulting from irrigation with 5.0 g · L − 1 water salinity was 51.18–82.86% higher than that resulting from 1.1–3.5 g L − 1 water salinity in 2017, and was 32.38–44.79% higher than that resulting from 1.1–2.0 g · L − 1 in 2018. Irrigation with brackish water reduced maize yield, and the reduction in yield in 2018 was greater than that in 2017, but irrigation with 2.0 g · L − 1 brackish water did not signiﬁcantly reduce maize yield in 2017. These results suggest that reducing the salinity of irrigation water may e ﬀ ectively reduce soil N 2 O emission, alleviate the degree of soil salinization, and increase crop yield.


Introduction
Situated in Inner Mongolia, Hetao Irrigation District (HID) is an important commodity grain-producing area in Northwest China. Due to the region's arid climate, agricultural production has always depended on local Yellow River irrigation. To promote agricultural production, mulched planting technology, due to its heat preservation, water saving, and yield increases, has been widely promoted in HID. However, with the development of the economy, the use of industrial water and domestic water in HID is increasing. Meanwhile, because of policy regulation, the amount of water diverted from the Yellow River will decrease by 20% (from 5 billion m 3 to 4 billion m 3 ) in the next 10 years [1]. The problem of insufficient agricultural water use is becoming increasingly serious. Therefore, it is imperative to develop methods for efficient water-saving irrigation. Simultaneously, alternative water sources for irrigation should be found for the sustainable development of agriculture in HID.

Experimental Site
We used the Shuguang Experimental Station (40 • 46 N, 107 • 24 E, and elevation of 1039 m) to conduct field experiments with spring maize for two seasons (2017 and 2018). The location of the experimental site is illustrated in Figure 1. The region has a typically arid continental climate with an average annual precipitation of 105 mm, average annual evaporation of 2306.5 mm, average annual temperature of 7.8 • C, and average annual relative humidity of 48.9% [2,27]. The soil properties are shown in Table 1

Experimental Site
We used the Shuguang Experimental Station (40°46′ N, 107°24′ E, and elevation of 1039 m) to conduct field experiments with spring maize for two seasons (2017 and 2018). The location of the experimental site is illustrated in Figure 1. The region has a typically arid continental climate with an average annual precipitation of 105 mm, average annual evaporation of 2306.5 mm, average annual temperature of 7.8 °C, and average annual relative humidity of 48.9% [2,27]. The soil properties are shown in Table 1

Experimental Design and Field Management
There were four treatments in this field experiment: (1) groundwater drip irrigation with salinity of 1.1 g·L −1 (T1), (2) brackish water drip irrigation with salinity of 2.0 g·L −1 (T2), (3) brackish water drip irrigation with salinity of 3.5 g·L −1 (T3), and (4) brackish water drip irrigation with salinity of 5.0 g·L −1 (T4). The four treatments were laid out in randomized complete blocks with three replicates for a total of 12 plots. There were 1-m buffer channels between adjacent plots to avoid interaction between the treatments.

Experimental Design and Field Management
There were four treatments in this field experiment: (1) groundwater drip irrigation with salinity of 1.1 g·L −1 (T1), (2) brackish water drip irrigation with salinity of 2.0 g·L −1 (T2), (3) brackish water drip irrigation with salinity of 3.5 g·L −1 (T3), and (4) brackish water drip irrigation with salinity of 5.0 g·L −1 (T4). The four treatments were laid out in randomized complete blocks with three replicates for a total of 12 plots. There were 1-m buffer channels between adjacent plots to avoid interaction between the treatments.
Seeds of the spring maize hybrid Ximeng No.3358, selected as the experimental crop, were sown on 26 April 2017 and 28 April 2018, and harvested on 12 September 2017 and 15 September 2018. The growth stages lasted 140 and 141 days, respectively. The layout of the experimental plot is shown in Figure 3. Two rows of spring maize were planted in each bed. The intra-row spacing was 0.3 m and the inter-row spacing was 0.8 m. The surface of the intra-row soil was covered with plastic mulch. The planting density of the maize was 60,000 plants·ha −1 . Drip irrigation belts were arranged in the intra-row center. The thickness of the drip irrigation belts applied in the experiment was 0.4 mm, and the lateral diameter of the drip irrigation belts was 16 mm. The drippers were classified as embedded type. The interval between the drippers was 0.3 m, and the average discharge of each dripper was 2.0 L·h −1 . To satisfy the demand for different levels of irrigation water salinity, we used the groundwater of the experimental station and added an equal ratio of NaCl/KCl (Table 2). On the basis of previous publications, the same amounts of irrigation (300 mm) and nitrogen (300 kg·ha −1 ) were applied for each treatment in the two seasons of the experiment [28,29]. The drip irrigation frequency was designed to be 6 days, and the irrigation amount was 15-25 mm each time [30]. Before they were sown, all plots were fertilized with 375 kg·ha −1 diammonium phosphate and 75 kg·ha −1 urea, and the fertilizers were applied to the soil tillage layer (0-20 cm) by machine [28,29]. The residual nitrogen was used for fertilization in the jointing stage, heading stage, and filling stage at a ratio of 2:2:1 in drip irrigation treatments [31]. The amounts of water and fertilizer were measured using a water meter (DN25, Aimeihui, Zhejiang, China) and a proportional fertilizer pump (Mixrite 2504, Tefen, Kibbutz Nahsholim, Israel) in each plot. The irrigation and fertilization dates for each treatment in 2017 and 2018 are shown in Figure 4.

Soil N2O Emission
Soil N2O gas samples were collected with a static chamber and analyzed by gas chromatography [32]. The sampling chamber was divided into a top chamber (50 × 50 × 50 cm) and a ground chamber (50 × 50 × 15 cm). The stainless-steel top chamber was covered with a foam board to prevent the temperature from rising too quickly, and two battery-operated fans were equipped to mix the gas. A thermometer slot was employed in the top chamber to measure the temperature during sampling. One week before sampling, the ground chamber was inserted in the center of each plot at a soil depth of 15 cm until the spring maize harvest. To represent the actual N2O emission flux during drip

Soil N 2 O Emission
Soil N 2 O gas samples were collected with a static chamber and analyzed by gas chromatography [32]. The sampling chamber was divided into a top chamber (50 × 50 × 50 cm) and a ground chamber (50 × 50 × 15 cm). The stainless-steel top chamber was covered with a foam board to prevent the temperature from rising too quickly, and two battery-operated fans were equipped to mix the gas. A thermometer slot was employed in the top chamber to measure the temperature during sampling. One week before sampling, the ground chamber was inserted in the center of each plot at a soil depth of 15 cm until the spring maize harvest. To represent the actual N 2 O emission flux during drip irrigation with film mulching, the ground chamber was buried according to the intraand inter-row proportions. In 2018, the ground chambers were also embedded in and between the rows of each plot to explore the relationship between N 2 O emissions and different locations. During sampling, the convex part of the top chamber was placed into the groove of the bottom chamber and water-sealed with no air exchange. A 100-mL gas sample was extracted from the chamber using a syringe 0, 10, 20, and 30 min after closing the chamber, and the samples were then transferred to sampling bags and tested within a week. The soil N 2 O samples were collected from 08:30 to 11:30 a.m. every 7-14 days, and there were additional samplings on the first, third, fifth, and seventh days after fertilization [33,34]. The concentration of N 2 O was measured by gas chromatography (Agilent 7890A, Agilent Technologies, Santa Clara, CA, USA), and its emission flux (F, µg·m −2 ·h −1 ) was determined by the following equation [35]: where ρ is the density of N 2 O (kg·m −3 ) in standard temperature and pressure, H is the top chamber height (m), ∆c/∆t is the slope of N 2 O concentration variation over time (ppb·h −1 ), and T is the average temperature during sampling (K). The N 2 O (kg·ha −1 ) cumulative flux (S) from spring maize soils was calculated by the following equation [36]: where Fi represents the N 2 O emission flux (kg·ha −1 ·day −1 ), Di represents days between two sampling times (days), and n represents the sample number.

Soil Properties
When sampling the gas, soil samples were acquired from a depth of 0-0.3 m, and sampling points were located between intra-row maize plants and from the center of inter-row spaces in each plot. Part of the soil was packed in an aluminum box and dried at 105 • C for 8-10 h. The remaining soil samples were air-dried, pulverized, and then sifted through a 1 mm sieve. A 10 g subsample was weighed and added to 50 mL of deionized water. After the mixture was shaken, left to stand, and centrifuged, the supernatant EC 1:5 (µS·cm −1 ) and pH were measured by a multi-parameter tester (SG23, Mettler Toledo, Shanghai, China). Another 5 g subsample was weighed and added to 50 mL of a KCl solution. After the mixture was shaken, left to stand, and filtered, the NO 3 − content (mg·kg −1 ) in the supernatant was measured by a continuous flowing analyzer (Alliance FUTURA, AMS, Frépillon, France). After the spring maize harvest in 2018, the saturated hydraulic conductivity (cm·s −1 ) of 0-20 cm of soil in the intra-row was measured by the constant-head method.

Spring Maize Yield and IWUE
At harvest time, 6 maize plants from each plot were chosen to measure yield. Maize kernel samples were weighed after sun drying and threshing, and the actual yield was converted according to the number of plants. The IWUE was calculated by the following equation: where E represents the IWUE (kg·m −3 ), Y represents the yield of spring maize (kg·ha −1 ), and Q represents the cumulative amount of irrigation water (m 3 ·ha −1 ).

Statistical Analysis
The ANOVA procedure in SPSS software (IBM, Armonk, NY, USA) was used to statistically analyze the data. A Duncan multiple comparisons test was used to make multiple comparisons of annual average values. Pearson's correlation analysis procedure in SPSS software (IBM, Armonk, NY, USA) was used to compare the correlations between soil N 2 O emission, soil properties, and maize yield. The confidence level is 95%. Figure 5 shows the average intra-and inter-row soil water content of spring maize under different treatments in 2017 and 2018. The soil water content in intra-row soil was higher than that of inter-row soil under T1-T4 treatments (p < 0.05). On the whole, the soil water content in 2018 was higher than in 2017 (p < 0.05), which might be attributed to the large volumes and high frequency of precipitation in 2018 ( Figure 2). The irrigation water salinity had a significant effect on the soil water content (p < 0.05). The intra-and inter-row soil water content under T4 treatment was much higher than that under T1-T3 treatments (p < 0.05), with intra-row increases of 25.55-35.43% and 19.66-29.09% and inter-row increases of 41.38-57.69% and 23.19-37.10% in 2017 and 2018, respectively. Figure 5 shows the average intra-and inter-row soil water content of spring maize under different treatments in 2017 and 2018. The soil water content in intra-row soil was higher than that of inter-row soil under T1-T4 treatments (p < 0.05). On the whole, the soil water content in 2018 was higher than in 2017 (p < 0.05), which might be attributed to the large volumes and high frequency of precipitation in 2018 ( Figure 2). The irrigation water salinity had a significant effect on the soil water content (p < 0.05). The intra-and inter-row soil water content under T4 treatment was much higher than that under T1-T3 treatments (p < 0.05), with intra-row increases of 25  ; different capital letters between treatments represent that the difference in intra-row is significant at the 0.05 level; different lowercase letters between treatments represent that the difference in inter-row is significant at the 0.05 level. Figure 6 shows the average EC1:5 in intra-and inter-row soil of spring maize under different treatments in 2017 and 2018. The EC1:5 in 2018 was lower than that in 2017 in both intra-and interrow soil (p < 0.05), which might be caused by the fact that the heavy precipitation has a leaching effect on the soil salt in 2018. The irrigation water salinity had a significant effect on the soil EC1:5 (p < 0.05). The intra-row soil EC1:5 under T3-T4 treatments was much higher than that under T1-T2 treatments (p < 0.05), with the former being higher than the latter by 43.59-85.33% in 2017 and 31.37-65.35% in 2018, and there was a significant difference in the soil EC1:5 between T3 and T4 treatments (p < 0.05). The inter-row soil EC1:5 under T4 treatment significantly increased by 62.92-77.60% in 2017 and 44.90-95.41% in 2018 compared with that under T1-T3 treatments (p < 0.05), and the difference between T3 treatment and T1-T2 treatments in 2018 was significant (p < 0.05). ; different capital letters between treatments represent that the difference in intra-row is significant at the 0.05 level; different lowercase letters between treatments represent that the difference in inter-row is significant at the 0.05 level. Figure 6 shows the average EC 1:5 in intra-and inter-row soil of spring maize under different treatments in 2017 and 2018. The EC 1:5 in 2018 was lower than that in 2017 in both intra-and inter-row soil (p < 0.05), which might be caused by the fact that the heavy precipitation has a leaching effect on the soil salt in 2018. The irrigation water salinity had a significant effect on the soil EC 1:5 (p < 0.05). The intra-row soil EC 1:5 under T3-T4 treatments was much higher than that under T1-T2 treatments (p < 0.05), with the former being higher than the latter by 43.59-85.33% in 2017 and 31.37-65.35% in 2018, and there was a significant difference in the soil EC 1:5 between T3 and T4 treatments (p < 0.05). The inter-row soil EC 1:5 under T4 treatment significantly increased by 62.92-77.60% in 2017 and 44.90-95.41% in 2018 compared with that under T1-T3 treatments (p < 0.05), and the difference between T3 treatment and T1-T2 treatments in 2018 was significant (p < 0.05).  ; different capital letters between treatments represent that the difference in intra-row is significant at the 0.05 level; different lowercase letters between treatments represent that the difference in inter-row is significant at the 0.05 level. Figure 7 shows the average pH value in the intra-and inter-row soil of spring maize under different treatments in 2017 and 2018. The pH value of soil tillage layer under all treatments in 2018 increased to varying degrees compared with that in 2017, and the range increased with the increase in irrigation water salinity (p < 0.05). Irrigation water salinity had a significant effect on pH value in the soil tillage layer (p < 0.05). In 2017, the pH value under T3-T4 treatments increased by 0.48-1.68% compared with that under T1-T2 treatments in intra-row soil (p < 0.05), and the difference between ; different capital letters between treatments represent that the difference in intra-row is significant at the 0.05 level; different lowercase letters between treatments represent that the difference in inter-row is significant at the 0.05 level.  Figure 7 shows the average pH value in the intra-and inter-row soil of spring maize under different treatments in 2017 and 2018. The pH value of soil tillage layer under all treatments in 2018 increased to varying degrees compared with that in 2017, and the range increased with the increase in irrigation water salinity (p < 0.05). Irrigation water salinity had a significant effect on pH value in the soil tillage layer (p < 0.05). In 2017, the pH value under T3-T4 treatments increased by 0.48-1.68% compared with that under T1-T2 treatments in intra-row soil (p < 0.05), and the difference between T3 and T4 treatments was significant in intra-row soil (p < 0.05), but there was no significant difference in pH between T1-T4 treatments in inter-row soil (p > 0.05). In 2018, the pH value under T3-T4 treatments increased by 2.13-4.06% compared with that under T1-T2 treatments in intra-row soil (p < 0.05), and the difference between T3 and T4 treatments was significant (p < 0.05). There were significant differences in pH values among T1-T4 treatments in inter-row soil (p < 0.05). ; different capital letters between treatments represent that the difference in intra-row is significant at the 0.05 level; different lowercase letters between treatments represent that the difference in inter-row is significant at the 0.05 level. Figure 7 shows the average pH value in the intra-and inter-row soil of spring maize under different treatments in 2017 and 2018. The pH value of soil tillage layer under all treatments in 2018 increased to varying degrees compared with that in 2017, and the range increased with the increase in irrigation water salinity (p < 0.05). Irrigation water salinity had a significant effect on pH value in the soil tillage layer (p < 0.05). In 2017, the pH value under T3-T4 treatments increased by 0.48-1.68% compared with that under T1-T2 treatments in intra-row soil (p < 0.05), and the difference between T3 and T4 treatments was significant in intra-row soil (p < 0.05), but there was no significant difference in pH between T1-T4 treatments in inter-row soil (p > 0.05). In 2018, the pH value under T3-T4 treatments increased by 2.13-4.06% compared with that under T1-T2 treatments in intra-row soil (p < 0.05), and the difference between T3 and T4 treatments was significant (p < 0.05). There were significant differences in pH values among T1-T4 treatments in inter-row soil (p < 0.05).    Figure 8 shows the average intra-and inter-row soil NO3 -content of spring maize under different treatments in 2017 and 2018. On the whole, the soil NO3 -content in inter-row soil was higher than that of intra-row soil (p < 0.05). Due to the leaching effect on soil NO3 − content by heavy precipitation in 2018 (Figure 2), NO3 − content in both intra-and inter-row soil was lower than that in 2017 (p < 0.05). Irrigation water salinity had no significant effect on the average NO3 − content in intraand inter-row soil (p > 0.05).

Saturated Soil Hydraulic Conductivity
The results of saturated soil hydraulic conductivity (Ksat) at harvest in 2018 are shown in Figure  9. Under all treatments, Ksat decreased to varying degrees compared with that before sowing (Table  1); this decrease was due to the tillage of the surface soil before sowing. Irrigation water salinity

Saturated Soil Hydraulic Conductivity
The results of saturated soil hydraulic conductivity (K sat ) at harvest in 2018 are shown in Figure 9. Under all treatments, K sat decreased to varying degrees compared with that before sowing (Table 1); this decrease was due to the tillage of the surface soil before sowing. Irrigation water salinity significantly affected K sat (p < 0.05). K sat decreased with the increase in irrigation water salinity: the value was 54.92-75.93% lower for the T2-T4 treatments than that for the T1 treatment. Treatment   T1  T2  T3  T4   0.00   Treatment   T1  T2  T3  T4 0.00

Saturated Soil Hydraulic Conductivity
The results of saturated soil hydraulic conductivity (Ksat) at harvest in 2018 are shown in Figure  9. Under all treatments, Ksat decreased to varying degrees compared with that before sowing (Table  1); this decrease was due to the tillage of the surface soil before sowing. Irrigation water salinity significantly affected Ksat (p < 0.05). Ksat decreased with the increase in irrigation water salinity: the value was 54.92-75.93% lower for the T2-T4 treatments than that for the T1 treatment.

Seasonal N2O Emission Flux
The soil N2O emission flux in 2017 and 2018 had a similar pattern during the growth stage of spring maize ( Figure 10). The soil was proven to be the emission source of N2O. We observed six peaks of N2O emission flux in total, and they appeared one to five days after fertilization in 2017 and 2018. Compared with the peak value in the one to five days after fertilization, the average peak values of N2O emissions in treatments T1, T2, T3, and T4 were 81.57, 89.30, 95.14, and 139.32 µg·m −2 ·h −1 in 2017, and there was a significant difference between the T4 treatment and T1-T3 treatments (p < 0.05). The average peak values of N2O emissions in treatments T1, T2, T3, and T4 were 100.62, 114.30, 154.45, and 184.23 µg·m −2 ·h −1 in 2018, and there was a significant difference between T3-T4 treatments and the T1 treatment (p < 0.05). Overall, the N2O peak emission flux increased with the increase in irrigation water salinity. There was no peak N2O emission with irrigation only. 3.6.2. Seasonal N2O Emission Flux from Different Locations Figure 11 show that the N2O emission flux from intra-and inter-row soil reached peaks from the first to fifth day after fertilization. The average peak values of N2O emissions in treatments T1, T2,  Figure 11 show that the N 2 O emission flux from intra-and inter-row soil reached peaks from the first to fifth day after fertilization. The average peak values of N 2 O emissions in treatments T1, T2, T3, and T4 were 178.09, 207.13, 224.77, and 305.32 µg·m −2 ·h −1 in intra-row soil, and there was a significant difference between the T4 treatment and T1-T3 treatments (p < 0.05). The average peak values of N 2 O emissions in treatments T1, T2, T3, and T4 were 57.04, 62.08, 114.89, and 116.11 µg·m −2 ·h −1 in inter-row soil, and there was a significant difference between T3-T4 treatments and T1-T2 treatments (p < 0.05). Meanwhile, the peak N 2 O emission flux from intra-row soil was higher than that of inter-row soil under T1-T4 treatments (p < 0.05).

N2O Cumulative Flux
The cumulative N2O flux under all treatments in 2017 and 2018 is shown in Figure 12a. The cumulative N2O flux under T4 treatment increased by 51.18-82.86% in 2017 and 19.83-44.79% in 2018 compared with that under T1-T3 treatments, and the difference between T4 treatment and T1-T2 treatments in the two years was significant (p < 0.05).
The cumulative N2O flux at different locations under all treatments in 2018 is shown in Figure  12b. The cumulative N2O flux under T4 treatment increased by 22.49-44.76% in intra-row soil and 14.77-46.38% in inter-row soil compared with that under T1-T3 treatments, and the difference between T4 treatment and T1 treatment in intra-and inter-row soil was significant (p < 0.05). Also, the cumulative N2O emission from intra-row soil was significantly higher than that of inter-row soil under T1-T4 treatments (p < 0.05). .77-46.38% in inter-row soil compared with that under T1-T3 treatments, and the difference between T4 treatment and T1 treatment in intra-and inter-row soil was significant (p < 0.05). Also, the cumulative N 2 O emission from intra-row soil was significantly higher than that of inter-row soil under T1-T4 treatments (p < 0.05).

Spring Maize Yield and IWUE
The spring maize yield and IWUE under all treatments in 2017 and 2018 is shown in Figure 13. The increase in irrigation water salinity significantly reduced spring maize yield and IWUE (p < 0.05). After irrigation with brackish water for one year, the yields under T3-T4 treatments significantly decreased by 11.81-23.93% compared with that under T1 treatment (p < 0.05). After two years, the yields under T2-T4 treatments significantly decreased by 13.86-31.47% compared with that under T1 treatment (p < 0.05). The average IWUE under T1 treatment was significantly larger than that under T3-T4 treatments after irrigation with brackish water in 2017 and 2018 (p < 0.05).

Correlation Analysis among Irrigation Water Salinity, Soil Properties, N2O Emissions and Maize Yield
The correlation between soil N2O emission flux and soil properties is shown in Figure 14. There was a significant linear correlation between soil N2O emission flux and soil temperature when the soil temperature ranged from 13.7 °C to 30.1 °C, indicating that the soil temperature was an important

Spring Maize Yield and IWUE
The spring maize yield and IWUE under all treatments in 2017 and 2018 is shown in Figure 13. The increase in irrigation water salinity significantly reduced spring maize yield and IWUE (p < 0.05). After irrigation with brackish water for one year, the yields under T3-T4 treatments significantly decreased by 11.81-23.93% compared with that under T1 treatment (p < 0.05). After two years, the yields under T2-T4 treatments significantly decreased by 13.86-31.47% compared with that under T1 treatment (p < 0.05). The average IWUE under T1 treatment was significantly larger than that under T3-T4 treatments after irrigation with brackish water in 2017 and 2018 (p < 0.05).

Spring Maize Yield and IWUE
The spring maize yield and IWUE under all treatments in 2017 and 2018 is shown in Figure 13. The increase in irrigation water salinity significantly reduced spring maize yield and IWUE (p < 0.05). After irrigation with brackish water for one year, the yields under T3-T4 treatments significantly decreased by 11.81-23.93% compared with that under T1 treatment (p < 0.05). After two years, the yields under T2-T4 treatments significantly decreased by 13.86-31.47% compared with that under T1 treatment (p < 0.05). The average IWUE under T1 treatment was significantly larger than that under T3-T4 treatments after irrigation with brackish water in 2017 and 2018 (p < 0.05).

Correlation Analysis among Irrigation Water Salinity, Soil Properties, N2O Emissions and Maize Yield
The correlation between soil N2O emission flux and soil properties is shown in Figure 14. There was a significant linear correlation between soil N2O emission flux and soil temperature when the soil temperature ranged from 13.7 °C to 30.1 °C, indicating that the soil temperature was an important

Correlation Analysis among Irrigation Water Salinity, Soil Properties, N 2 O Emissions and Maize Yield
The correlation between soil N 2 O emission flux and soil properties is shown in Figure 14.
There was a significant linear correlation between soil N 2 O emission flux and soil temperature when the soil temperature ranged from 13.7 • C to 30.1 • C, indicating that the soil temperature was an important factor causing differences in seasonal N 2 O emission flux. The soil water content and soil N 2 O emission fluxes were significantly correlated in 2017 but not in 2018, which is attributed to the heavy precipitation in 2018. N 2 O emission was significantly correlated with the soil water content in intra-row soil, but not with inter-row soil, indicating that the intra-row soil water content was an important factor driving differences in seasonal N 2  The correlation among irrigation water salinity, soil properties, soil N 2 O emission flux and maize yield is shown in Tables 3 and 4. The result indicated that high-salinity brackish water irrigation significantly increased the water content of the soil tillage layer by affecting the crop's absorption of water. High-salinity brackish water also increased the salinity content and pH value, and decreased the saturated hydraulic conductivity of the soil tillage layer, thus accelerating the soil salinization and alkalization process. The combined effect of the above results lead to a decrease in the yield of spring maize. Note: IWS is irrigation water salinity, SWC is soil water content; ** and * represents a significant correlation at the 0.01 and 0.05 levels, respectively. Note: IWS is irrigation water salinity, SWC is soil water content, K sat is saturated soil hydraulic conductivity; ** and * represents a significant correlation at the 0.01 and 0.05 levels, respectively. row soil, but not with inter-row soil, indicating that the intra-row soil water content was an important factor driving differences in seasonal N2O emissions. In 2017 and 2018, soil N2O emissions were significantly correlated with soil NO3 − . The concentration of soil NO3 − was 33.07-51.76 mg·kg −1 and 25.50-39.78 mg·kg −1 when N2O reached its emission peak in 2017 and 2018. Soil N2O emission flux was significantly correlated with soil NO3 − in intra-row soil but not in inter-row soil, indicating that the NO3 -in intra-row soil was the main factor causing differences in seasonal N2O emission.   Note: ** and * represents a significant correlation at the 0.01 and 0.05 levels, respectively.

Soil Properties
Our results are consistent with previous studies, which indicate that irrigation with brackish water had adverse effects on soil properties (Tables 3 and 4) [20,37,38]. These results are attributed to irrigation with high-salinity brackish water changing the soil osmotic potential, making it difficult for plants to absorb water and slowing transpiration [3]. High-salinity brackish water is rich in Na + ( Table 2). The sodium adsorption ratio of irrigation water increased, leading to an increase in soil pH (Figure 7), a decrease in soil hydraulic conductivity, and a decrease in the soil infiltration rate (Figure 9) [39,40]. As a result, soil compaction hindered the movement of water and salt in soil, resulting in an increase in the water and salinity content of the intra-row soil. There was a gradient of water potential between the intra-and inter-row soil under film mulching conditions ( Figure 5). As a result of this gradient, salt migrated with water from intra-row soil to inter-row soil. This explains why brackish water irrigation increased the soil water content, salinity, and pH value in inter-row soil.

Soil N 2 O Emission Dynamics
Nitrification and denitrification, in which soil microorganisms and enzymes participate, are the main pathways for N 2 O emission in farmland soil [13]. Our result is consistent with the conclusion of Scheer et al. [41], who indicated that soil N 2 O emissions increased significantly when fertilization and irrigation were simultaneously applied. In our study, the soil N 2 O emission peaks occurred one to five days after simultaneous irrigation and fertilization; after peaking, N 2 O emissions gradually returned to normal levels, i.e., the level before fertilization. This may be due the significant linear correlation observed between N 2 O emission flux and soil NO 3 − (Figure 14). Sufficient substrates were provided for soil nitrification and denitrification after topdressing urea, which promotes soil N 2 O emission.
Our results also show that the soil N 2 O emission dynamics were significantly correlated with the soil water content and soil temperature ( Figure 14). This finding is consistent with Chen et al.'s conclusion from their greenhouse tomato irrigation experiment, which indicated that the soil water content and soil temperature were the dominant factors driving the soil N 2 O emission dynamics when fertilizer was not applied [42].

Effect of Irrigation Water Salinity on N 2 O Emissions
Our study shows that irrigation water salinity had no effect on soil NO 3 − content (Figure 8), indicating that soil nitrification might not be affected by irrigation water salinity. We also found that the soil N 2 O emission increased with the increase in irrigation water salinity. Although there was no correlation between the seasonal N 2 O emission flux and seasonal soil EC 1:5 , the average N 2 O emission flux had a significant positive correlation with the average soil EC 1:5 (Tables 3 and 4), which is contrary to the correlation between N 2 O emission and soil NO 3 − content. This indicates that the effect of salinity on N 2 O emission was durative rather than explosive. Although opposing results have been reported in some other papers, the discrepancy might be due to a threshold soil salinity level that directed the influence on nitrification and denitrification rates (EC 1:5 = 1130 µS·cm −1 ) [43]. When the soil salinity was lower than the threshold, the increased nitrification and denitrification rates were attributed to the increase in soil salinity, whereas the nitrification and denitrification rates decreased when the soil salinity was higher than the threshold [43]. However, this value was higher than the soil EC 1:5 measurements in our experiment (Figure 6), indicating that salinity might increase the nitrification and denitrification rate. The increase in soil salinity might also result in reduced N 2 O reductase activity, resulting in the promotion of N 2 O accumulation in the denitrification process [44]. Because Na + , K + , and higher pH values have stronger salting-out potential, the solubility of N 2 O in soil solution is expected to decrease after irrigation with high-salinity brackish water [45]. These processes could be a pathway by which salinity promotes N 2 O emissions.
Because crop water uptake was restricted by irrigation water salinity, the water content in the soil tillage layer was significantly improved by using high-salinity brackish water irrigation compared with other treatments in the actual irrigation process ( Figure 5). There was a positive correlation between soil N 2 O emission flux and soil water content (Tables 3 and 4, and Figure 14). This agrees with the findings of previous studies [12,18,34,36,46], which reported that soil N 2 O emission flux increased as the soil's saturated water content increased from 30% to 90%. This occurs because increased water content reduces the oxygen concentration and thereby induces soil nitrification and denitrification and thus the production of N 2 O. This may be another pathway through which salinity promotes N 2 O emission. In conclusion, 2.0 g·L −1 brackish water irrigation has the potential to significantly reduce soil N 2 O emission while saving freshwater resources.
Wang et al. observed the effect of alternating brackish water and freshwater irrigation on soil N 2 O emission under the same experimental conditions as ours (Table 5) [26]. Their results indicated that irrigation water salinity and the alternating irrigation method had no significant effect on soil N 2 O emission. Although this is contrary to our result, the reason may be that alternating irrigation reduces the effect of salinity on soil physical properties. Comparing this with our results, alternating brackish water and freshwater irrigation can potentially reduce soil N 2 O emission.

Effect of Different Location on N 2 O Emissions
The proportion of uncovered area in the farmland system cannot be neglected under film mulching conditions. Until now, research was lacking on N 2 O emissions at different locations under film mulching conditions. Our results demonstrate a positive correlation between soil N 2 O emission and soil NO 3 − content in intra-row soil, but there was no correlation with soil NO 3 − content in inter-row soil ( Figure 14).
After simultaneous irrigation and fertilization, the N 2 O emission flux in intra-and inter-row soil of each treatment was improved to different extents, and this may have been caused by the horizontal diffusion of N 2 O [47]. The average N 2 O emission flux in intra-and inter-row soil increased as the irrigation water salinity increased, which might also be the result of the interaction between soil water and salinity (Tables 3 and 4). The result also indicates that soil N 2 O emission in intra-row soil was larger than that in inter-row soil. This phenomenon may be ascribed to the fertilization position. In drip irrigation treatments, fertilizers and water can be applied to the root zone directly through drippers, causing intra-row soil to have better water and fertilizer conditions than inter-row soil.

Effect of Irrigation Water Salinity on Spring Maize Yield
Various studies have been conducted on the impact of irrigation water salinity on crop yield [37,38,48,49]. Zhu et al. and Feng et al. reported that the maize yield decreased with the increase in irrigation water salinity [37,49]. Our results show the same phenomenon. Compared with T1 treatment, the yield under T2-T4 treatments decreased by 7.60-23.93% and 13.86-31.47% in 2017 and 2018, respectively. This may be caused by the relatively poor resistance of maize to Na + during the rapid growth stage [48]. The increasing irrigation water salinity caused an increase in Na + in the soil, and irrigation with high-salinity water was found to likely lead to ion toxicity due to Na + osmotic stress, resulting in a reduction in maize yield [50]. However, there was no significant difference between T1 and T2 treatments in 2017, indicating that it is feasible to use 2.0 g·L −1 brackish water for irrigation in the short term when freshwater resources are in short supply.
Wang et al. also observed the effect of alternating irrigation with brackish water and freshwater on spring maize yield under the same experimental conditions as ours (Table 6) [26]. Comparing their results with ours, we surmise that the yield of spring maize could improve by alternating irrigation using brackish water and freshwater compared with the yield when using brackish water irrigation alone. Alternating irrigation can reduce the accumulation of soil salinity and ion toxicity, thus increasing the yield. This means that alternating irrigation with brackish water and freshwater is another way to use brackish water resources in areas with scarce freshwater resources. Note: 1:1 means to irrigate with freshwater once after one instance of brackish water irrigation; 1:2 means to irrigate with freshwater once after two instances of brackish water irrigation.

Conclusions
Irrigation with 3.5-5.0 g·L −1 brackish water promoted the accumulation of soil salt, increased the soil pH value, and reduced the soil water conductivity. By affecting the water absorption of crops, the soil water content increased after irrigation with 5.0 g·L −1 brackish water. Irrigation with high-salinity brackish water accelerated the degree of soil salinization and alkalization. On this basis, due to the interaction between soil water and soil salinity, soil N 2 O emissions significantly increased by irrigation with 5.0 g·L −1 brackish water. Irrigation with 3.5-5.0 g·L −1 brackish water significantly reduced the yield of spring maize.
Compared with continuous irrigation with 3.5-5.0 g·L −1 brackish water, continuous irrigation with 2.0 g·L −1 brackish water or alternating irrigation with brackish water and freshwater could slow the degree of soil salinization and alkalization, reduce soil N 2 O emissions, and increase maize yield. Thus, these are potential measures to solving the current agricultural water crisis in Hetao Irrigation District. These results can also provide references for areas considering the use of brackish water.
Funding: This research was funded by the National Natural Science Foundation of China, grant numbers 51679239.