Responses of Soil N₂O and CO₂ Emissions and Their Global Warming Potentials to Irrigation Water Salinity

Abstract

Irrigation using marginal quality water (brackish, saline, or treated wastewater, with a salinity of 2–8 g L⁻¹) instead of fresh water alters the soil carbon and nitrogen cycle, and thus, soil greenhouse gas emissions. To reveal the responses of soil nitrous oxide (N₂O) and carbon dioxide (CO₂) emissions and their global warming potentials (GWPs) to irrigation water salinity, a pot experiment was conducted at three levels (2, 5, and 8 g L⁻¹, namely S2, S5, and S8). The results show that the cumulative soil CO₂ emissions were reduced with increases in the irrigation water salinity and were 11.6–28.1% lower than those from the fresh water-irrigated treatment (CK). The cumulative N₂O emissions from S2 and S8 decreased by 22.7% and 39.6% (p < 0.05), respectively, in comparison to CK, whereas those from S5 increased by 87.7% (p < 0.05). The cumulative GWPs from S2 and S8 were 19.6% and 44.1% lower than those from CK, while those from S5 were significantly higher (p < 0.05). These findings indicate that reducing the salinity of brackish water from 5 to 2 g L⁻¹ before using it for irrigation is a potential strategy to mitigate soil GHGs and solve water resource scarcity. The response of soil greenhouse gas (GHG) emissions to salinity may be significantly different among irrigation water salinity ranges. The results have an important guiding significance for exploring greenhouse gas emission reduction measures, and sustainable utilization models of water and soil resources.

1. Introduction

Marginal quality water (e.g., brackish water (2–5 g L⁻¹), marginal saline water (5–8 g L⁻¹), or treated wastewater) is increasingly used for irrigation due to water scarcity in arid and semi-arid regions of the world [1,2]. It is generally high in salinity and sometimes in sodicity, and thus, potentially induces soil salinization [3,4], which not only inhibits plant growth and production [5] but also affects carbon (C) and nitrogen (N) transformation [6,7], as well as soil microbial activity [8]. Nitrous oxide (N₂O) and carbon dioxide (CO₂) are the byproducts of soil N and C transformation and two important atmospheric trace gases, contributing greatly to global climate change [9,10,11,12]. Therefore, understanding the response of soil C and N transformation, especially the greenhouse gas (GHG) emissions associated with this process, to irrigation water salinity is essential for the reasonable use of marginal quality water for agricultural production.

Soil C and N transformation, i.e., GHG emissions, are influenced by various environmental factors [13,14,15,16], including soil salinity [12], soil nutrient availability, moisture, temperature [15], microbial activity [14,16], and so on. Among these, soil moisture is considered to play a dominant role in the temporal process of GHG emissions [15,17,18,19]. When soil is irrigated with brackish or marginal saline water, the salt accumulation will inevitably lead to changes in the biochemical and microbiological processes [2], and ultimately affect N₂O and CO₂ emissions [20]. In recent years, some researchers have investigated the impact of irrigation water salinity on agricultural GHG emissions [2,12,20], but sometimes their results have been contradictory. For example, Maucieri et al. [2] concluded that CO₂ fluxes were reduced by 27.6% and N₂O increased by 59.5%, with irrigation water salinity levels from 0 to 10 dS m⁻¹ in incubated soil columns. Zhang et al. [20] found that increasing irrigation water salinity significantly increased the N₂O emissions from a drip-irrigated cotton field. Meanwhile, Kontopoulou et al. [12] indicated that irrigation water salinity had no significant effect on the N₂O and CO₂ emissions from a bean field. Even so, there is little information on the response of N₂O and CO₂ emissions and their GWP to irrigation water salinity (2–8 g L⁻¹), and there is poor knowledge about GHG emissions and irrigation water salinity, which also limits the ability to predict the GWP of N₂O and CO₂ from irrigated soils with marginal quality water. Therefore, more research is required to utilize two important resources—water and soil.

Assuming that irrigation water salinity within a specific range could reduce soil N₂O and CO₂ emissions and their GWPs, a pot incubation experiment was conducted at three salinity levels (2, 5, and 8 g L⁻¹). The soil moisture, electrical conductivity (EC), and N₂O and CO₂ emissions were monitored. The objectives of this study were to: (1) quantify the N₂O and CO₂ fluxes and their GWPs from soils irrigated with water at different salinity levels; (2) reveal the responses of N₂O and CO₂ to various irrigation water salinity levels; and (3) discuss the implication of the reasonable utilization of marginal quality water for irrigation.

2. Materials and Methods

2.1. Experimental Design

The experiments were conducted in silty clay soil in 2016 at the State Key Laboratory of Hydrology—Water Resources and Hydraulic Engineering in Nanjing (32° 03’ N, 118° 46’ E), China. In May 2016, soil samples were air-dried, ground, sieved, homogenized, and uniformly packed into cylindrical soil columns (internal diameter (I.D.) = 0.3 m, height (H) = 0.5 m) to a bulk density of 1.25 g cm⁻³ for a 0–0.05 m layer and 1.33 g cm⁻³ for a 0.05–0.50 m layer, according to the local soil bulk density. The soil columns were left undisturbed for more than 60 days to avoid the influence of soil disturbance on GHG emissions. The soil columns were placed outdoors in a completely randomized pattern and covered by an awning on rainy days only to avoid the influence of rainfall.

Detailed soil physical and chemical properties can be found in Wei et al. [21,22]. The soil organic matter (OM), total nitrogen (TN), and total phosphorus (TP) in the soil samples were 25.2, 1.4, and 0.04 g kg⁻¹, respectively. The soil ammonium nitrogen (NH₄⁺-N) and nitrate nitrogen (NO₃⁻-N) were 8.8 and 34.2 mg kg⁻¹. The soil pH_H2O was 6.2, and the soil EC was 0.4 ds m⁻¹. Among these, the soil OM was determined by the potassium dichromate concentrated sulfuric acid oxidation method [23]. The soil TN was determined by the modified Kjeldahl method [24]. The soil TP was determined by the alkali fusion-Mo-Sb anti-spectrophotometric method [25]. The soil NH₄⁺-N and NO₃⁻-N concentrations were determined by using the colorimetric method [26,27]. The soil pH_H2O and electrical conductivity (EC) were measured by using a pH meter and an electrical conductivity meter. The soil samples (10 g) were air-dried, sieved (2 mm and 1 mm for the pH and EC, respectively), mixed with deionized water (soil water = 1:2.5 and 1:5 for the pH and EC), and then stirred (3 min) and left standing (30 min). The supernatant was extracted, and finally, the electrical conductivity (EC) was measured using a pH meter (PHS-3E, Shanghai, China, accuracy of 0.01) and electrical conductivity meter (TES-1381K, Taiwan, China, accuracy of 0.01 dS m⁻¹).

To investigate the impacts of the salinity of the marginal quality water (e.g., brackish water or marginal saline water) on the soil greenhouse gas (GHG) emissions, the lower limit (2 g L⁻¹) and upper limit (5 g L⁻¹) for the brackish water, as well as the upper limit (8 g L⁻¹) for the marginal saline water were selected as the three salinity levels (namely S2, S5, and S8, with EC values of 3.60, 8.13 and 12.66 dS m⁻¹, respectively). The saline water was produced by adding NaCl and CaCl₂ at a mass ratio of 1:1 to fresh water. Fresh water samples with no salt added were included as a control treatment (CK) in the experiment. In the current study, deionized water was applied as fresh water to produce saline water, and its indicators were as follows: the TN, TP, NH₄⁺-N, and NO₃⁻-N were 0.18, 0.01, 0.10, and 0.03 mg L⁻¹, respectively; the soil pH_H2O was 6.94; the EC was 0.34 dS m⁻¹. All indicators reached class I of “water quality standards in the environmental quality standard for surface water (GB 3838-2002, PRC)” [28], and both the saline water and fresh water could be used without harmful effect. Each treatment was replicated six times (three for gas sampling and the other three for soil sampling). For each treatment (initial soil moisture contents of about 11.2% WFPS), 3530 and 3210 mL of water were added by using surface irrigation to replenish the soil moisture at a depth of 0–20 cm to the field capacity for the first and second irrigation, respectively.

2.2. Gas Sampling and Greenhouse Gas Emissions

Gas samples from each soil column were collected at intervals of 2 days by using the static chamber (H = 30 cm, I.D. = 30 cm, Figure 1). The sampling chamber was 8 mm polyvinyl chloride (PVC), equipped with a thermometer and an electric fan for measuring the air temperature and mixing air. Moreover, it was wrapped with a heat insulation sponge to minimize the effect of inner air heating by solar radiation. A rubber tube was inserted into the top of the chamber and connected to a three-way stopcock, which was used for gas sampling. Four gas samples were collected from each chamber using a 10 mL syringe at intervals of 10 min between 9:45 and 10:15 h on each sampling day [24]. When gas samples were collected, soil temperatures at the 0, 5, 10, and 15 cm layers were measured using mercury thermometers (accuracy of 0.1 °C).

The N₂O and CO₂ concentrations were analyzed immediately (within 48 h) after sampling by using a gas chromatograph system (Agilent 7890A, Agilent Technologies Inc., California, USA, accuracies of N₂O and CO₂ were 0.1 and 2.0 ppb, respectively). A thermal conductivity detector (TCD) and an electron capture detector (ECD) were used for measuring the CO₂ and N₂O concentrations, respectively. The gas fluxes were calculated according to the slope of the linear regression between the gas concentration and time using Equation (1), as described by Hou et al. [29].

$$F=\rho \cdot h\cdot \frac{273}{273+T}\cdot \frac{dC}{dt}$$

(1)

where F is the gas emission flux (μg m⁻² h⁻¹ for N₂O; mg m⁻² h⁻¹ for CO₂); ρ is the gas density in the standard state (1.973 kg m⁻³ for both N₂O and CO₂); h is the effective height of the static chamber above the soil surface (m, h = 0.3 m); T is the mean air temperature inside the chamber during the sampling period (°C); and dC/dt is the linear increase rate of the gas concentration over time (µL m⁻³ h⁻¹ for N₂O; mL m⁻² h⁻¹ for CO₂).

2.3. Global Warming Potentials (GWPs)

The GWP is an index defined as the cumulative radiative forcing between the present and a later time, caused by gas emitted per unit mass. This index is used to compare the effectiveness of greenhouse gas to trap heat in the atmosphere relative to some standard gas (e.g., CO₂). The GWP for N₂O was 265 when the GWP value for CO₂ was set as 1. The GWP (equivalent to the total equivalent CO₂ emissions) of each soil column was calculated using Equation (2) [11]:

$$GWP=\mathsf{C}{\mathsf{O}}_2+265*{\mathsf{N}}_2\mathsf{O}$$

(2)

The cumulative emissions of N₂O and CO₂ and their GWP were calculated by integrating the daily values along the sampling period.

2.4. Soil Moisture Measurement

Soil samples (about 10–20 g each) from depths of 0, 5, 10, and 15 cm were collected at 10:00 h on each sampling day using a stainless steel sampler (I.D. = 10 mm). The soil moisture content was determined gravimetrically by drying at 105 °C for 24 h. The water-filled pore space (WFPS) was adopted to describe the soil moisture, which was calculated as the ratio of the soil volumetric moisture content (calculated by multiplying the gravimetric soil moisture content with the bulk density) to the soil porosity, with the total porosity calculated from the soil bulk density and particle density (2.65 g cm⁻³) [30].

2.5. Data Analysis and Statistics

Fisher’s least significant difference (LSD) test was used to detect the differences in the gas (N₂O and CO₂) fluxes and their GWPs among treatments at a significance level of p = 0.05 [31]. Statistical analyses of the data were performed using SPSS19.0 (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Soil Environmental Variables

Soil moisture at depths of 0–15 cm showed similar patterns and fell within the same ranges for the treatments irrigated with either saline water or fresh water (Figure 2). Generally, the soil moisture varied significantly in the early period, and then decreased until the following irrigation. Over the whole observing period, the WFPS at depths of 0–15 cm varied in the ranges of 10–93%, 14–93%, 13–94%, and 16–95% for the CK, S2, S5, and S8 treatments, with averages of 50%, 52%, 53%, and 54%, respectively (

Table 1. Soil result values at depths of 0–15 cm.

Treatment	EC	WFPS	T	pH
	(dS m−1)	(%)	(°C)	(1)
S2	0.6 ± 0.0 b	52 ± 4 a	33 ± 1 a	6.9 ± 0.0 a
	(0.3–0.9)	(14–93)	(24–36)	(6.5–7.0)
S5	1.0 ± 0.1 c	53 ± 5 a	34 ± 1 a	6.9 ± 0.0 a
	(0.5–1.4)	(13–94)	(24–37)	(6.5–7.0)
S8	1.5 ± 0.1 d	54 ± 4 a	34 ± 1 a	6.9 ± 0.0 a
	(0.9–2.2)	(16–95)	(24–37)	(6.6–7.0)
CK	0.1 ± 0.0 a	50 ± 3 a	34 ± 1 a	6.4 ± 0.0 a
	(0.1–0.2)	(10–93)	(24–37)	(6.3–6.4)

EC, WFPS, T, and pH represent electrical conductivity, water-filled pore space, soil temperature, and potential of hydrogen, respectively. S2, S5, and S8 represent saline water irrigation at 2, 5, and 8 g L⁻¹; CK represents fresh water irrigation. The numbers outside brackets represent averages of soil result values (EC, WFPS, T, and pH) at depths of 0–15 cm over the whole observation period. The numbers between brackets represent ranges of soil result values (EC, WFPS, T, and pH) at depths of 0–15 cm (0, 5, 10, and 15 cm) in different observation periods. Different letters (^a,b,c,d) in each column represent significant differences at the 5% level.

). Among the different depths, the soil WFPS varied more significantly in the topsoil (0 and 5 cm) than at deeper depths (10 and 15 cm).

The soil EC values were found to be significantly higher with saline water than those with fresh water. Specifically, the soil EC increased remarkably with water salinity from 2 to 8 g L⁻¹. Among these, the soil EC levels from S2 (average of 0.6 dS m⁻¹) were 40% and 60% lower than those from S5 and S8 (p < 0.05). When it comes to the pH and soil temperature, no significant differences were found between the irrigation with marginal saline water and that with fresh water (Table 1).

3.2. Gas Emissions

3.2.1. N₂O Emissions

Two pulse N₂O emissions caused either by saline water or fresh water were observed in the current study (Figure 3). For the saline water-irrigated treatments, pulse N₂O emissions were observed in the 3 days immediately following the first irrigation, while the peak N₂O fluxes from S5 were 57.5% and 179.9% higher than those from S2 and S8 (p < 0.05). During the whole observation period, the cumulative N₂O emissions were mainly in the range of 586–1822 g m⁻² (

Table 2. The cumulative N₂O and CO₂ emissions and their integrative GWPs for each treatment.

Gas	Treatment				Relative Change to CK (%)
	S2	S5	S8	CK	S2	S5	S8
N2O	751 ± 126 b	1822 ± 141 a	586 ± 120 c	971 ± 116 b	−22.7 ns	87.7 **	−39.6 *
(mg N2O m−2)
CO2	240 ± 13 a	218 ± 16 a	209 ± 15 a	268 ± 17 a	−11.6 ns	−23.2 ns	−28.1 ns
(g CO2 m−2)
GWP	439 ± 46 b	700 ± 53 a	365 ± 47 c	525 ± 48 b	−19.6 ns	33.3 *	−44.1 *
(g CO2 m−2)

The symbols of * and ** indicate significant differences at the p < 0.05 and p < 0.01 levels, whereas ns represents no significance. S2, S5, and S8 represent saline water irrigation at 2, 5, and 8 g L⁻¹; CK represents fresh water irrigation. Different letters (^a,b,c) in each column represent significant differences at the 5% level.

), of which the values of S2 and S8 decreased by 22.7% and 39.6% (p < 0.05) compared to CK, while that of S5 improved by 87.7% (p < 0.05).

3.2.2. CO₂ Emissions

The soil respiration from saline water-irrigated soil showed a pattern distinct from that of N₂O emissions. Meanwhile, similar patterns were observed for CO₂ emissions between the saline water and fresh water, that is, only one pulse of CO₂ emissions (911, 777, 860, and 1091 mg CO₂ m⁻² h⁻¹ for CK, S2, S5, and S8, respectively) was observed at 3 days after the initial irrigation (Figure 4). Over the whole experimental period, the CO₂ fluxes mainly ranged from 42 to 1091 mg CO₂ m⁻² h⁻¹. Generally, the CO₂ fluxes were reduced with water salinity from 2 to 8 g L⁻¹, and the values in the saline water-irrigated soil were lower than that in the fresh water-irrigated soil. Consequently, the cumulative CO₂ emissions from S2, S5, and S8 decreased by 11.6%, 23.2%, and 28.1%, respectively, in comparison to CK (Table 2).

3.3. GWPs of N₂O and CO₂ and Their Contributions

The daily GWPs of the N₂O and CO₂ emissions from the saline water- or fresh water-irrigated treatments varied with similar temporal patterns, with maximum values of 2871–6407 mg CO₂ m⁻² h⁻¹ observed 3 days after the first irrigation (Figure 5). For the saline water, the daily GWPs were primarily in the ranges of 135–4666, 217–6407, and 115–2871 mg CO₂ m⁻² h⁻¹ for S2, S5, and S8. Compared to CK, the cumulative GWPs from S2 and S8 decreased by 19.6% and 44.1 (p < 0.05), while that from S5 was enhanced by 33.3% (Table 2). Since no significant differences were found in the soil moisture and temperature among the treatments (Table 1), these results suggest that the irrigation water salinity might be a critical factor in the variation of the soil GWPs among treatments.

As shown in Figure 6a, the N₂O contribution to the GWP was high immediately after the irrigation events. However, for CO₂, its contribution to the GWP seems to have been the opposite, and the proportions (averages of 52.0–75.8%) were greater than that of N₂O (averages of 24.2–48.0%). Moreover, the contributions of N₂O or CO₂ to the GWPs were also different among the irrigation water salinity levels (Figure 6b). Compared to CK, 2 g L⁻¹ of saline water irrigation elevated the contributions of the N₂O and CO₂ by 13.1% and 2.8% in a pulse period but decreased the N₂O contribution by 16.8% in a non-pulse period. When it comes to 5 g L⁻¹ of saline water irrigation, the N₂O contribution increased by 34.2% and −14.2% in the pulse and non-pulse periods, and the CO₂ contribution decreased by 8.4% and 11.6%, respectively. These results indicate that irrigation water salinity played an important role in partitioning the contribution of either CO₂ or N₂O to the cumulative GWPs between the pulse and non-pulse periods. Additionally, the differences in the contributions of the N₂O or CO₂ to the GWPs between pulse (when the soil moisture is generally high) and non-pulse periods suggest that the soil moisture condition affected the response of the soil microbial to salinity, and the impact degree may have fluctuated among the irrigation water salinity ranges.

3.4. Relations between GHGs Fluxes and Soil EC or Soil Temperature

Either GHG (N₂O and CO₂) emissions or their GWPs were reduced exponentially (y = a exp (-bx)) with an increase in the irrigation water salinity. For CO₂, no significant relationships were found with the soil EC at almost all soil depths. Meanwhile, the N₂O emissions from the saline water-irrigated treatments were exponentially correlated with soil EC at a 5 cm depth or a 0–15 cm average at a significance level of p < 0.01. Consequently, the GWPs of the N₂O and CO₂ emissions from S2 and S5 were significantly (p < 0.05) related to the soil EC at 5 cm depths or a 0–15 cm average. On the whole, the determination coefficients (R²) between the soil EC and N₂O flux were higher than those between the soil EC and CO₂ flux or the GWPs (

Table 3. Coefficient of determination for the relationships between GHG (N₂O, CO₂) fluxes or their GWP and soil EC or temperature at depths of 0–15 cm.

Items	Treatment (n = 15)	Coefficient of Determination (R2)
		EC					Temperature
		0	5	10	15	0–15	0	5	10	15	0–15
N2O	S2	0.33 *	0.60 **	0.10	0.30 *	0.58 **	0.03	0.00	0.00	0.01	0.00
	S5	0.16	0.63 **	0.03	0.13	0.55 **	0.14	0.00	0.00	0.01	0.00
	S8	0.12	0.68 **	0.12	0.25	0.61 **	0.28	0.02	0.00	0.00	0.02
	CK	0.05	0.07	0.17	0.09	0.15	0.10	0.00	0.01	0.01	0.00
CO2	S2	0.28 *	0.09	0.11	0.29 *	0.15	0.30 *	0.43 **	0.41 **	0.45 **	0.43 **
	S5	0.26	0.02	0.08	0.30 *	0.06	0.40 *	0.66 **	0.69 **	0.70 **	0.67 **
	S8	0.15	0.03	0.18	0.29 *	0.06	0.37 *	0.64 **	0.64 **	0.63 **	0.62 **
	CK	0.04	0.02	0.00	0.08	0.01	0.46 **	0.68 **	0.69 **	0.69 **	0.68 **
GWP	S2	0.32 *	0.34 *	0.09	0.33 *	0.32 *	0.03	0.11	0.15	0.18	0.13
	S5	0.19	0.40 *	0.02	0.18	0.38 *	0.01	0.06	0.11	0.13	0.06
	S8	0.09	0.22	0.17	0.30 *	0.24	0.00	0.18	0.25	0.28	0.17
	CK	0.05	0.07	0.13	0.1	0.12	0.01	0.15	0.21	0.21	0.14

The symbols of * and ** indicate correlation is significant at p < 0.05 and p < 0.01, respectively. S2, S5, and S8 represent saline water irrigation at 2, 5, and 8 g L⁻¹; CK represents fresh water irrigation.

). This implies that the soil salinity may have had a greater effect on the soil nitrification or de-nitrification processes than the soil respiratory activity.

Both the GHG (N₂O and CO₂) emissions and their GWPs showed exponential correlations to the soil temperature, and the determination coefficients of CO₂ and its GWPs generally increased with an increase in the soil depth (Table 3). For N₂O and its GWPs, no significant relationships with the soil temperature were found. Meanwhile, the CO₂ emissions from all treatments were significantly related to the soil temperature. The coefficients between the soil temperature and CO₂ flux were greater than those of the soil temperature and N₂O flux or GWP, especially for the CO₂ flux and soil temperature in fresh water. Therefore, this seems to confirm that the soil temperature had a greater impact on the soil respiratory activity than on the nitrification or denitrification processes. The saline water irrigation could be helpful for reducing the influence of temperature on soil respiration.

4. Discussion

4.1. N₂O and CO₂ Respond Differently to Irrigation Water Salinity

Soil microbial activity often decreases with increases in soil salinity due to the osmotic desiccation of microbial cells [8,31,32,33,34]. In the present study, the soil CO₂ flux was reduced with irrigation water salinity from 2 to 8 g L⁻¹. However, no significant differences in the CO₂ fluxes were found among the irrigation water salinity levels (Table 2). Since the soil EC increased remarkably with salinity from 2 to 8 g L⁻¹ and no significant regression relationships with the soil EC were found at almost all layers (Table 3), it can be inferred that salinity might not determine the magnitude of CO₂ emissions from saline water-irrigated soils. When it comes to N₂O, it was quite different from the CO₂. By increasing the water salinity from 2 to 8 g L⁻¹, the soil N₂O fluxes increased first and then decreased (Table 2). The soil N₂O emissions from S5 behaved significantly differently from those from S2 and S8, and a similar phenomenon can also be found when synthesizing the previous results in the previous literature [20,35,36] (

Table 4. Summary of results on the effect of salinity on N₂O emissions.

Type of Test	Salt Type	Salinity Level	Moisture Conditions	N2O Sequence	Reference
A drip-irrigated cotton field experiment	NaCl + CaCl2 (1:1)	0.4 (FW), 4.6 (BW), 8.0 (SW) d (dS m−1 in water)	10–25% (w/w)	FW < BW < SW (N0) a	[20]
				SW < FW < BW (N360) b
A soil microcosm experiment	NaCl	0 (L),1 (M), 5 (H) e (10^-9 g L−1 in water)	Permanent flooding	L > M > H (ns) c	[36]
			Wet–dry cycle	L > M > H (ns)
An incubation experiment	NaCl + NH4Cl	0 (T1), 5.9 (T2), 11.7 (T3), 23.4 (T4), 35.1 (T5) f (g L−1 in soil solution)	65% WHC	T2 > T3 > T5 > T1 > T4 (ns)	[35]
			45%WHC	T2 > T1 > T5 > T3 > T4
	Na2SO4 + (NH4)2SO4		65% WHC	T1 > T2 > T5 > T4 > T3 (ns)
			40%WHC	T1 > T2 > T5 > T3 > T4 (ns)
	Ca(NO3)2	0 (T1), 5.9 (T2), 17.6 (T3), 29.3 (T4), 46.8 (T5) (g L−1 in soil solution)	70%WHC	T2 > T1 > T3 > T4 > T5 (ns)
			50%WHC

^a N0 represents treatment without nitrogen application; ^b N360 represents treatment with 360 kg N ha⁻¹ application. ^c ns means the difference in N₂O fluxes between treatments was not significant. ^d FW, BW, and SW represent saline water irrigation at 0.35, 4.61, and 8.04 dS m⁻¹, respectively. ^e L, M, and H represent saline water irrigation at sodium chloride levels of 0, 1, and 5 ppt. ^f T1-T5 represent soil solutions at 0, 0.1, 0.2, 0.4, and 0.6 M chlorides (NaCl and NH₄Cl) or sulphates (Na₂SO₄ and (NH₄)₂SO₄). WHC means water holding capacity.

). For example, Wei et al. [36] found that the soil N₂O cumulative flux resulting from irrigation with 5.0 g L⁻¹ of water salinity was 51.2–82.9% and 32.4–44.8% higher than that resulting from 1.1–3.5 g L⁻¹ and 1.1–2.0 g L⁻¹ of water salinity in 2017 and 2018, respectively. Inubushi et al. [35] indicated that the N₂O emissions from 5.9 g L⁻¹ of saline water-irrigated soils were greater than those from fresh water- and 11.7 g L⁻¹ of saline water-irrigated soils when the water-holding capacity (WHC) was in the range of 45–65%.

4.2. Effect of Soil Environmental Variables on GHG (N₂O and CO₂) Emissions

It is known that soil GHG (N₂O and CO₂) emissions are simultaneously affected by multiple factors, such as the soil moisture, temperature, salinity, and so on [20,30,37,38,39,40]. Generally, appropriate soil environmental conditions could induce peak GHG (N₂O and CO₂) emissions [40,41,42,43,44,45]. Some previous studies in the literature have researched the suitable scope of the soil WFPS, temperature, or EC corresponding to peak GHG emissions [35,43,45,46], but the conclusions have been somewhat different. For example, Lewczulk et al. [43] indicated that the pulse N₂O and CO₂ emissions from soybean fields took place at about 45% WFPS and 25 °C, and 50% WFPS and 20 °C, respectively. We et al. [45] concluded that the highest peak N₂O and CO₂ fluxes from surface watering soil were observed at 60.2% WFPS and 29.9 °C, and 50.3% WFPS and 33.6 °C, respectively. Cardoso et al. [46] found that the peak N₂O and CO₂ emissions occurred with a water content and temperature near 70–80% WFPS and 20–25 °C. Inubushi et al. [35] showed that higher N₂O emissions from Yellow Soil and Andosol usually occurred when the soil moisture and salinity were around a 40–70% water holding capacity (WHC) and 0–5.9 g L⁻¹ (0–9.49 dS m⁻¹).

In the current study, the soil CO₂ emissions from all treatments showed distinct patterns to that of the N₂O emissions, that is, two pulse N₂O emissions were observed after two irrigation events, while only one peak CO₂ emission was found after the first irrigation (Figure 3 and Figure 4). It may be that the soil-related factors (69–78%WFPS, 31–37 °C, and 0.1–1.1 dS m⁻¹ for the soil moisture, temperature, and EC, respectively) corresponding to the N₂O pulse emissions fell within the optimal range [35,45,46] that favors N₂O production. However, for CO₂, since no significant relationships with the soil moisture or EC were found at almost any depth, the more suitable soil temperature (36–37 °C) may have led to its pulse emission (

Table 5. Coefficient of determination for the exponential relationships between GHG (N₂O, CO₂) fluxes and soil moisture at depths of 0–15 cm.

Gas	Treatment (n = 15)	Coefficient of Determination (R2)
		Depth
		0	5	10	15	0–15
N2O	S2	0.50 **	0.57 **	0.58 **	0.69 **	0.58 **
	S5	0.40 *	0.43 **	0.44 **	0.45 **	0.44 **
	S8	0.57 **	0.66 **	0.59 **	0.62 **	0.62 **
	CK	0.16	0.17	0.09	0.21	0.16
CO2	S2	0.008	0.004	0.002	0.02	0.001
	S5	0.08	0.05	0.02	0.03	0.04
	S8	0.19	0.13	0.16	0.09	0.16
	CK	0.18	0.12	0.16	0.03	0.02

The symbols of * and ** indicate correlation is significant at p < 0.05 and p < 0.01, respectively. S2, S5, and S8 represent saline water irrigation at 2, 5, and 8 g L⁻¹; CK represents fresh water irrigation.

). The average soil temperatures after the second irrigation decreased by 11.5–12.3% for all treatments, which may have led to the reduction in the respiration rate, and thus the lower CO₂ emissions [43,45]. The above reasons may account for the differences in the emission patterns between N₂O and CO₂.

Moreover, the present study shows that GHGs were significantly produced immediately following irrigation either with saline water or fresh water, and then decreased until the following irrigation (Figure 3 and Figure 4). This was possibly due to the strong alternation between drying and wetting after irrigation, which promotes soil microorganism activity and carbon and nitrogen transformation, resulting in higher GHG emissions [29,47]. However, with the decrease in the soil moisture and the consumption of soil matrix (carbon and nitrogen), the microbial activity decreased gradually, and thus the emissions were reduced.

4.3. N₂O Emissions Varied among Irrigation Water Salinity Levels

There is no consistent conclusion on the effect of salinity on soil N₂O emissions in the previous literature (Table 5). Sometimes it increased remarkably with the increase in irrigation water salinity [20], but sometimes it was unaffected [12,35,36]. This implies that the response of N₂O emissions to irrigation water salinity might vary among different soil conditions (such as soil moisture and mineral nitrogen contents). For example, under a 65% water holding capacity (WHC), the N₂O emissions from soils at an 11.7 g L⁻¹ salinity level were greater than that at 35.1 g L⁻¹, while showing the contrary under a 40% WHC [35]. Yet, the response of soil N₂O emissions to irrigation water salinity under permanent flooding was found to be similar to that under wet–dry cycles [36]. Moreover, under a 0 kg N ha⁻¹ nitrogen level, the N₂O flux at an 8.0 dS m⁻¹ irrigation water salinity level was greater than that at 0.4 dS m⁻¹, but it showed the opposite under 360 kg N ha⁻¹ [20]. In the current study, the N₂O fluxes from S2 and S8 were significantly lower than that from CK, while those from S5 were significantly higher (Table 2). Hence, we suspect that the response of the N₂O emissions to irrigation water salinity might have varied among the salinity ranges due to no significant differences in the soil-related factors among the irrigation water salinity levels.

Meanwhile, the soil moisture significantly affects soil aeration and redox state, and ultimately, the nitrification or denitrification process [48,49]. It is generally considered that nitrification is the main pathway of N₂O production in aerobic soil conditions [50] or when soil moisture is less than 70% WFPS [51,52,53]. On the contrary, when the soil moisture is higher than 70% WFPS, denitrification is considered to be the main source of N₂O [54,55]. In our study, the average soil moisture at depths of 0–15 cm for all treatments was higher than 70% WFPS within 3 days after each irrigation event, meaning that denitrification played a dominant role in the N₂O production within this period. Additionally, some previous researchers have concluded that the denitrification rate generally decreases with increases in salinity [56,57,58]. However, the current results show that N₂O emissions that were dominated by denitrification (first 3 days after each irrigation) increased first and then decreased with irrigation water salinity levels from 2 g L⁻¹ to 8 g L⁻¹, which is quite different from the results in the above literature. Yet, this is similar to the results drawn by Zhang et al. [59], who indicated that the relationship between the denitrification rate and salinity was closely related to the incubation period, and soil denitrification activity increased first and then decreased with increases in salinity when the incubation period was 3 days. This phenomenon may be related to the complex denitrification process of saline water irrigation because water salinity not only directly affects denitrifying bacteria and microbial metabolic enzyme activity but also indirectly affects denitrification through pH, conductivity, and soluble organic carbon content. In addition, previous studies on the effect of pH on N₂O production in denitrification have shown that when pH ≤ 5, the denitrification rate and N₂O emissions decreased with increases in the soil pH [60,61]. When the pH was within the range of 5–7, it showed the opposite, that is, the increasing pH promoted N₂O emissions [60]. In the current study, the soil pH increased with the increase in the irrigation water salinity (the initial soil sample was acidic, pH = 6.2), and was mainly at a neutral level (6.9, Table 1). However, the N₂O emissions first increased and then decreased with the increase in the pH (or irrigation water salinity), which is different from the results in the above literature. These results likely indicate that the influence of salinity on N₂O production during denitrification might mask the effect of pH.

Soil moisture lower than 70% WFPS accounted for more than 80% of the experimental period. These moisture levels often favor soil nitrification [29,62], during which ammonia oxidation and nitrite oxidation are two important steps dominating the production of N₂O and are both sensitive to soil salinity [63,64,65]. The current study indicates that 2 g L⁻¹ of saline water irrigation reduced the soil N₂O emissions and 5 g L⁻¹ of saline water irrigation enhanced it compared to CK. Thus, it can be inferred that the nitrogen nitrification steps that participate in N₂O production, namely ammonia oxidation and nitrite oxidation, might have different salinity sensitivity among salinity levels or ranges. Similar results showing that the ammonia bacteria were more sensitive to the NaCl concentration than nitrite oxidizers in solutions when the Cl^- fell within a range of 0–5 g L⁻¹ were found by Moussa et al. [63] in a mixed bacteria solution culture, and it was more sensitive to the NaCl concentration in solutions with 0–5 g L⁻¹ Cl^- than 5–10 g L⁻¹ Cl^-. Ammonia oxidation might be inhibited to a greater degree than nitrite oxidation at low water salinity levels, while it was the opposite for the high salinity case. This may be the main reason for the variations in the N₂O emissions at the 2 and 5 g L⁻¹ irrigation water salinity levels.

4.4. Implication for Safe Use of Saline Water for Irrigation

As an alternative source of fresh water, marginal quality water will be widely used for agricultural irrigation in the near future. The previous literature mainly focused on the effects of marginal quality water on soil salinization, crop growth, yield, and quality. With the development of ecological green agriculture, it is particularly important to study the ecological effects of farmland caused by saline irrigation. The current study mainly researched the effects of irrigation water salinity on soil water and salt dynamics and the GWPs of N₂O and CO₂ emissions. The results show that 2 g L⁻¹ of saline water irrigation could be helpful for reducing N₂O and CO₂ emissions and mitigating their GWPs. Meanwhile, 5 g L⁻¹ of saline water irrigation enhanced their GWPs by significantly increasing the soil N₂O emissions (Table 2). These results suggest that desalinating brackish water can help reduce the environmental impacts of greenhouse gas emissions and achieve cleaner production in crop cultivation and sustainable development of land resources for marginal quality water irrigation.

The current study mainly focused on the impacts of marginal quality water (2–8 g L⁻¹) on soil N₂O, CO₂ emissions, and their GWPs, and the conclusions were drawn based on a short period incubation experiment without plant cultivation, which is an important factor linked to soil N₂O emissions [66]. Several researchers have reported that salinity stress affects seed germination either by decreasing the rate of water uptake (osmotic effect) or by facilitating the intake of toxicity ions, which may change enzymatic or hormonal activities inside the seed [67,68]. Thus, further research should be conducted with more detailed water salinity levels under field conditions with plant cultivation to investigate the effect of salinity on seed germination, crop growth, and GHGs, and explore the marginal water salinity threshold that can reduce soil GHG emissions while also maintaining a higher seed germination percentage and crop yields. Moreover, the results likely indicate that the response of soil greenhouse gas emissions to salinity might be significantly different among irrigation water salinity ranges. Therefore, the impact of irrigation water salinity on the microbiological processes associated with the soil C and N transformation process, by measuring the number of microbial populations, should also be conducted in the future.

5. Conclusions

The N₂O and CO₂ emissions were significantly affected by marginal quality water (2–8 g L⁻¹) irrigation but exhibited different variations among the irrigation water salinity ranges. Compared to CK, 2 and 8 g L⁻¹ of saline water decreased the cumulative N₂O emissions by 22.7% and 39.6% (p < 0.05), whereas 5 g L⁻¹ of saline water enhanced it by 87.7% (p < 0.05). Marginal quality water irrigation limited the soil respiration and resulted in lower CO₂ emissions as compared to CK. Saline irrigation at 2 and 8 g L⁻¹ reduced the integrative GWPs of N₂O and CO₂, while at 5 g L⁻¹, it improved significantly (p < 0.05). The response of soil C and N transformation to salinity, especially the soil GHG emissions associated with this process, might be significantly different among irrigation water salinity ranges. Using desalinated brackish water for irrigation is a promising strategy that can solve water resource scarcity and reduce the short-term integrative greenhouse effect. The results will provide references for exploring greenhouse gas emission reduction measures and sustainable utilization models of water and soil resources.