C3 and C4 Grass Species: Who Can Reduce Soil Nitrous Oxide Emissions in a Continental Arid Region?

In order to relieve grazing pressure, drought-tolerant grass species are widely cultivated in arid regions. However, soil N emission is largely neglected while pursuing forage yield. We carried out a randomized block study to investigate whether and how the C3 and C4 grass species differ in soil N emission in a typical salinized field with temperate continental arid climate in the northwest inland regions, China. We quantified soil N2O flux from two C3 (barley and rye) and two C4 grass species [corngrass and sorghum hybrid sudangrass (SHS)] in fields during the growing season (from May to September) by using the static box method, and then determined the relationships between soil N2O fluxes and forage yield and soil properties. Results show that soil available nitrogen, soil temperature, pH, soil organic carbon, and total nitrogen were correlated, but soil water was anti-correlated with soil N2O fluxes. In addition, N2O flux increased significantly faster with soil temperature in C4 than in C3 grass fields. Although the lower total N2O emission fluxes were detected for C3 species, the lower yield-scaled N2O was detected for C4 species. Our study provided insights into the determination of grass species and the understanding of mechanisms regulating N2O fluxes in C3 and C4 species in the continental arid regions.


Introduction
Crops and livestock are the core components of an agricultural system, and their interaction drives the continuous evolution of agricultural systems. Forage crops sustain more than 70% of sheep and goat, as well as 50% of meat products in the world [1]. In China, more than 60% of cultivated land is located in arid and semi-arid areas [2], where stall-feeding rather than grazing is the main model of livestock production due to the serious seasonal drought [3]. Therefore, in recent years crop forages have been widely planted to meet the increasing demand for meat products and release grazing pressure in these areas. Gramineae grass is one kind of the most important forage crops, and some annual grass species are very productive even under arid conditions [4][5][6]. Compared to C3 species, C4 species are more suitable for growing in an arid region due to shrinking pore diameter and then reducing transpiration and water loss under drought. Land-use in which intensified C3 species are strategically diversified with C4 species may lead to benefits for soil microbiological diversity [7], soil water content and N use efficiencies [8], soil carbon sequestration [9], and other ecosystem services including saline-alkali land improvement [10]. Despite considerable recent research try to better understand these factors, the differences between both species of grass system in term of N 2 O emission

Climate and Soil
The field experiments were performed in 2016 and 2017 at the Linze Grassland Agricultural Trial Station of Lanzhou University, Linze County, Zhangye City and Gansu Province, China. The experimental site is located at latitude 39 • 15 and longitude 100 • 02 E at an elevation of 1390 m above sea level. The northwest inland arid area of the research station is a secondary salinization meadow with temperate continental climate. The average annual temperature is 7.6 • C and the average hour of sunshine is 3042 h/yr. There is a frost-free period of approximately 180 days/year and an average annual mean rainfall of 121 mm yr −1 , with over 60% of the rainfall occurring in summer and fall. The average annual free water evaporation is 2430 mm. The amount of precipitation was 50 mm in 2016 and 87 mm in 2017. The monthly precipitation during the experimental years is shown in Figure 1. The soil in the research site is classified as Aquisalids according to USDA soil taxonomy (salt 0.7-0.9%). Before experiment, the initial soil properties measured at 30 cm soil depth were: soil organ carbon 9.34 g/kg, total N 1.07 g/kg, pH 8.5, bulk density 0. 93 g/cm 3 .

Experimental Design
Two C4 grass species, sorghum hybrid sudangrass (SHS, Sorghum bicolor × S. sudanenese) and corngrass I (Zea mays × Zea mexicana)), and two C3 grass species, barley (Hordeum vulgare L.) and rye (Secale cereale L.) were chosen to establish plots for measuring N2O emission and forage yield. This study was conducted with a randomized complete block design with three replications (plots) for each grass species. The length and width of each plot were 15 m and 6 m, respectively (area = 90 m 2 ), and a 1.5-m-wide isolation belt and a ridge were placed between each plot to prevent nutrient and water leakage. Traditional flat planting of 50 cm in width was used for the two C4 species with different seeding spacing (25 cm for corngrass and 20 cm for SHS) and evaluated density (82,500 plants/ha corngrass and 10,500 plants/ha for SHS). Broadcasting sowing was used for barley and rye with a seeding rate of 375 kg/ha and 300 kg/ha, respectively.
Forages were sown on May 20 in 2016 and May 24 in 2017 due to the local unstable weather condition (e.g., air temperature declining suddenly) in spring and no cold tolerance of chosen C4 species, using a hole-sowing machine with a seeding depth of 3-4 cm and manual broadcasting sower. A base fertilizer containing 150 kg/ha P2O5 was spread evenly over the furrow. Irrigation (about 120 mm) was applied twice during growing season ( Table 1). The corngrass and SHS were harvested three times per year at the stubble height of 15 cm, while the plant heights were 100-130 cm and 130-160 cm for corngrass and SHS, respectively. Barley and rye also were harvested three times per year at the stubble height of 5 cm, while the plant grew to 25-30 cm. The schedule for harvests was listed in Table 1. Weeds were controlled manually during each growing season. Monthly precipitation irrigation Monthly mean air temperature

Experimental Design
Two C4 grass species, sorghum hybrid sudangrass (SHS, Sorghum bicolor × S. sudanenese) and corngrass I (Zea mays × Zea mexicana), and two C3 grass species, barley (Hordeum vulgare L.) and rye (Secale cereale L.) were chosen to establish plots for measuring N 2 O emission and forage yield. This study was conducted with a randomized complete block design with three replications (plots) for each grass species. The length and width of each plot were 15 m and 6 m, respectively (area = 90 m 2 ), and a 1.5-m-wide isolation belt and a ridge were placed between each plot to prevent nutrient and water leakage. Traditional flat planting of 50 cm in width was used for the two C4 species with different seeding spacing (25 cm for corngrass and 20 cm for SHS) and evaluated density (82,500 plants/ha corngrass and 10,500 plants/ha for SHS). Broadcasting sowing was used for barley and rye with a seeding rate of 375 kg/ha and 300 kg/ha, respectively.
Forages were sown on 20 May 2016 and 24 May 2017 due to the local unstable weather condition (e.g., air temperature declining suddenly) in spring and no cold tolerance of chosen C4 species, using a hole-sowing machine with a seeding depth of 3-4 cm and manual broadcasting sower. A base fertilizer containing 150 kg/ha P 2 O 5 was spread evenly over the furrow. Irrigation (about 120 mm) was applied twice during growing season ( Table 1). The corngrass and SHS were harvested three times per year at the stubble height of 15 cm, while the plant heights were 100-130 cm and 130-160 cm for corngrass and SHS, respectively. Barley and rye also were harvested three times per year at the stubble height of 5 cm, while the plant grew to 25-30 cm. The schedule for harvests was listed in Table 1. Weeds were controlled manually during each growing season. Three successive sunny days avoiding the irrigation and harvest events in each month during growing season were chosen for measuring the N 2 O emission, and the specific times were listed in Table 1.
Measurement of N 2 O was carried out once at 9:00-11:00 of each chosen day in each plot. A static opaque chamber was used to sample N 2 O in each plot. The static opaque chamber (30 cm × 30 cm × 30 cm) was constructed from stainless steel and sheathed with 2-cm-thick foam plastic for improving temperature stability. It was fitted with an internal battery-operated fan to mix the air and with a silica gel catheter (2 mm diameter × 200 mm length) on the top of chamber for gas sampling. Gas samples were drawn through a three-way stopcock, using a 50-mL plastic syringe, and then transferred into 300-mL aluminum foil gas-collecting bags. For each sampling event, four gas samples of approximately 300 mL were taken at time interval of 10 min (i.e., 0, 10, 20, and 30 min). The chamber was also equipped with an electronic thermometer. The temperature inside the chamber was recorded during gas sampling and applied to calculate gas flux; soil temperature (ST) was also measured by a mercurial thermometer inserted 5 cm into the soil at the sampling site before and after gas sampling and the mean temperature was applied to detect its effect on GHG emissions.
Samples were brought back to the laboratory, and gas concentration was measured within 24 h by a N 2 O analyzer (Model No. 908-0015-0000, Los Gatos Research, San Jose, CA, USA). The exchange flux of N 2 O describes the change of gases in unit time in the sampling box, which was calculated according to Liu et al. [31] and Ning et al. [32]. The total N 2 O flux in growing season was calculated according to Ning et al. (2020) [32]: total N 2 O flux = mean daily N 2 O flux × test days. N 2 O emission intensity (NEI) measuring the N 2 O flux per unit hay yield (NEI hay ) and per unit crude protein yield (NEI CP ) was estimated according to Dyer et al. [33].

Forage Yield and Soil Property
To determine the forage yield, three quadrats (50 cm × 50 cm) were set up in each plot for each C3 species, and eight typical plants of each C4 species were randomly selected in each plot, and then these fresh grasses mowed were oven dry for a minimum of 48 h. After weighting, the forage samples were smashed and extracted by passing through a 0.25-mm sieve, then the CP was measured by Kjeldahl method [34].
Three sites in each plot were randomly selected for collecting soil at a 0-15 cm depth using the bucket auger (5 cm diameter) after each N 2 O collection. The soil water content (SWC) was estimated as: (original wet weight − soil dry weight)/original dry weight × 100%. Part of fresh soil was used to measure soil available nitrogen (SAN) (NH 4 + -N and NO 3 − -N) by colorimetric method [35]. The other part of soil samples was air-dried then ground through a 0.25-mm sieve. Soil organic carbon (SOC) was measured by chromic acid REDOX titration [36], soil pH (SpH) was measured by potential method [37], and soil total nitrogen (STN) was determined by following the methods of Bremner and Mulvaney [34].

Statistical Analysis
All other statistical analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC, USA). Results of a Shapiro-Wilk test (UNIVARIATE Procedure) indicated that data collected from this study were normally distributed. As the interactions between year and species were not significantly different, a one-way ANOVA (GLM Procedure) was applied to compare the mean difference in soil N 2 O flux between months or between species (Figures 2 and 3), and that in forage production between species ( Figure 5), with a least significant difference (LSD) test for multiple comparisons. N 2 O flux in relation to each soil property (Figure 4) was determined for each species using a linear regression (GLM Procedure). As there was no significant difference in soil properties and N 2 O flux between 2016 and 2017, data were pooled before analyses. The regression slopes of different species were compared using an analysis of covariance (ANCOVA, GLM Procedure) with a CONTRAST statement for pairwise comparison. As there was no significant difference in slope between the two C3 or C4 species, we also determined the relationships between N 2 O flux and soil properties for the C3 and C4 species (Figure 4; Tables 2 and 3). A multiple variate linear regression was used to test the soil properties affecting the N 2 O flux, while only the significant factors (properties) remained in the final optimal model (Table 4). study were normally distributed. As the interactions between year and species were not significantly different, a one-way ANOVA (GLM Procedure) was applied to compare the mean difference in soil N2O flux between months or between species (Figures 2 and 3), and that in forage production between species (Figure 5), with a least significant difference (LSD) test for multiple comparisons. N2O flux in relation to each soil property (Figure 4) was determined for each species using a linear regression (GLM Procedure). As there was no significant difference in soil properties and N2O flux between 2016 and 2017, data were pooled before analyses. The regression slopes of different species were compared using an analysis of covariance (ANCOVA, GLM Procedure) with a CONTRAST statement for pairwise comparison. As there was no significant difference in slope between the two C3 or C4 species, we also determined the relationships between N2O flux and soil properties for the C3 and C4 species (Figure 4; Tables 2 and 3). A multiple variate linear regression was used to test the soil properties affecting the N2O flux, while only the significant factors (properties) remained in the final optimal model (Table 4).

Soil N2O Emission in Barley, Rye, Corngrass, and SHS
For each species, soil N2O fluxes significantly increased from early growing season (May) to mid-growing season (July), then significantly decreased (P < 0.05) (    The total soil N 2 O quantities during the growing seasons were significantly greater from SHS than from rye in 2016 (P < 0.05) and significantly greater from SHS and corngrass than from rye in 2017 (P < 0.05) (Figure 3). The total soil N 2 O quantity in C3 fields during the growing seasons was 15.08% lower compared to that in C4 fields.
The total soil N2O quantities during the growing seasons were significantly greater from SHS than from rye in 2016 (P < 0.05) and significantly greater from SHS and corngrass than from rye in 2017 (P < 0.05) (Figure 3). The total soil N2O quantity in C3 fields during the growing seasons was 15.08% lower compared to that in C4 fields.

Soil N2O Emission in Relation to Soil Property
Increasing soil temperature, organic carbon, pH, available nitrogen, and total nitrogen was significantly promoted, but increasing soil water content significantly decreased N2O emissions (P < 0.05), except that soil organic carbon did not have a significant effect on N2O emission from barley ( Figure 4, Table 2). For a given soil property, the regression slope was not significantly different between barley and rye or between corngrass and SHS (P > 0.05). The soil temperature, water content, organic carbon, pH, available nitrogen, and total nitrogen respectively explained 69.37%, 76.03%, 22.54%, 70.26%, 74.00%, and 35.55% variation of N2O emission from C3 species and 74.48%, 66.09%, 38.10%, 66.45%, 70.26%, and 18.78% variation of N2O emission from C4 species (Figure 4, Table 3). N2O emission increased significantly faster with increasing ST in C4 fields than in C3 fields (P = 0.0083).

Soil N 2 O Emission in Relation to Soil Property
Increasing soil temperature, organic carbon, pH, available nitrogen, and total nitrogen was significantly promoted, but increasing soil water content significantly decreased N 2 O emissions (P < 0.05), except that soil organic carbon did not have a significant effect on N 2 O emission from barley ( Figure 4, Table 2). For a given soil property, the regression slope was not significantly different between barley and rye or between corngrass and SHS (P > 0.05). The soil temperature, water content, organic carbon, pH, available nitrogen, and total nitrogen respectively explained 69.37%, 76.03%, 22.54%, 70.26%, 74.00%, and 35.55% variation of N 2 O emission from C3 species and 74.48%, 66.09%, 38.10%, 66.45%, 70.26%, and 18.78% variation of N 2 O emission from C4 species (Figure 4, Table 3). N 2 O emission increased significantly faster with increasing ST in C4 fields than in C3 fields (P = 0.0083).

Species
Intercept   Table  2 for each species and in Table 3 for C3 and C4 species.    Table 2 for each species and in Table 3 for C3 and C4 species.
Amount soil properties, soil available nitrogen, soil water content, and soil temperature were the main factors that significantly affected N 2 O fluxes and explained 88.23% variation of N 2 O fluxes in C3 fields (Table 1); while only two factors (i.e., soil available nitrogen and temperature) significantly affected N 2 O flux and explained 82.55% variation of N 2 O flux in C4 field. Soil-available nitrogen accounted for ≥ 70% variation of N 2 O fluxes for both field types (Table 4).

Soil N 2 O Emission and Forage Production
The hay and crude protein yields of SHS were significantly greater than that of corngrass in 2016 and 2017, and both were significantly greater than that of barley and rye (P < 0.05) ( Figure 5A,B). The total hay and crude protein yields of C4 species were, respectively, 1.61 and 1.37 times greater than that of C3 species. The NEI hay and NEI CP were significantly lower in corngrass and SHS than in barley in 2016 and 2017, and significantly lower in SHS than in corngrass and in rye than in barley in 2017 (P < 0.05) ( Figure 5C,D). The NEI hay and NEI CP of C4 species were, respectively, 26.50% and 13.44% lower than that of C3 species.

Discussion
When soil-available N is large, a great plant N capture will reduce soil mineral availability, substrate for denitrifiers, and the abundance of nitrate-reducing microorganisms, which will increase soil N2O emission [20,38,39]. Therefore, the productivity of biomass and crude protein are the critical factors regulating plant-soil interactions in terms of N2O emission [19]. However, there were contrasting results between C3 and C4 grasses, i.e., C4 grass species had a relatively higher soil N2O flux and higher crude protein and biomass yields than C3 species (Figures 3 and 5B). These discrepancies were probably because of species particularities and the limited soil N concentration.

Discussion
When soil-available N is large, a great plant N capture will reduce soil mineral availability, substrate for denitrifiers, and the abundance of nitrate-reducing microorganisms, which will increase soil N 2 O emission [20,38,39]. Therefore, the productivity of biomass and crude protein are the critical factors regulating plant-soil interactions in terms of N 2 O emission [19]. However, there were contrasting results between C3 and C4 grasses, i.e., C4 grass species had a relatively higher soil N 2 O flux and higher crude protein and biomass yields than C3 species (Figures 3 and 5B). These discrepancies were probably because of species particularities and the limited soil N concentration. C4 grass species (such as the corngrass and SHS in the study) require more water, which may cause soil temperature to be more sensitive to the change of sunshine, while soil temperature is positively correlated with N 2 O emission [40,41]. Meanwhile, the competitions of soil microbial communities in high-N soil are more effective for N source than that in in low-N soil [42,43]. Therefore, it is likely that grass species with great water requirement moderate N 2 O emission by decreasing soil temperature and enhancing the N competitions of soil microbes instead of controlling N uptake and then diminishing the abundance or activity of soil microbes in low-available-N or no-N-addition soils. Our results show that soil-available N explained > 70% variance of soil N 2 O emission, and the sensitivities of soil N 2 O emission to available N between C3 and C4 species was not significantly different (Table 3). Soil-available N contributed greater impact to N 2 O emission than soil temperature and soil water content (Table 4). Fertilizing N significantly increases soil N 2 O emission because it actually increases the soil-available N as a substrate for nitrification and denitrification [29]. Our results prove again that available N is the most important factor that directly reflects the soil N 2 O emission no matter what type of grass species is used in the land [15,29].
The most parsimonious multiple linear regression model unraveled the two most important factors, beside available N, as the key predictors of N 2 O emission, the soil temperature and soil water content (Table 4). Soil temperature is the major driver regulating soil N 2 O fluxes mainly via soil respiration and microbial activity [16][17][18]. As reported in studies on N 2 O fluxes in sorghum and wheat fields in arid region [40,41], we found that increasing soil temperature elevated N 2 O fluxes ( Figure 4). These results agree with the general conclusions of previous studies [16][17][18]44]. Therefore, it may be prevalent that N 2 O fluxes start to increase in spring (May) and peak in summer (July) (Figure 2A,B), because the soil warming promotes microbial activity [16,17]. In the inland arid region of northwest China, the optimum water requirement for wheat is 414 mm, which is less than that for corn (e.g., 570 mm) [42]. Grass species have different water requirements, and nutrient use efficiency is likely to have an effect on soil properties including soil temperature [4,5], and then to vary the richness and diversity of microbes regulating soil N 2 O emission [43,44]. The greater richness and diversity of microbes in C4 grass fields is likely to result in the higher sensitivity of soil N 2 O emission to soil temperature in C4 than in C3 grass fields (Table 3).
Contrary to other findings [45,46], a significantly negative impact of soil water content on soil N 2 O emission was detected in the study (Figure 4). A greater soil water content may be beneficial for microbe-regulating soil N 2 O emission [38,39]. However, it also activates soluble salts in soil that negatively affects the mineralization and nitrification. In fact, soil water content influences N 2 O flux via changing gas diffusion rate and oxygen availability or regulating microbial communities because they require water for physiological activities [17]. However, different soil types may have specific soil moisture to optimize N 2 O flux [47]. N 2 O transport is restricted when moisture exceeds the optimum level [48,49] leading to anaerobic conditions, whereas suboptimal moisture levels will limit N 2 O flux due to water stress of soil microbes [16]. In addition, the high soil water content may restrict the soil temperature, which may significantly suppress soil N 2 O emission. Therefore, depending on the region, the change of soil water content may make the soil N 2 O emission absolutely different. N 2 O emission intensity (NEI, yield-scaled N 2 O emissions, i.e., cumulative emissions divided by hay yield or crude protein yield [28]) gives an effective measure to balance potential trade-offs between food production and environmental sustainability [50]. We found that C4 grass species may respectively lead to 26.5% and 13.4% reductions on N 2 O emission per unit of hay and crude protein production ( Figure 5C,D). In arid regions, traditional annual C3 grass species are planted more in order to release grazing pressure in cool seasons due to its short growing season and abundant biomass. Our study shows that in the current context of widespread C3 grass species use, C4 species should be promoted to develop animal husbandry that retain N more efficiently and contribute optimally to climate change mitigation. Therefore, improving cultivation of C4 species will be a huge challenge or an opportunity for future agriculture.

Conclusions
Our results indicated that in the arid regions with higher soil salt, soil in C4 fields emitted more N 2 O than that in C3 fields; however, C4 grass species produced greater hay and crude protein yield with lower hay/crude protein yield-scaled N 2 O emission. Therefore, we may conclude that C4 species contribute optimally to climate change mitigation in arid regions. We revealed that soil-available N was the most important factor regulating soil N 2 O flux in both C3 and C4 grass fields, and soil temperature and soil water content were the other two key factors. Our results assume that C4 species with great water requirement moderate N 2 O emission by decreasing soil temperature and enhancing the N competitions of soil microbes instead of controlling N uptake and then diminishing the abundance or activity of soil microbes in low available N or no N addition soils. Further investigations into this assumption should be conducted in the similar environmental region.