Nitrous Oxide Emission and Crop Yield in Arable Soil Amended with Bottom Ash

Bottom ash (BA), a byproduct of coal combustion from electric power plants with a porous surface texture and high pH, may influence the physical and chemical properties of upland arable soil associated with nitrous oxide (N2O) emission from upland soil. This study evaluated the use of BA in mitigating N2O emissions from upland arable soil and increasing the crop yield. In a field experiment, N2O emitted from the soil was monitored weekly in a closed chamber over a 2-year period (2018–2019). BA was applied to upland soil at the rates of 0, 200, and 400 Mg·ha−1. Cumulative N2O emission significantly decreased with increasing BA application rate; it decreased by 55% with a BA application rate of 400 Mg·ha−1 compared with the control. Yield-scaled N2O emission decreased with increasing BA application rates of up to 200 Mg·ha−1. Water-filled pore spaces (WFPS) were 70.2%, 52.9%, and 45.3% at the rates of 0, 200, and 400 Mg·ha−1, respectively, during the growing season. For economic viability and environmental conservation, we suggest that BA application at a rate of 200 Mg·ha−1 reduces N2O emissions per unit of crop production.


Introduction
Nitrous oxide (N 2 O) has 298-fold higher global warming potential than carbon dioxide (CO 2 ) over a 100-year time horizon, contributing up to 6% of global warming [1,2]. Agricultural soils are the primary source of global anthropogenic N 2 O emissions, accounting for approximately 60% of the total N 2 O emissions [3], primarily owing to the application of synthetic nitrogen (N) fertilizers and manures. N 2 O is mainly produced as a result of microbial nitrification and denitrification in soils. Although nitrification requires O 2 , denitrification relies on its absence or limitation and is attributed to anoxic conditions. Water-filled pore spaces (WFPSs) of soils can be an indicator of the activity of aerobic and anaerobic microorganisms. Here, higher WFPS values indicate that more air replaced by water in pores, thus removing O 2 from the soil. Changes in WFPS value comprise an important factor influencing N 2 O emissions from arable soils [4][5][6][7]. Nitrification that requires NH 4 + as an inorganic N substrate for aerobic respiration is the predominant process of N 2 O production from soils with a WFPS value of <60%. Conversely, denitrification that requires NO 3 − as an alternative electron acceptor for anaerobic respiration is a predominant process of N 2 O production from soils with a WFPS value of >60% [6,[8][9][10][11][12]. N 2 O emissions increase dramatically when both processes occur simultaneously as the soil WFPS value increases to almost 60% [13]. Soil pH is considered a critical factor for controlling various chemical and biological soil properties. Furthermore, soil pH has a significant effect on microbial processes responsible for the production and consumption of N 2 O in soils [14]. Primarily, the last step of denitrification (N 2 O → N 2 )

Characterization of BA
The BA used in this study was a byproduct of the coal-fired power plant from Korea South-East Power Co., Ltd. in Yeongheung, Republic of Korea. The BA was produced at a high temperature of approximately 1200 • C. Expanded X-ray diffraction (XRD) pattern analysis was performed to characterize the BA crystal structures as shown in Figure 1a. The surface morphology of BA and its physical and chemical properties are shown in Figure 1b and Table 2, respectively. BA had a pH of 9.3, total carbon of 6.6 g·kg −1 and total nitrogen of 0.9 g·kg −1 . It was porous (58% porosity) and had a large surface area (10.1 m 2 ·g −1 ).

Characterization of BA
The BA used in this study was a byproduct of the coal-fired power plant from Korea South-East Power Co., Ltd. in Yeongheung, Republic of Korea. The BA was produced at a high temperature of approximately 1200 °C. Expanded X-ray diffraction (XRD) pattern analysis was performed to characterize the BA crystal structures as shown in Figure 1a. The surface morphology of BA and its physical and chemical properties are shown in Figure 1b and Table 2, respectively. BA had a pH of 9.3, total carbon of 6.6 g·kg −1 and total nitrogen of 0.9 g·kg −1 . It was porous (58% porosity) and had a large surface area (10.1 m 2 ·g −1 ).  Silicate (SiO2) and calcium carbonate (CaCO3) showed distinct characteristic peaks, indicating high purity and crystallinity. However, calcium oxide (CaO) showed several weak characteristic peaks. As shown during XRD pattern analysis, the relative percentages of SiO2, CaCO3, and CaO were 60%, 39%, and 1%, respectively.

Experimental Design and Crop Management
The BA was applied at the rates of 0, 200, and 400 Mg·ha −1 to the experimental plots before crop cultivation on 1 April 2018. In total, 12 plots (3 m × 4 m for each plot) were arranged in a randomized complete block design with four replicates. First, radish seeds were sown and then maize seeds. Radish seeds were sown and harvested on 30 March  Silicate (SiO 2 ) and calcium carbonate (CaCO 3 ) showed distinct characteristic peaks, indicating high purity and crystallinity. However, calcium oxide (CaO) showed several weak characteristic peaks. As shown during XRD pattern analysis, the relative percentages of SiO 2 , CaCO 3 , and CaO were 60%, 39%, and 1%, respectively.

Experimental Design and Crop Management
The BA was applied at the rates of 0, 200, and 400 Mg·ha −1 to the experimental plots before crop cultivation on 1 April 2018. In total, 12 plots (3 m × 4 m for each plot) were arranged in a randomized complete block design with four replicates. First, radish seeds were sown and then maize seeds. Radish seeds were sown and harvested on

N 2 O Emission Measurements
A closed chamber method [27] was used to measure N 2 O emission from soil. Static PVC column chambers (headspace; 10.8 L, 25 cm diameter × 22 cm height) were placed at the center of each plot on 18 March 2018. The collar was placed between crops (radish and maize), and crops were not planted inside the chamber. All weeds that grew inside the chamber were removed during the experimental period. After 2 weeks of acclimation, the collar was closed with a fitted lid and air vent and rubber septa to collect gas samples. The plants grown inside the chamber were removed during the experimental period. Gas samples were collected using 30-mL syringes at 0, 20, and 40 min after covering the lid between 10:00 and 12:00 h throughout the year-long study period. Gas samples were collected once a week during the growing season of radish and maize, twice a week at special events, such as chemical fertilizer application and heavy rainfall, and once every 2 weeks during the fallow season. Gas sampling was performed throughout the study period from April 2018 to March 2020. The temperature in the chamber during gas sampling was measured using a portable thermometer (WT-1, Elitech, London, UK). The concentration of N 2 O was analyzed using a gas chromatograph-mass spectrophotometer (GC-MS QP2020, Shimadzu, Kyoto, Japan). N 2 O fluxes were calculated based on the slope of the linear increase in concentration during the chamber-closure period.
The N 2 O flux was calculated using the following formula: where ρ is the gas density of N 2 O (g·m −3 ), V is the volume of the chamber (m 3 ), A is the soil area covered by the chamber (m 2 ), ∆c/∆t is the rate of change in gas concentration in the chamber (g·m −3 ·min −1 ), T is the average temperature in the chamber ( • C), k is the time conversion coefficient (min·day −1 ), and a (10,000 m 2 ·ha −1 ) is the area conversion coefficient. Cumulative N 2 O emission was calculated as follows (kg·N 2 O·ha −1 ): where F represents the N 2 O flux (g·N 2 O·ha −1 ·day −1 ), i is the ith measurement, and (t i+1 − t i ) is the number of days between two adjacent measurements. Yield-scaled N 2 O emission was calculated as follows [28]:

Soil Sampling and Analysis
In each plot, soil sampling was performed before starting the experiment, after base and additional fertilizer applications, and after crop harvesting in the radish and maize growing season. Samples of the soil layers at 0-15 cm depth were collected using a hand auger (5 cm diameter) and a core sampler (100 cm 3 ) to measure soil bulk density. Samples were then air-dried, passed through a 2 mm sieve, and kept in a plastic zipper bag for chemical analysis. Soil pH and EC were measured in a 1:5 soil:distilled water suspension using a pH and electrical conductivity meter (Orion Star A215, Thermo Scientific Orion, Mansfield, TX, USA). Organic matter content was analyzed using the Walkley and Black method [29], total nitrogen was analyzed using the Kjeldahl method, and available phosphorus was analyzed using the Lancaster method (5 g soil was extracted with 20 mL of 0.33 M CH 3 CHOOH, 0.15 M lactic acid, 0.03 M NH 4 F, 0.05 M (NH 4 ) 2 SO 4 , and 0.2 M NaOH at a pH of 4.25) [30]. Analysis of soil inorganic nitrogen content was performed by adding 25 mL of 2 M KCl to 5 g of dry soil samples and shaking for 30 min, after which sample solutions were filtered using Whatman No. 2 filter paper. Ammonium nitrogen (NH 4 + ) was measured using the indophenol-blue colorimetric method [31], and nitrate-nitrogen (NO 3 − ) was measured using the brucine method [32].
Soil temperature and moisture at 10 cm soil depth were measured for each treatment at 3 h intervals using sensors (5TE Water Content, Temperature, and Electrical Conductivity, Decagon Devices, Inc., Pullman, WA, USA). Weather data, such as daily average temperature and precipitation during the experimental period were obtained from the local weather observation data (Aanderaa, Automatic weather station, xylem, Bergen, Norway) of the Korean Meteorological Administration (latitude: 35 • 29 N; longitude: 128 • 44 E; distance from the experimental field: 7.48 km).
WFPS (%) was calculated every day for a year using the following equation: where θ is the volumetric moisture content (m 3 ·m −3 ). Soil porosity (m 3 ·m −3 ) was calculated using a particle density value of 2.65 g·m −3 and soil bulk density values. Bulk density was determined using soil samples of all plots collected from a depth of 0-15 cm once a month during the study period. Samples were collected by a fixed-volume core (94.64 cm 3 ) and dried at 105 • C.

Statistical Analysis
Statistical analysis was performed using R software package V. 3.6.2 (http://www. R-project.org, accessed on12 June 2020). The mean values of cumulative N 2 O emissions, radish and maize yields, and yield-scaled N 2 O emissions were compared using pairwise comparison. Differences among parameters were determined using two-way analysis of variance. Application rate (R) and year (Y) were considered to be fixed effects. Owing to application rate × year interactions for cumulative N 2 O emission during radish and maize growing seasons, each year was separately analyzed. The least significant difference was used for multiple comparisons between the means and performed only when the F-test result was significant (p < 0.05).

N 2 O Flux
N 2 O flux patterns differed with air temperature patterns over the study period, as shown in Figure 2a,b. Although flux was relatively low during the cold and dry fallow season, it did not peak as air temperature reached its maximum in the August of both Years 1 and 2. N 2 O flux peaks appeared only during the radish and maize growing seasons in Years 1 and 2 when basal and additional N fertilizers were applied, but peaks did not appear during fallow seasons. More N 2 O flux peaks appeared during the maize growing season than during the radish growing season in Years 1 and 2.
Daily WFPS values varied over 2 years and increased after high rainfall and irrigation events during this period (Figure 2c). BA application rate resulted in a change in soil WFPS values. Daily WFPS values of soil amended with 0 Mg·ha −1 of BA were always the highest, followed by those of soil amended with 200 and 400 Mg·ha −1 of BA over the 2 years.  , and water-filled pore space (c) after the application of bottom ash at various rates for 2 years. Arrows in the graph represent fertilizer addition and irrigation, and BF and AF denote base and additional fertilizer applications, respectively.

Cumulative N2O Emission
There was a significant R × Y interaction for cumulative N2O emission during the radish growing season (Table 3). Cumulative N2O emission during the radish growing season decreased significantly following application of 400 Mg·ha −1 of BA in Year 1 and with increasing BA application rate in Year 2 (Table 4).
A significant R × Y interaction was noted for cumulative N2O emission during the maize growing season (Table 3). Cumulative N2O emission during the maize growing , and water-filled pore space (c) after the application of bottom ash at various rates for 2 years. Arrows in the graph represent fertilizer addition and irrigation, and BF and AF denote base and additional fertilizer applications, respectively.

Cumulative N 2 O Emission
There was a significant R × Y interaction for cumulative N 2 O emission during the radish growing season (Table 3). Cumulative N 2 O emission during the radish growing season decreased significantly following application of 400 Mg·ha −1 of BA in Year 1 and with increasing BA application rate in Year 2 (Table 4).  A significant R × Y interaction was noted for cumulative N 2 O emission during the maize growing season (Table 3). Cumulative N 2 O emission during the maize growing season increased significantly with an increasing BA application rate in Year 1 and with a BA application rate of up to 200 Mg·ha −1 in Year 2 (Table 4).
Application rate, but not year, significantly influenced cumulative N 2 O emission during the entire year ( Table 3). The mean value of cumulative N 2 O emission during the entire year across both Years 1 and 2 decreased significantly with increasing BA application rate (Table 4).

Changes in the Physical and Chemical Properties of Soil
Mean WFPS values during the radish growing season, maize growing season, and the entire year for Years 1 and 2 decreased with BA application (  The mean value of soil pH at maize harvest time across both Years 1 and 2 increased significantly with a BA application rate of up to 200 Mg·ha −1 , but no further increase was observed with additional BA application (Table 6).

Crop Yield and Yield-Scaled N2O Emission
BA application rate did not influence radish and maize yields in this study (Tables 3  and 7).

Crop Yield and Yield-Scaled N 2 O Emission
BA application rate did not influence radish and maize yields in this study (Tables 3 and 7).
BA application rate significantly influenced yield-scaled N 2 O emissions (Table 3). Radish yield-scaled N 2 O emission decreased significantly with increasing application rates of BA (Table 7). Maize yield-scaled N 2 O emission and total yield-scaled N 2 O emission decreased significantly following a BA application rate of up to 200 Mg·ha −1 , but no further decrease was noted with additional BA application.

Discussion
Precipitation affects N 2 O emissions from arable soil. Some studies have reported that N 2 O flux peaked after a high-rainfall event, which led to soil adopting an anoxic state for N 2 O production via denitrification [33][34][35][36]. However, in this study, the peak of the N 2 O flux did not appear after high-rainfall events, despite several occurrences of such events over the 2 years. All N 2 O flux peaks appeared only during the radish and maize growing season. The peak of the N 2 O flux did not appear soon after N-fertilizer application, but rather occurred 3-23 days later after urea application. Further, daily N 2 O flux is associated with soil WFPS values. Nitrification is the predominant process for N 2 O emission from soils with <60% WFPS, whereas denitrification is the predominant process of N 2 O emission from soil with >60% WFPS [10][11][12]. When soil WFPS value is approximately 60%, N 2 O production increases considerably owing to simultaneous nitrification and denitrification [37]. When soil WFPS value increases by >70%, soil environmental conditions favor denitrification and N 2 is emitted instead of N 2 O [13]. In the present study, all N 2 O flux peaks after N-fertilizer application during the growing season in both years appeared when WFPS value was approximately 65%, as shown in Figure 2a,c. For example, the peak of the daily N 2 O flux did not appear following application of 0 Mg·ha −1 of BA on 29 May 2018, when a basal N fertilizer was applied during the maize growing season in Year 1 because daily WFPS value was 86% on that day. Under these O 2 -limited conditions, N 2 is primarily emitted instead of N 2 O. However, the first and second peaks of daily N 2 O flux following application of 0 Mg·ha −1 of BA appeared on 21 June and 17 July after basal N-fertilizer application when daily WFPS values were 66% and 62%, respectively. NO 3 − concentrations in soil increased with increasing BA application rate during radish and maize growing seasons (Figure 3), whereas that of NH 4 + decreased with increasing BA application rate. This implies that nitrification is the predominant N 2 Oproducing process rather than denitrification in BA-amended soils as BA application decreases soil WFPS values. A 35%-60% WFPS value constitutes favorable soil water conditions for nitrification. The mean WFPS value during the entire year across Years 1 and 2 decreased from 70.2% with 0 Mg·ha −1 of BA to 52.9% and 45.3% with 200 and 400 Mg·ha −1 of BA, respectively ( Table 5).
As mentioned above, this study examined three hypotheses. The first hypothesis was that the application of porous BA decreases bulk density and WFPS value of soil to Agriculture 2021, 11, 1012 11 of 15 render soil conditions unfavorable for microorganisms associated with N 2 O-production processes, such as nitrification and denitrification. Results from this study confirmed the first hypothesis. The bulk density of soil at maize harvest time decreased significantly following BA application at 200 Mg·ha −1 (Table 6). This decrease in the bulk density of soil following BA application was owing to the physical properties of BA, including its high porosity and large surface area (Table 2). Subsequently, this reduced the WFPS of soil. The mean WFPS value during the entire year across both Years 1 and 2 decreased from 70.2% following application of 0 Mg·ha −1 of BA up to 45.3% with 400 Mg·ha −1 of BA (Table 5). Specifically, daily WFPS value with 200 and 400 Mg·ha −1 of BA was mostly below 35% during the entire year, except for high rainfall and irrigation events, as shown in Figure 2c. A WFPS value of 35-60% constitutes favorable soil water conditions for nitrification. However, nitrification and denitrification reduced in water-limited conditions involving a WFPS value of <35% [12,38]. BA application may ensure water-limited conditions for microorganisms involved in nitrification and denitrification and thus decrease N 2 O emission from soil. Similar results with this study were observed by other researchers using ash materials, such as biochar and charcoal. Carvalho et al. [39] reported that WFPS value decreased significantly by approximately 10% following the application of 32 Mg·ha −1 of wood biochar in a bean-rice rotated cultivation system. They observed a positive correlation between N 2 O fluxes and WFPS value, indicating that WFPS was a relevant soil variable related to N 2 O emission. In addition, Yanai et al. [40] reported that suppressed N 2 O emissions after adding charcoal stemmed from changes in WFPS values rather than the addition of Cl − and SO 4 2− , which were the major anions in charcoal based on laboratory experiments. In the current study, a large decrease in WFPS value (from 70.2% to 45.3%) with BA application was observed compared with the results of other studies [39,40], owing to a greater BA application rate (400 Mg·ha −1 ). Therefore, we observed a further decrease in cumulative N 2 O emission by up to 54.8%, i.e., from 17.7 kg·N 2 O·ha −1 to 8.0 kg·N 2 O·ha −1 .
The second hypothesis was that alkaline BA application increases soil pH, promotes reduction of N 2 O to N 2 , and decreases N 2 O emission. Soil pH increased following BA application (Table 6) owing to the chemical properties of BA such as presence of large amounts of CaCO 3 and CaO (Table 2). Soil pH is a primary factor influencing N 2 O production and consumption processes in soil [41,42]. Several studies have reported that the abundance of nitrogen-cycling genes and the rates of nitrification and denitrification are strongly regulated by soil pH [43][44][45][46]. Notably, Nos activity is more sensitive to low pH than other reductases in denitrification [47]. Therefore, under low soil pH conditions, more N 2 O is produced than N 2 [48]. The ratios of N 2 O/(N 2 + N 2 O) showed a significant negative correlation with soil pH within the generally observed pH range of 5-8 in agricultural soils [45,48]. Increasing pH of acidic soil decreases N 2 O emissions from nitrification by increasing ammonia oxidizing bacteria (AOB) gene copy numbers [49]. In this study, soil pH increased from 5.4 at 0 Mg·ha −1 of BA up to 5.9 following application of 200 Mg·ha −1 of BA (Table 6). Therefore, an increase in soil pH with BA application may help reduce N 2 O emission from the soil by decreasing AOB abundance during nitrification and increasing Nos activity during denitrification. Kim et al. [21] reported that soil pH increased from 6.03 to 6.16 with 30 Mg·ha −1 of BA application in upland soil used for growing lettuce. In the current study, a greater increase in soil pH (from 5.4 to 5.9) with BA application was observed compared with their results as a higher BA rate (200 Mg·ha −1 ) was applied. Some studies have reported that liming affects acidic soil by suppressing N 2 O emissions. Shaaban et al. [50] observed that soil pH of two acidic soils (5. 25  The third hypothesis was that BA comprising various elements provides available nutrients for plant growth and increases crop yield. However, this hypothesis was not confirmed based on our results. Radish and maize yields did not increase with increasing BA application rate. Rather, BA application decreased bulk density and WFPS value of soil in the present study (Tables 5 and 6), implying improved physical properties, such as porosity and aeration. However, several studies have reported that crop yield was not influenced by changes in physical soil properties for a short period of <5 years [51][52][53][54]. This study was conducted for a short period of 2 years. Improved crop yield through changes in physical properties of soil via BA application may have been observed for a study period of >5 years. In addition, macronutrient concentration, such as total nitrogen, available phosphate, and exchangeable K and Ca in soil for plant growth, did not increase with increasing BA application rates. However, BA used in this study comprised nitrogen, phosphorus, CaCO 3 , and CaO (Tables 2 and 6). We assumed that the low solubility of BA in the soil did not enrich the soil with sufficient plant nutrients to increase radish and maize yields.
In this study, daily N 2 O fluxes were measured using a static chamber without growing any plants, including radish, maize, and weeds, for the entire experiment period. This may incorrectly reflect the cumulative N 2 O emission from the real crop cultivation system because the effect of rhizosphere soil on N 2 O emission was excluded. Rhizosphere soil encompasses the narrow zone of contact between the roots and soil particles and plays an important role for both plant growth and N 2 O emissions [55,56]. Rapid nitrogen transformations and translocations occur in the rhizosphere soil via root uptake and microbial activities mainly from roots and microorganisms interacting and competing with each other for nutrients [57]. Plant roots directly capture NH 4 + , NO 3 − , and amino acids according to their growth demands while excreting sugars, organic acids, and amino acids, which stimulate microbial growth and influence carbon and the nitrogen biogeochemical cycles. Despite the importance of the rhizosphere and its management for developing effective mitigation strategies, studies quantifying N 2 O emission in the rhizosphere are surprisingly scarce owing to the lack of appropriate tools for performing measurements in such microenvironments. Measuring N 2 O emission in microenvironments remains challenging owing to soil heterogeneity and the gaseous nature of N 2 O [58,59].
Agricultural practices can be associated with N 2 O emission by estimating cumulative N 2 O emission based on crop yield, which are referred to as yield-scaled N 2 O emission. From global environment and food security perspectives, sustainable agriculture in the future should explore systems with low N 2 O emissions and high crop yield to ensure food security. Yield-scaled N 2 O emission provides essential information for estimating the environmental impacts of intensive agricultural production systems [60]. The mean value of total yield-scaled N 2 O emission across both Years 1 and 2 decreased significantly following BA application at 200 Mg·ha −1 and no further decrease was observed after additional BA application (Table 7). A lower value of yield-scaled N 2 O emission indicates less N 2 O emission for the same unit of crop production. Therefore, applications of up to 200 Mg·ha −1 of BA may be environmentally and economically beneficial in soil management to reduce N 2 O emissions while maintaining crop production.