Reduction of Nitrogen through Anaerobic Processes in Chinese Rice Paddy Soils

Abstract

Comprehending the anaerobic nitrogen transformations, including denitrification, anaerobic ammonium oxidation (anammox), and anaerobic ammonium oxidation linked with iron reduction (Feammox) in soil, is essential for improving soil fertility and minimizing the environmental impacts of nitrogen loss. Despite this, research on anaerobic nitrogen transformations, particularly Feammox in paddy soil, is sparse. This study examined soil denitrification, anammox, and Feammox, along with their respective contributions to nitrogen loss in paddy soil at various depths, under different fertilization and irrigation treatments. It utilized 15N isotope labeling to investigate the limiting factors of these anaerobic nitrogen transformations and their interactions. The findings showed that denitrification rates ranged from 0.41 to 2.12 mg N kg⁻¹ d⁻¹, while anammox rates ranged from 0.062 to 0.394 mg N kg⁻¹ d⁻¹, contributing 84.3% to 88.1% and 11.8% to 15.7% of total soil nitrogen loss, respectively. Denitrification was identified as the predominant pathway for nitrogen loss across different soil depths. Fertilization and irrigation had more pronounced impacts on anaerobic nitrogen transformations than did soil depth, potentially affecting these processes through both abiotic and biotic pathways. This study identified significant correlations among the three types of anaerobic nitrogen transformations. These findings offer a theoretical foundation for optimizing nitrogen management strategies to mitigate losses in agricultural systems.

1. Introduction

Paddy fields are a major part of agricultural ecosystems and soil nutrients determine agricultural production indicators such as grain production and crop quality [1]. Paddy fields are subjected to long-term fertilization and perennial irrigation, making them a hot spot for anaerobic nitrogen (N) transformation, resulting in soil N loss and agricultural yield damage [2,3]. In general, the pathways of anaerobic N transformation in the soil include denitrification and anammox. The denitrification process was traditionally considered the main pathway of N loss through the generation of soil N gases such as nitrous oxide (N₂O) or di-nitrogen (N₂) [4]. A previous study demonstrated that denitrification accounts for more than 50% of the total N loss from the soil ecosystem [5]. The anammox mechanism was seen in the denitrifying reactor [6]. The anammox process has been confirmed by many researchers [7,8] and was initially used in wastewater treatment as a tool for removing ammonia. In a previous study, the range of anammox contribution was identified as different in different environments: anammox contributes 36.8–79.5% in wetlands interface and 4–37% in paddy soils [5,9]. Except for these processes, it has recently been discovered that anaerobic ammonium (NH₄⁺) ion oxidation can be coupled with ferric iron (Fe (III) reduction (termed Feammox) with either N₂, nitrite (NO₂⁻), or nitrate (NO₃⁻) as the end product in several environments [10]. For instance, anaerobic oxidation of NH₄⁺ to NO₂⁻ has been detected under iron-reducing conditions in wetland soils [11]. Additionally, Yang, Weber and Silver [10] found evidence of N₂, NO₂⁻, or NO₃⁻ production via Feammox in tropical forest soils. These new findings illustrate alternative pathways of soil N loss [12]. There may be a synergistic relationship between Feammox and denitrification, as well as competition between Feammox and anammox for NH₄⁺. To date, however, there have been few reports on the interactions among these processes in natural ecosystems.

Nitrogen fertilizers have been utilized extensively for higher yield in paddy soils worldwide, especially in China [1]. According to FAO, China is currently the largest producer and consumer of N fertilizer; the amount of N fertilizers used is almost 40% of the global level [13]. The anthropogenic sources of N fertilizers in agriculture soils contribute approximately 60% N₂O, which is the intermediate product of the denitrification pathway [5]. The use of slow-release fertilizers can regulate N release from the soil according to plant demand, and can reduce 58% N losses by leaching and volatilization [14]. Fertilizer management follows water management, which is more crucial nowadays due to increasing water scarcity. Continuous irrigation (CI) is a common practice in paddy soils worldwide, resulting in denitrification and methane production [15]. Research studies which have investigated the effect of wetting and drying irrigation (AWD) on rice production, N use efficiency (NUE), and N₂O emissions found that there was no increase in N₂O production but that the rice yield increased by 27%, and NUE increased by 176%. In addition it was found that the AWD technique reduced water use by 40% without any loss in the yield of rice [16]. Limited research has been carried out on the impact of various agricultural measures on anaerobic N transformation. In particular, it is not clear whether Feammox, as a newly discovered N species, will be affected by fertilization in the process of N loss.

Soil ecological factors are the main elements involved in evaluating soil quality. Agricultural ecosystems are vulnerable to agricultural production activities, especially fertilizer application and crop cultivation, which have a significant impact, resulting in drastic changes in various physical, chemical, and biological characteristics of the soil, hence affecting its N transformations [14]. Studies have shown that the application of N fertilizer reduces soil pH and increases soil inorganic N content (SIN) [10,17]. The application of phosphorus (P) fertilizer has little effect on soil physicochemical characteristics, but it has an impact on denitrification [18]. Crop planting affects the soil environment at the roots of the crop, resulting in fluctuations in soil SIN content [19]. It has been shown that fertilization accelerates topsoil mineral decomposition to increase soil NH₄⁺-N content, and NH₄⁺-N content also fluctuates with crop fertility [20]. Theoretically, the anaerobic N transformation can be regulated by agricultural activities to alleviate the N overload in farmland, but in practice, fertilization is still needed to maintain the growth of the crops. If the anaerobic N transformations keep increasing, N loss which is detrimental to crop output may result [21]. The search for the N balance between the N loss process and crop demand has been a hot research topic for these reasons. The study of the mechanism of anaerobic N transformations which control soil N loss processes is a key entry point to studying the above problems.

The primary objective of the current study was to determine denitrification, anammox, and Feammox rates and their contribution to the total nitrogen loss in paddy soils of southern China. The dominant environmental factors affecting the anaerobic N transformation were considered. The findings should provide reference data for the study of the anaerobic N transformation mechanisms and generate constructive suggestions for promoting soil N utilization and reducing negative environmental impact.

2. Materials and Methods

2.1. Site Description and Experiment Design

The field experiments were carried out between 2018 and 2020 at an independent site located in Ningbo City, Zhejiang Province (29°39′9.05″ N, 121°35′0.07″ E). There are rice and potato rotations in the region of the experiment. The properties of the experimental site are mentioned in

Table 1. The properties of the experimental site.

Mean annual temperature	16.2 °C
Mean annual precipitation	1564 mm
pH	5.1
Organic matter content	41.9 g kg−1
Electrical conductivity	61.7 μs cm−1
Alkali hydrolysable N	277 mg kg−1
Available phosphor	90.2 mg kg−1
Available potassium	140 mg kg−1

.

The fields were treated with different irrigation and fertilization methods [22]. The detailed irrigation treatments were alternate, while the fertilization treatments were unfertilized control (CK), farmer’s use of traditional fertilizer (FTF; 500 kg organic fertilizer, 30 kg urea, and 50 kg calcium superphosphate), pig manure fertilizer (PMF; 146 kg pig manure, 17.3 kg urea, 7.25 kg calcium superphosphate, and 8 kg potassium chloride), and slow-release fertilizer (SRF; 3.33 kg slow-release fertilizer, 17.3 kg urea, no calcium superphosphate, 8.4 kg potassium chloride). Fertilizers were added as base fertilizers and broadcast applications. The irrigation rates of AWD and CI were 6404 m³ ha⁻¹ and 10,966 m³ ha⁻¹, respectively. The presence of nitrogen compounds in the irrigation water was about 8 mg/L to 30 mg/L. The average TN runoff loss during the rice growing season varied from 3.48 to 21.28 kg N/ha depending on the year and treatment. CI plots were irrigated sufficiently—an average of twice every three days—to maintain a mean ponded water depth of 20–30 mm. Plots were continuously flooded from one week before to one week after flowering. AWD was imposed from 10 to 55 days after transplanting. The maximum and minimum water depths following rewatering of AWD plots were 80 mm above the soil surface at 35 days after transplanting and 120 mm below the soil surface at 24 days after transplanting. The specific fertilization methods are summarized in Table S1, and the distribution of experimental fields is shown in Figure 1. Each treatment had 3 replications (Figure 1; Table S1).

2.2. Soil Sampling and Analyses

Soil samples were collected in November 2018. Specifically, five 1 m × 1 m quadrats were set up in each field. The aboveground rice was collected from each quadrat, oven dried at 60 °C and weighed as the crop yield. Soil was extracted from each quadrat at the following depths: 0–20 cm, 20–40 cm, 40–60 cm, and 60–80 cm. After being sieved by 2 mm mesh, the soil sample was divided into two halves. One subsample was stored at 4 °C for the isotope labeling experiment and the remaining sample was returned to the laboratory and air-dried to measure the soil properties.

After air drying, soil pH (1: 2.5 soil: H₂O) was measured using a soil pH meter (Figure S2) [23]. Soil organic carbon (SOC) and total N (TN) content were determined using an elemental analyzer (Vario EL III, Elementar, Langenselbold, Germany) after inorganic C was removed by soaking in 0.1 M hydrochloric acid (HCl) (Figure S3) [24]. Soil NO₃⁻ and NH₄⁺ contents were extracted with 2 M potassium chloride (KCl) and determined by microplate-colorimetric techniques using the vanadium-chloride and salicylate-nitroprusside methods, respectively (Figures S4 and S5) [25]. Available P was determined by the Mo-Sb colorimetric method, and available potassium was determined by flame photometry (FP640; INASA, Shanghai, China) (Figure S6) [26]. The fresh soil was utilized for measuring the soil nitrification potential by the shaken-soil-slurry method [27,28].

2.3. ¹⁵N-Labelled Incubations

The production rates of denitrification, anammox, and Feammox were determined by ¹⁵NO₃⁻-N and ¹⁵NH₄⁺-N isotope labeling [29]. The detailed steps were as follows: before incubation, we conducted preincubation to activate soil microorganisms and eliminate oxygen (O₂) and NO_x⁻ in the soil. The soil was stored in jars at 4 °C, and the soil-to-water ratio was adjusted to 1:3 (dry soil: water). The jars were then incubated in the dark in an anaerobic incubator at 25 °C for 4 days. After preincubation, 12 g slurry was transferred into 50 mL serum vials (Labco, Loxstedt, Germany), which were further filled with helium (He) to isolate O₂. The experiment included three treatments (three replicates for each): (1) CK (addition of sterile anoxic deionized water only); (2) ¹⁵NH₄Cl (¹⁵N at. % = 99.9, NH₄⁺); (3) Na¹⁵NO₃ (¹⁵N at. % = 99.9, NO₃⁻). The highest concentration of ¹⁵NH₄⁺ or ¹⁵NO₃⁻ was 100 μM in (2) and (3). The vials were incubated at 25 °C for 24 h. We extracted gas at 0, 6, 12, and 24 h to determine the production rate of ³⁰N₂ and ²⁹N₂ and conducted destructive soil sampling at 0 and 24 h to measure the content of Fe(III). The production rates of ³⁰N₂ and ²⁹N₂ were determined using a gas chromatograph mass spectrometer (GC-MS) (Agilent 7890A/5975C inert MSD; Agilent, Santa Clara, CA, USA), while Fe(III) was measured by the hydrochloride-phenanthroline method.

2.4. Calculation of Anaerobic N Transformations

The potential rates of Feammox (R_f), denitrification (R_d), and anammox (R_a) were calculated. These are shown in

Table 2. The potential rates of Feammox (R_f), denitrification (R_d), and anammox (R_a) of Anaerobic N Transformations.

Rf	Δ30N2 = 30N2(2) − 30N2(1)
Rd	Δ30N2 = 30N2(3) − 30N2(1)
Ra	Δ29N2 = 29N2(3) − 29N2(1)

below:

In this table, ³⁰N₂₍₃₎, ³⁰N₂₍₂₎, and ³⁰N₂₍₁₎ are the production rates of ³⁰N₂ of treatments (1), (2), and (3), respectively. The ²⁹N₂₍₃₎ and ²⁹N₂₍₂₎ for the same.

2.5. Statistical Analysis

All data were analyzed using Excel 2016 and SPSS 20.0. OriginPro 8.5 was utilized for drawing figures. Duncan’s multiple range tests were utilized for analyzing statistical significance at the level of p < 0.05.

3. Results

3.1. Anaerobic N Transformations under Different Treatments

The denitrification rate, anammox rate, and the reduction rate of Feammox were 0.41–2.12 mg N kg⁻¹ d⁻¹, 0.062–0.394 mg N kg⁻¹ d ⁻¹, and 0–0.96 mg kg⁻¹ d⁻¹, respectively. The results of variation partitioning analysis (VPA) showed the most influential treatment of these three anaerobic N transformations and indicated that the impact size of different treatments was as follows: irrigation > fertilization > depth (Figure 2).

Under different irrigation methods, the three anaerobic N transformations were distinct. The CK and FTF treatments, the 0–20 cm (D1) and 40–60 cm (D3) of denitrification showed obvious differences, and the 20–40 cm (D2) changed significantly in the FTF treatment (p < 0.05). In the AW, all depths varied significantly in the CK treatment. In the FTF treatment, the D3 did not show any obvious difference. Feammox showed significant differences in D2, 60–80 cm (D4) in the CK treatment, and D1, D3, and D4 changed significantly in the FTF treatment (p < 0.05) (Figure 3a,b).

After different fertilization methods, in the AWD treatment, the denitrification rate was significantly different in D2, D3, and D4 (p < 0.05), whereas, in the CI treatment, the four depths were all significantly different (p < 0.05). The AW of the four depths changed significantly in the AWD treatment (p < 0.05), while only D1 and D2 showed significant differences after the CI treatment. Interestingly, Feammox had variations at all depths except D3 in the AWD treatment, but in the CI treatment, only D3 changed (Figure 3c,d).

3.2. Influencing Factors of Anaerobic N Transformations

The anaerobic N transformations exhibited strong associations with various soil properties across different treatments (Figure 4). Under fertilization conditions, NH₄⁺, NO₃⁻, and available phosphorus (P) were significantly correlated with the anaerobic N transformation pathways, with available P being the predominant factor in both CK and FTF treatments (Figure 4a,b). Post-irrigation, the AWD treatment showed significant associations between anaerobic N transformations and NH₄⁺, NO₃⁻, available P, soil organic carbon (SOC), and soil water content (WC). In the CI treatment, denitrification, Feammox, and anammox were notably influenced by SOC, NO₃⁻, and pH (Figure 4c,d).

The analysis of the interactions of anaerobic N transformations revealed that the rate of denitrification, Feammox, and anammox all had a significantly positive influence on each other, and the correlation between denitrification and anammox was particularly close (Figure 5).

3.3. Determinants of Anaerobic N Transformations

The results of the linear model indicated the dominant factors of these three anaerobic N transformations (Figure 6). For denitrification, anammox had the most significant effect on the variations. Among these variables, denitrification and anammox had the largest effect on anammox and Feammox, respectively, while the impact of soil physicochemical properties was less than the rate of all anaerobic N transformations.

4. Discussion

4.1. Effect of Irrigation on the Three Anaerobic N Conversion Pathways

In the current study, the effect of agricultural methods (irrigation and fertilization) on the three anaerobic N conversion pathways was much greater than the effect of depth (Figure 2). We speculate that the reason for this phenomenon is due to soil specificity. Since the sampled soil is planted with rice all year round, the extension of the rice root system forms a special soil environment [30]. During the growth of the rice, aeration tissues release O₂ through respiration, which makes the soil near the surface root form an aerobic condition, while the surrounding soil around the root system, or the deep soil, is all anoxic [30]. Numerous studies have demonstrated that the rates of denitrification and anammox are significantly higher in rhizosphere soils than in non-rhizosphere soils [31]. The root system secretes metabolites to provide nutrients to soil microorganisms, resulting in some variations in the activities of soil microorganisms in different soil profiles, with microorganisms in rhizosphere soils being more active than those in non-rhizosphere soils [32,33]. The formation of aerobic-anoxic interface and microbial activities result in the topsoil being more vulnerable to external disturbances, while the deep soil remains more stable.

4.2. Effect of Fertilization on the Three Anaerobic N Conversion Pathways

Many studies have been conducted on the effect of fertilization on N loss. In particular, researchers have studied the influence of fertilization on microbial communities associated with N cycling, providing a mechanistic explanation for this phenomenon [14,18,34]. It has been suggested that the application of fertilizer changes soil pH and N content and alters the microbial communities associated with anaerobic N transformation [18,35]. In the present study, denitrification, anammox, and Feammox were significantly correlated with NH₄⁺-N, NO₃^—N, and available P, whether the fertilizer was applied or not (Figure 4a,b). However, the degree of explanation of these three environmental factors was reduced after FTF. We suspected that the application of fertilizer would not significantly affect the soil environment, but would affect the soil microbial community. We also hypothesized that microorganisms do not change due to the shift in the soil environment, but rather due to other N cycle processes (e.g., nitrification), which are also influenced by fertilization, leading to changes in the reaction substrate for anaerobic N conversion.

4.3. The Role of Water Content in Soil Nitrification and Denitrification

Water content is essential to soil nitrification and denitrification [36]. The WC has a positive correlation with the intensity of denitrification, but a negative correltion with nitrification [37]. Clearly, irrigation is a crucial agricultural tool for regulating the soil N cycle in agricultural fields. Since CI is a common irrigation method, we considered CI a blank control and AWD an irrigation treatment. We found that there was no significant difference in the topsoil after the AWD treatment, but that changes were obvious in the deeper soil (Figure 3), which explains the inferior effect of irrigation on anaerobic N transformation to fertilization in the current study. The soil properties affecting anaerobic N transformation also changed significantly after the application of different irrigation methods. The AWD treatment caused significant effects of NH₄⁺-N, available P, and WC on anaerobic N transformations, with the contribution of NH₄⁺-N being particularly obvious (Figure 4c). The reason for this phenomenon is that, compared to CI, AWD increases aerobic conditions in the soil and inhibits denitrification and other anaerobic reactions, and the concentration of NH₄⁺-N, the substrate for nitrification, anammox, and Feammox, is prone to fluctuate under changing oxygen conditions, thus affecting the N transformation process. The effect of soil pH was not as significant as the change in WC, but was more important in controlling the anaerobic N transformation by affecting other nutrient contents. We believe that irrigation acts on anaerobic N transformation by affecting the soil environment, while soil biological factors are not as heavily involved.

4.4. The Relationship of the Interactions of Denitrification, Feammox, and Anammox together on Anaerobic N Transformations

In the present study, there was a significant positive correlation between the three anaerobic N transformation pathways (Figure 5), the interrelationship exceeding the effect of soil environment for all three pathways (Figure 6). Based on the pathways between the three anaerobic N transformations, the products and reaction substrate of these three pathways are closely linked to each other. Among them, the products of Feammox can produce NO₂⁻ and NO₃⁻ for denitrification and anammox, respectively [38]. A significantly positive correlation between Feammox and the other two pathways is evident in the present study and in previous studies [38,39]. It is worth noting that, while Feammox and anammox have a competitive relationship for NH₄⁺, the relationship between Feammox and anammox was positive in the present study. We speculated that the reason for this phenomenon is that the soil of the experimental site has high alkali-hydrolyzable N content, indicating that the soil has sufficient nutrients, so it does not cause strong competition between Feammox and anammox. Instead, the NO₃⁻ produced by Feammox can be utilized more effectively by anammox. Theoretically, denitrification and anammox would compete for NO₃⁻. However, in this study, denitrification and anammox were positively correlated and affected each other (Figure 5 and Figure 6). We speculate that this was due to the action of microorganisms. The extant microorganisms that can advance anammox all belong to Phyllostomycetes [40,41]. Some scholars believe that Phyllostomycetes may also have a certain denitrification function [37], and the connection through this microorganism gives these two anaerobic N transformation processes a certain synergistic effect. For the sampled soil with high inorganic N content, competition between the two pathways is alleviated.

5. Conclusions

In this study, soil denitrification, anammox, and Feammox rates and their contribution to soil N loss in different depths of a paddy soil with different fertilization and irrigation regimes were estimated. We showed that, as the soil deepened, denitrification, anammox, and Feammox rates successively reached their peaks. Denitrification contributed most heavily to soil N loss, followed by anammox and Feammox. Denitrification was also the most significant cause of soil N loss at various depths. We also noted that Feammox occurred at various depths in the paddy soil, and contributes to N loss and NH₄⁺ removal in paddy soils. Compared to depth, fertilization, and irrigation had a more significant effect on anaerobic N transformations. This study provides a theoretical basis on which to view N loss and agricultural measures in the agricultural system. This work is of great significance to the study of the N biogeochemical cycle.