Hole Application of Urea Inhibited Nitrification in the Zone around the Fertilizer Point by Reducing the Abundance of Nitrification Genes

Abstract

The present study investigated the interactions among nitrogen transformation and soil bacteria along the direction of diffusion of hole-applied urea. To this end, a lab incubation trial was conducted on sandy loam and silty loam soils. Soil bacterial communities were analyzed via 16S rRNA high-throughput sequencing, and soil chemical properties were measured at 8, 20, and 60 d after urea application. The treatments were the fertilizer point and 0–4 cm, 4–8 cm, 8–12 cm, and 12–16 cm horizontally distant from the fertilization point. They were designated FP, 0–4, 4–8, 8–12, and 12–16, respectively. The pre-culture and pre-incubation soil sample was used as a control. Soil NH₄⁺ concentration was the key factor influencing the soil bacterial community. For the sandy loam, the FP and 0–4 treatments reduced the putative abundance of amoA by 38.9–83.4% and 40.7–67.6%, amoB by 38.9–83.4% and 40.6–67.6%, and amoC by 41.1–84.1% and 43.6–69.9%, respectively, compared with the control group. For the silty loam, the FP and 0–4 treatments reduced the putative abundance of amoA by 85.0–87.3% and 28.9–82.6%, amoB by 84.6–87.2% and 29.1–82.5%, and amoC by 81.9–87.1% and 27.5–82.7%, respectively, compared with the control group. The fertilizer core region was <4 cm from the fertilizer point and maintained high NH₄⁺ concentrations for >60 d, which strongly inhibited nitrification. Overall, the fertilizer core region slowly released nitrogen and inhibited nitrification. For these reasons, hole application of urea may serve as a long-acting nitrogen fertilizer.

1. Introduction

Nitrogen (N) is an essential and limiting plant macronutrient and is vital to crop growth. It plays crucial roles in improving crop yield and quality [1]. Farmers often apply large amounts of chemical N fertilizer in agricultural production [2]. However, the amount used may greatly exceed actual crop demand [3]. Overuse of chemical N fertilizer may result in aquatic eutrophication [4] and the emission of greenhouse gases such as nitrous oxide (N₂O) [5,6]. To reduce the negative environmental impact of excessive N fertilizer application in agricultural production, the Chinese government has promoted a zero-growth policy for chemical fertilizer use and mandated other fertilizer conservation measures.

Loss of N fertilizer lowers nitrogen-use efficiency (NUE) and exacerbates agricultural non-point source pollution [7,8]. Nitrogen fertilizer that is not absorbed by crop roots may be lost to ammonia (NH₃) volatilization [9] and nitrate (NO₃⁻) runoff and leaching [10]. The latter may contaminate local groundwater. Therefore, N losses must be reduced during the effort to increase crop yield. Split application is the traditional N fertilizer method used by farmers [11]. However, this technique decreases NUE, increases N loss, has a negative environmental impact, and increases topdressing costs [12,13,14].

Implementation of the appropriate N fertilizer application method helps promote crop yield and has environmental benefits [15,16]. Hole application of N fertilizer delivers higher crop yield and better NUE than split N fertilizer application [1,6,17]. Previous studies demonstrated that one-time hole application of N fertilizer in the rhizosphere can meet the N demand for the entire crop growth period, reduce the number of N fertilizer applications per season, and lower topdressing costs [11,18]. N fertilizer hole application can also mitigate NH₃ volatilization, N₂O emission, NO₃⁻ leaching, and the risk of agricultural non-point source pollution [5].

Proper fertilization position is vital to the efficacy of hole application of N fertilizer [19]. After the fertilizer is applied to the soil, it dissolves and diffuses into the zone around the fertilizer point. Placement of the N fertilizer remote from the rhizosphere will prevent it from being absorbed in the early stages of crop growth [20]. By contrast, fertilizer application too near the rhizosphere could create high local fertilizer concentrations and chemically burn the seedlings. It is of practical significance to examine the transformation and migration of hole application of N fertilizer in the soil and evaluate the impact of this fertilization pattern on the reduction of N loss, mitigation of the negative environmental effects of N fertilizer overuse, and improvement of NUE.

Previous studies endeavored to identify the optimal distance between the fertilization position and the crop root system. Hole application of urea at 5–8 cm distance from rice, summer maize, and oilseed rape roots [1,18,21] prevented seedling burn and improved crop yield and NUE. Liu et al. [1] and Siddique et al. [22] found that hole application of N fertilizer can maintain high NH₄⁺ concentrations near the fertilization point for >60 d. They also discovered that a NH₄⁺ concentration gradient centered on the fertilization point decreased with distance from it. Hence, hole application of N fertilizer is long-acting, particularly at the early crop-growth stage. Furthermore, it promotes root proliferation, intensive root system formation around the fertilization point, and active plant nutrient uptake [23]. However, little information is available on the biochemical mechanisms of N transformation and migration of N fertilizer in response to hole N fertilizer application.

Soil microorganisms play important roles in soil N fertilizer transformation [24,25]. Soil bacteria are highly diverse, abundant, and multifunctional [26]. Soil bacteria participate in N fixation [27], ammonia oxidation [28], nitrification [25], and denitrification [24]. Changes in soil bacterial community and/or function can affect N fertilizer availability and pollution risk [29]. The inhibition of nitrification blocks nitrifier activity in the soil, reduces the soil nitrification rate [30], improves fertilizer NUE [31], and reduces N fertilizer loss and environmental degradation [32]. The nitrification inhibitor (e.g., dicyandiamide), combined with urea, can inhibit the activity of ammonia-oxidizing microorganisms by interfering with the utilization of the substrate by ammonia monooxygenase during the oxidation process [33]. Changes in the ambient environment caused by fertilization may influence soil bacterial community structure and function and alter nutrient availability [34,35]. The metabolic functions of the bacterial genes are implicated in soil N transformation and finally affect crop root N uptake [36,37]. However, previous studies focused mainly on the relationships between fertilizer dosage or type of fertilizer and soil bacteria under traditional fertilization management. By contrast, little is known about the interactions between N fertilizer and soil microorganisms in response to hole application of N fertilizer. Clarifying the biochemical mechanisms of N migration and transformation following hole application of N fertilizer may facilitate the optimization of fertilizer position, spacing, and other parameters.

Soil properties in the vicinity of the fertilizer point dramatically change in response to hole application of N fertilizer. Soil bacterial communities are highly sensitive to changes in their ambient environment [34,38]. It is well-known that high NH₄⁺ concentration is detrimental to plant and animal cells [39]. A previous study reported that high soil NH₄⁺ concentration had a toxic effect on the growth of soil microorganisms [40]. However, the effect of such a high NH₄⁺ concentration around the fertilizer point on soil microbes remains unclear. We hypothesized that high soil N concentrations would alter soil bacterial communities and inhibit the activity of ammonia-oxidizing microorganisms, thereby inhibiting nitrification. In the present work, we designed and executed an incubation trial to study the relationships among N transformation, N migration, and soil bacterial community structure and function in response to hole application of urea. We also conducted the trial on two different soil textures to broaden the applicability of this trial. We investigated the (1) effects of N diffusion on soil bacterial community and function, (2) scope and duration of the impact, and (3) relationships among soil nitrification and changes in soil bacterial community structure and function.

2. Materials and Methods

2.1. Soil Samples and Trial Design

The soils used in the incubation trial were derived from the topsoil (0–20 cm) of two typical farmlands in the Yangtze River Delta Region of China. One was located in Qidong (QD) County, Jiangsu Province (31.94° N, 121.73° E). Soybean-rapeseed (dryland) is the main crop rotation in this region. The soil texture was sandy loam (52.6% sand (0.05–2 mm), 33.0% silt (0.002–0.05 mm), and 14.4% clay (<0.002 mm)). The QD soil had pH = 7.61, soil organic matter (SOM) = 13.8 g kg⁻¹, total N = 0.65 g kg⁻¹, NH₄⁺-N = 7.44 mg kg⁻¹, NO₃⁻-N = 4.74 mg kg⁻¹, available P = 18.3 mg kg⁻¹, and available K = 237 mg kg⁻¹. The second was located in Dangtu (DT) County, Anhui Province (31.33° N, 118.60° E). Rice-rapeseed (paddy-upland) is the main crop rotation in this region. The soil texture was silty loam (31.0% sand (0.05–2 mm), 54.4% silt (0.002–0.05 mm), and 14.6% clay (<0.002 mm)). In the 0–20 cm soil layer, the soil pH, SOM, total N, NH₄⁺-N, NO₃⁻-N, available P, and available K were 7.02, 28.7 g kg⁻¹, 1.81 g kg⁻¹, 8.64 mg kg⁻¹, 2.28 mg kg⁻¹, 14.3 mg kg⁻¹, and 116 mg kg⁻¹, respectively. Air-dried soil samples were passed through 2 mm mesh sieves to remove visible stones and plant residues in preparation for the subsequent lab incubation trial.

The incubation trial was performed in a custom-made wooden box 5 cm high × 9 cm wide × 25 cm long. Nine hundred grams of dried soil were added to the box, and the soil water content was adjusted with sterile distilled water to 40% of the maximum water-holding capacity. The soil was pre-cultured in an incubator in the dark at 26 ± 1 °C for 4 d to activate soil microorganisms. The pre-cultured soil was then placed in a self-sealing bag and manually shaken repeatedly to mix the soil and water and reduce experimental error caused by soil and water heterogeneity. The soil was then returned to the box, and the soil water content was adjusted with sterile distilled water to 60% of the maximum soil water-holding capacity. The weighing method was used for this purpose.

Four grams of urea (>99.0% purity, Sinopharm Chemical Reagent Co., Ltd., Beijing, China) were hole-applied at 3 cm to the left of the culture box and at 2 cm depth. The four grams of urea were turned into a cylinder (2 cm in diameter × 1 cm high) by a granulator to simplify the process of hole application. The cylindrical urea was placed in a polyvinyl chloride pipe (2 cm in diameter and functions like an injection syringe) and pressed to the pipe bottom using a piston. A ruler was used to accurately check for the urea position. In the direction of horizontal urea diffusion, the fertilizer point was taken as the zero point, and the 4 cm width was taken as the layer-by-layer sampling unit. Each sampling layer was considered a treatment, and there were five sampling layers per incubation box. The treatments were the fertilizer point and 0–4 cm, 4–8 cm, 8–12 cm, and 12–16 cm horizontally distant from the fertilization point. They were designated FP, 0–4, 4–8, 8–12, and 12–16, respectively. The part that was 1 cm away from the inner wall of the box was discarded at the sampling time to avoid border effect. The soil-sampling process is shown in Figure 1. The incubation box was sealed with plastic wrap perforated with five small holes to create aerobic conditions while preventing rapid evaporation. The soil samples were destructively collected on days 8, 20, and 60 after the hole application of urea. There were three replicates per sampling period and 18 culture boxes for both soil types. The soil on culture day 0 (pre-culture and pre-incubation) was taken as the control. Water lost by evaporation was replenished with sterile distilled water every 4 d by the weighing method. The soil moisture content was maintained at 60% of the soil maximum water-holding capacity throughout the incubation trial. Each soil sample was divided into two portions. One was stored at −40 °C until DNA extraction and high-throughput sequencing, while the other was stored at 4 °C for soil chemical property analysis.

2.2. Soil Physical and Chemical Analysis

The detailed procedure for measuring the water-holding capacity was as follows: (1) A ring-sampler (ring-samplerA, 100 cm³) was used to collect undisturbed soil and was filled with the soil sample. (2) The ring-samplerA was covered with a piece of filter paper from the bottom and soaked in water for one day to saturate the soil with moisture. (3) Another soil sample was collected from the same soil layer and passed through a 1 mm sieve after air-drying; then, this sample was used to fill another ring-sampler (ring-samplerB, 100 cm³). (4) Ring-samplerA (with the filter paper) was placed on ring-samplerB while ensuring close contact between the two. (5) After 8 h, an aluminum box was used to extract 10–20 g of soil, and the soil water content in the aluminum box was determined (i.e., maximum water-holding capacity).

The pH of 1:2.5 soil:distilled water suspensions were measured with a pH meter (No. FE28, Mettler Toledo, Zurich, Switzerland). Soil ammonium (NH₄⁺-N) and nitrate (NO₃⁻-N) were extracted from 1:8 soil:water suspensions with 2 M KCl (>99.0% purity, Sinopharm Chemical Reagent Co., Ltd., Beijing, China). The suspensions were shaken for 1 h, and the supernatants were passed through filter paper. The NH₄⁺-N and NO₃⁻-N concentrations were measured with a continuous-discrete analyzer (SmartChem 200; Westco Scientific Instruments, Brookfield, CT, USA).

2.3. DNA Extraction and Bacterial 16S rRNA Gene Amplification

The DNA in 0.5 g soil samples from the incubation trial was extracted with a FastDNA SPIN Kit for Soil (MP Biomedicals LLC, Santa Ana, CA, USA) according to the manufacturer’s instructions. Genomic DNA quality was checked on 1% agarose gels, and DNA concentration and purity were determined by UV–Vis spectrophotometry (NanoDrop ND-2000; Thermo Fisher Scientific, Wilmington, DE, USA). Qualified DNA samples were delivered to Guangdong MAGIGENE Technology Co., Ltd. (Shenzhen, China) for high-throughput sequencing of the 16S rRNA gene on an Illumina MiSeq PE 300 platform (Illumina, San Diego, CA, USA). The bacterial 16S rRNA V4-V5 region was amplified with 515F and 907R primers [41].

2.4. Illumina Sequencing Data Processing

Raw sequences were merged with FLASH (Fast Length Adjustment of Short Reads) software (http://0-ccb-jhu-edu.brum.beds.ac.uk/software/FLASH/ accessed on 21 April 2022) [42]. Bioinformatics analysis of the high-throughput sequencing data was conducted in the QIIME (Quantitative Insights into Microbial Ecology) v. 1.9.1 (qiime.org) pipeline [43]. Low-quality sequences (average quality score <25) and those <300 bp in length were discarded. Chimeras were detected and filtered by the UCHIME algorithm (http://drive5.com/usearch/manual8.1/uchime_algo.html accessed on 10 June 2022) [44]. Qualified sequences from different samples were clustered into operational taxonomic units (OTUs) at 97% sequence identity by the UCLUST method (http://www.drive5.com/usearch/manual/uclust_algo.html accessed on 11 June 2022). The most abundant read in each OTU was selected as the representative sequence. Taxonomic annotations were assigned to bacterial OTUs against the SILVA 132 database (https://www.arb-silva.de accessed on 11 June 2022) by the UCLUST method. All representative sequences were aligned with SILVA reference sequences using the Python Nearest Alignment Space Termination (PyNAST) tool (https://github.com/biocore/pynast accessed on 12 June 2022). Non-bacterial sequences were discarded, and 724,194 high-quality bacterial sequences were obtained. The remaining sequences from all samples were rarefied to the same sequencing depth (5562) for β-diversity comparisons based on the minimum sequence number. Two α-diversity indices, including Faith’s phylogenetic diversity index (PD) [45] and Shannon diversity index [46], were calculated by QIIME. Sequences obtained in this study were submitted in the National Genomics Data Center (NGDC) Genome Sequence Archive (https://bigd.big.ac.cn/gsub/ accessed on 7 September 2022) with accession number CRA008085.

2.5. Statistical Analysis

One-way analysis of variance (ANOVA) was run in SPSS v. 17.0 (IBM Corp., Armonk, NY, USA) to determine differences among treatments in terms of their soil chemical properties. Non-metric multidimensional scaling (NMDS) ordination plots based on Bray–Curtis dissimilarities were used to display differences among treatments in terms of their bacterial community compositions. Permutational multivariate analysis of variance (PERMANOVA) was used to determine significant differences among treatments in terms of their bacterial communities using the “Adonis” function in R v. 4.0.2 (http://cran.r-project.org accessed on 16 June 2022). Correlations among soil chemical properties and bacterial communities were determined by a Mantel test. The NMDS, PERMANOVA, and Mantel test were implemented using the “vegan” package v. 2.4-2 in R v. 4.0.2. Relationships among soil chemical properties and bacterial α-diversity were determined by Spearman’s correlation analysis in R v. 4.0.2. Bacterial functional genes were generated using the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) program (https://github.com/picrust/picrust2 accessed on 14 June 2022) based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (https://www.genome.jp/kegg/ accessed on 14 June 2022) [27,37].

3. Results

3.1. Soil Mineral Nitrogen and pH

Figure 2a and Figure 3a show the distribution of horizontally migrating NH₄⁺-N following hole application of urea at different incubation times. For QD at the same incubation time, the highest NH₄⁺ concentration was observed in the FP treatment. The soil NH₄⁺ concentration gradually decreased with increasing distance from the fertilizer point. With increasing incubation time, the soil NH₄⁺ concentration decreased in the FP, 0–4, and 4–8 treatments but initially increased and then decreased in the 8–12 and 12–16 treatments. In each incubation period, the soil NH₄⁺ concentration was higher in the 12–16 treatment than the control (3.18 mg kg⁻¹). The distribution of the NH₄⁺ concentration in the DT soil resembled that for the QD soil. At the same incubation time, however, the NH₄⁺ concentration was higher in the DT soil than the QD soil except for the 12–16 treatment at day 8.

Figure 2b and Figure 3b show the changes in the content of horizontally migrating soil NO₃⁻ following hole application of urea at different incubation times. For QD at the same incubation time, the soil NO₃⁻ content increased and then decreased with increasing distance from the fertilizer point. The soil NO₃⁻ content was highest in the 8–12 treatment. By day 8, the soil NO₃⁻ content was low (<10 mg kg⁻¹) in the FP, 0–4, and 12–16 treatments, While the soil NO₃⁻ content was higher in the 4–8 and 8–12 treatments than the control (6.78 mg kg⁻¹). The soil NO₃⁻ concentrations in the 8–12 and 12–16 treatments were significantly higher on day 20 than on day 8. For all treatments, the soil NO₃⁻ concentration was significantly higher on day 60 than it was on days 8 and 20. For DT, the soil NO₃⁻ concentration in the control was 12.3 mg kg⁻¹. The soil NO₃⁻ concentration was low (<10 mg kg⁻¹) in all treatments on day 8. By day 20, the soil NO₃⁻ content was low (<12 mg kg⁻¹) in the FP, 0–4, and 4–8 treatments, while a certain amount of NO₃⁻ was produced in the 8–12 (99.8 mg kg⁻¹) and 12–16 (124 mg kg⁻¹) treatments. After 60 d fertilization, the soil NO₃⁻ concentration increased with increasing distance from the fertilizer point. The maximum NO₃⁻ concentration was measured for the 12–16 treatment (207 mg kg⁻¹). For all treatments, the soil NO₃⁻ concentrations were higher in the QD soil than the DT soil after 60 d fertilization.

Figure 2c and Figure 3c show that for QD at the same incubation time, the FP treatment had the highest soil pH value. However, the pH gradually decreased with increasing distance from the fertilizer point. With increasing incubation time, the pH decreased in all treatments. The soil pH in each treatment was higher than that of the control (7.78) except for the 8–12 and 12–16 treatments at day 60. Both the QD and DT soils presented with similar pH-change patterns.

3.2. Bacterial Community Composition and α-Diversity

The responses of the six most abundant bacterial phyla to the various treatments are shown in

Table 1. The abundance of dominated bacteria phyla under different treatments and incubation times for Qidong (QD) soil. Data are means ± SE, n = 3.

Treatments	Proteobacteria (%)	Bacteroidetes (%)	Acidobacteria (%)	Actinobacteria (%)	Firmicutes (%)	Chloroflexi (%)
Control Day 8 ‡	39.4 ± 1.4	23.1 ± 0.91	9.91 ± 0.29	6.01 ± 0.59	13.3 ± 0.86	3.04 ± 0.16
FP	8.20 ± 0.7 d	2.53 ± 0.33 c	2.81 ± 0.47 c	3.22 ± 0.21 b	80.6 ± 1.9 a	0.82 ± 0.11 b
0–4	12.7 ± 0.3 c	4.51 ± 0.87 b	3.49 ± 0.16 c	5.15 ± 0.78 b	70.6 ± 1.4 a	1.13 ± 0.05 b
4–8	17.9 ± 2.0 b	7.38 ± 1.45 b	6.79 ± 0.52 b	5.15 ± 0.58 b	57.7 ± 3.4 b	1.63 ± 0.01 b
8–12	33.2 ± 1.6 a	23.9 ± 1.7 a	12.2 ± 0.1 a	9.16 ± 1.27 a	11.5 ± 1.0 c	3.54 ± 0.03 a
12–16	36.8 ± 1.2 a	25.9 ± 1.8 a	12.0 ± 0.5 a	8.57 ± 1.11 a	7.36 ± 0.75 c	3.11 ± 0.63 a
Day 20 §
FP	7.6 ± 0.12 c	1.85 ± 0.25 d	1.34 ± 0.17 c	5.99 ± 0.94 b	80.4 ± 1.5 a	1.02 ± 0.10 b
0–4	10.0 ± 0.3 c	4.72 ± 1.53 cd	1.62 ± 0.15 c	6.65 ± 0.53 b	73.3 ± 1.5 a	1.20 ± 0.08 b
4–8	19.5 ± 2.8 b	8.41 ± 0.81 bc	5.59 ± 1.88 b	7.50 ± 0.87 ab	53.3 ± 6.7 b	1.87 ± 0.70 b
8–12	32.0 ± 0.8 a	13.8 ± 4.07 ab	10.5 ± 0.74 a	8.55 ± 1.00 ab	26.4 ± 5.0 c	3.05 ± 0.34 a
12–16	36.0 ± 2.8 a	20.0 ± 1.07 a	11.6 ± 0.78 a	10.8 ± 1.70 a	11.5 ± 0.9 d	3.60 ± 0.36 a
Day 60 †
FP	12.6 ± 0.4 e	7.23 ± 2.21 b	1.47 ± 0.26 c	10.3 ± 1.9 b	63.7 ± 2.2 a	1.00 ± 0.20 b
0–4	17.6 ± 0.9 d	9.26 ± 1.19 b	1.04 ± 0.30 c	17.8 ± 2.7 a	44.9 ± 1.7 b	1.27 ± 0.05 b
4–8	23.2 ± 1.1 c	20.2 ± 3.4 a	2.01 ± 0.23 c	13.1 ± 1.4 ab	33.5 ± 4.1 c	2.29 ± 0.46 b
8–12	38.5 ± 1.5 a	21.6 ± 2.9 a	4.29 ± 0.54 b	10.4 ± 0.4 b	16.7 ± 2.2 d	3.90 ± 0.91 a
12–16	32.5 ± 0.7 b	21.9 ± 0.4 a	10.2 ± 0.62 a	9.60 ± 0.9 b	14.5 ± 0.4 d	4.64 ± 0.32 a

Different letters in the same column for each treatment indicate significant differences according to one-way ANOVA (LSD, p < 0.05) when comparing treatments within the same incubation time. Control: samples at Day 0 (pre-culture and pre-incubation). ^‡: Day 8, samples at 8 days after urea application. ^§: Day 20, samples at 20 days after urea application. ^†: Day 60, samples at 60 days after urea application.

and

Table 2. The abundance of dominated bacteria phyla under different treatments and incubation times for Dangtu (DT) soil. Data are means ± SE, n = 3.

Treatments	Proteobacteria (%)	Bacteroidetes (%)	Acidobacteria (%)	Actinobacteria (%)	Firmicutes (%)	Chloroflexi (%)
Control Day 8 ‡	34.2 ± 0.9	34.1 ± 1.4	2.20 ± 0.20	6.64 ± 0.85	15.6 ± 0.5	5.03 ± 0.52
FP	13.8 ± 2.0 d	3.53 ± 0.2 c	1.13 ± 0.29 c	3.93 ± 0.58 cd	72.8 ± 3.4 a	3.12 ± 0.41 cd
0–4	12.6 ± 2.2 d	4.77 ± 1.3 c	1.06 ± 0.13 c	3.25 ± 0.57 d	74.6 ± 4.6 a	2.34 ± 0.23 d
4–8	23.8 ± 2.6 c	8.51 ± 1.4 c	1.94 ± 0.21 b	5.49 ± 0.79 bc	53.6 ± 4.4 b	4.17 ± 0.19 bc
8–12	35.3 ± 1.9 b	22.5 ± 3.5 b	2.85 ± 0.09 a	6.70 ± 0.15 ab	23.3 ± 3.8 c	5.99 ± 0.7 a
12–16	42.6 ± 1.3 a	32.0 ± 1.1 a	2.53 ± 0.29 ab	7.86 ± 0.49 a	6.40 ± 0.2 d	5.19 ± 0.17 ab
Day 20 §
FP	5.26 ± 0.4 b	1.10 ± 0.1 d	0.28 ± 0.02 c	2.11 ± 0.28 b	89.0 ± 0.9 a	1.37 ± 0.10 c
0–4	7.94 ± 0.5 b	2.83 ± 0.8 d	0.37 ± 0.01 c	2.20 ± 0.24 b	84.2 ± 0.8 a	1.28 ± 0.12 c
4–8	30.4 ± 5.6 a	13.9 ± 3.5 c	1.40 ± 0.32 b	6.35 ± 0.37 a	41.8 ± 9.9 b	3.54 ± 0.71 b
8–12	36.5 ± 0.4 a	24.5 ± 0.9 b	3.00 ± 0.23 a	6.03 ± 0.46 a	19.5 ± 1.3 c	6.50 ± 0.17 a
12–16	37.7 ± 0.7 a	30.2 ± 1.2 a	3.16 ± 0.27 a	7.48 ± 0.96 a	12.0 ± 0.9 c	5.77 ± 0.44 a
Day 60 †
FP	12.0 ± 2.6 b	15.8 ± 5.9 a	0.14 ± 0.03 b	6.77 ± 1.52 a	61.5 ± 6.9 a	0.77 ± 0.05 c
0–4	18.1 ± 3.4 b	15.3 ± 5.2 a	0.26 ± 0.06 b	8.75 ± 2.93 a	53.2 ± 3.7 ab	1.26 ± 0.33 bc
4–8	30.0 ± 2.9 a	12.7 ± 1.2 a	0.40 ± 0.11 b	8.40 ± 1.47 a	43.4 ± 4.6 bc	2.18 ± 0.36 b
8–12	34.8 ± 1.1 a	17.6 ± 1.1 a	1.63 ± 0.38 a	6.76 ± 0.55 a	31.1 ± 2.5 c	4.78 ± 0.64 a
12–16	37.1 ± 1.2 a	25.6 ± 0.9 a	2.18 ± 0.35 a	9.17 ± 0.56 a	16.6 ± 0.7 d	5.28 ± 0.26 a

Different letters in the same column for each treatment indicate significant differences according to one-way ANOVA (LSD, p < 0.05) when comparing treatments within the same incubation time. Control: samples at Day 0 (pre-culture and pre-incubation). ^‡: Day 8, samples at 8 days after urea application. ^§: Day 20, samples at 20 days after urea application. ^†: Day 60, samples at 60 days after urea application.

. Proteobacteria, Bacteroidetes, Acidobacteria, Actinobacteria, Firmicutes, and Chloroflexi predominated and accounted for >91.7% and >96.0% of the bacterial sequences in the QD and DT soils, respectively. For QD, the relative abundances of Proteobacteria, Bacteroidetes, Acidobacteria, Actinobacteria, and Chloroflexi in FP, 0–4, and 4–8 treatments were less than those in 8–12 and 12–16 treatments throughout the entire incubation period (except for Actinobacteria at day 60). The relative abundance of Firmicutes decreased with increasing distance from the fertilizer point at the same incubation time. With increasing incubation time, the relative abundance of Proteobacteria, Bacteroidetes, Acidobacteria, Actinobacteria, and Chloroflexi increased in the FP, 0–4, and 4–8 treatments. Although the relative abundance of Firmicutes in FP, 0–4, and 4–8 treatments decreased over incubation time, Firmicutes predominated in these treatments. In addition, the relative abundance of the dominant phyla in FP, 0–4, and 4–8 treatments greatly changed compared with the control, especially at days 8 and 20. In contrast, the relative abundance of the dominant phyla in 8–12 and 12–16 treatments changed slightly. The change patterns in the relative abundance of the dominant phyla for DT soil were similar to those for QD soil.

The results of one-way ANOVA of the soil bacterial diversity indices are shown in Figure 4 and Figure 5. The bacterial α-diversity was affected by treatments and incubation time. For QD in the control, the PD index was 71.2, and the Shannon index was 9.05. The PD and Shannon indices in FP, 0–4, and 4–8 treatments were less than those in 8–12 and 12–16 throughout the trial. The 12–16 treatment increased PD index by 0.33–5.50% and Shannon index by 2.02–4.56%, respectively, compared with the control throughout the trial. In contrast, the FP, 0–4, and 4–8 treatments decreased PD and Shannon indices compared with the control. In addition, the bacterial α-diversity did not regularly change with increasing incubation time under the same treatment. For DT in the control, the PD index was 62.2, and the Shannon index was 7.80. The bacterial diversity indices in FP, 0–4, and 4–8 treatments were less than those in 8–12 and 12–16 throughout the trial (except for Shannon index at day 8). The 12–16 treatment increased PD and Shannon indices throughout the incubation period, with increases of 1.14–7.00% and 5.33–12.8%, respectively, compared with the control. The FP, 0–4, and 4–8 treatments reduced the PD index by 14.3–31.0%, 17.1–26.6%, and 4.79–12.0%, respectively.

PERMANOVA indicated that incubation time (QD: R² = 0.21, p = 0.001; DT: R² = 0.19, p = 0.001) and treatment (QD: R² = 0.37, p = 0.001; DT: R² = 0.43, p = 0.001) significantly influenced bacterial community structure. We used NMDS analysis (Figure 6a,b) based on the Bray–Curtis distance to evaluate the soil bacterial community structure under different treatments. For QD at the same incubation time, the soil bacterial community horizontally migrated in the direction of the hole application of urea, and all samples from the same treatment could be combined. There was minimal differences among the 12–16 treatment and the control in terms of their bacterial community structures under the same incubation time. There was substantial differences between the FP treatment and the control in terms of their bacterial community structures. Moreover, the difference in the bacterial communities within the same treatment increased with incubation time. The change trends in the bacterial community structure were similar for both the DT and QD soils.

3.3. Putative Nitrification Genes

We used PICRUSt2 to predict the putative abundances of nitrification genes under different treatments (

Table 3. The putative abundance of nitrification genes under different treatments and incubation times for Qidong (QD) soil.

Treatments	amoA (%)	amoB (%)	amoC (%)	hao (%)
Control	2.01 × 10−3	2.01 × 10−3	4.16 × 10−3	0.38 × 10−3
Day 8 ‡
FP	0.33 × 10−3 b	0.33 × 10−3 b	0.66 × 10−3 c	0.08 × 10−3 d
0–4	0.75 × 10−3 b	0.75 × 10−3 b	1.42 × 10−3 bc	0.09 × 10−3 d
4–8	1.04 × 10−3 b	1.04 × 10−3 b	2.07 × 10−3 b	0.25 × 10−3 c
8–12	2.40 × 10−3 a	2.40 × 10−3 a	4.86 × 10−3 a	0.71 × 10−3 a
12–16	2.54 × 10−3 a	2.54 × 10−3 a	5.12 × 10−3 a	0.48 × 10−3 b
Day 20 §
FP	0.54 × 10−3 b	0.54 × 10−3 b	1.05 × 10−3 b	0.07 × 10−3 b
0–4	0.65 × 10−3 b	0.65 × 10−3 b	1.25 × 10−3 b	0.10 × 10−3 b
4–8	1.37 × 10−3 b	1.37 × 10−3 b	2.72 × 10−3 b	0.41 × 10−3 b
8–12	3.36 × 10−3 a	3.36 × 10−3 a	6.80 × 10−3 a	1.72 × 10−3 a
12–16	3.29 × 10−3 a	3.29 × 10−3 a	6.53 × 10−3 a	1.66 × 10−3 a
Day 60 †
FP	1.23 × 10−3 c	1.23 × 10−3 c	2.45 × 10−3 c	0.12 × 10−3 c
0–4	1.20 × 10−3 c	1.20 × 10−3 c	2.34 × 10−3 c	0.11 × 10−3 c
4–8	3.72 × 10−3 bc	3.72 × 10−3 bc	7.53 × 10−3 b	2.97 × 10−3 b
8–12	7.94 × 10−3 a	7.94 × 10−3 a	16.1 × 10−3 a	6.48 × 10−3 a
12–16	6.58 × 10−3 ab	6.58 × 10−3 ab	13.3 × 10−3 a	3.42 × 10−3 b

Different letters in the same column for each treatment indicate significant differences according to one-way ANOVA (LSD, p < 0.05) when comparing treatments within the same incubation time. Putative abundance is the proportion of each nitrification gene to the total number of genes in the sample. Data are mean value of three repetitions of the same treatment. Control: samples at Day 0 (pre-culture and pre-incubation). ^‡: Day 8, samples at 8 days after fertilization. ^§: Day 20, samples at 20 days after fertilization. ^†: Day 60, samples at 60 days after fertilization.

and

Table 4. The putative abundance of nitrification genes under different treatments and incubation times for Dangtu (DT) soil.

Treatments	amoA (%)	amoB (%)	amoC (%)	hao (%)
Control	2.59 × 10−3	2.60 × 10−3	4.99 × 10−3	0.22 × 10−3
Day 8 ‡
FP	0.39 × 10−3 b	0.40 × 10−3 b	0.90 × 10−3 b	0.11 × 10−3 c
0–4	0.54 × 10−3 b	0.54 × 10−3 b	1.09 × 10−3 b	0.09 × 10−3 c
4–8	0.73 × 10−3 b	0.74 × 10−3 b	1.42 × 10−3 b	0.13 × 10−3 b
8–12	1.81 × 10−3 a	1.83 × 10−3 a	3.47 × 10−3 a	0.23 × 10−3 a
12–16	1.64 × 10−3 a	1.65 × 10−3 a	3.20 × 10−3 a	0.19 × 10−3 ab
Day 20 §
FP	0.33 × 10−3 c	0.33 × 10−3 c	0.64 × 10−3 c	0.06 × 10−3 c
0–4	0.45 × 10−3 c	0.46 × 10−3 c	0.86 × 10−3 c	0.04 × 10−3 c
4–8	1.49 × 10−3 b	1.50 × 10−3 b	2.75 × 10−3 b	0.16 × 10−3 bc
8–12	2.50 × 10−3 a	2.50 × 10−3 a	4.72 × 10−3 a	0.85 × 10−3 a
12–16	2.94 × 10−3 a	2.95 × 10−3 a	5.63 × 10−3 a	0.32 × 10−3 b
Day 60 †
FP	0.36 × 10−3 c	0.37 × 10−3 c	0.81 × 10−3 c	0.08 × 10−3 c
0–4	1.84 × 10−3 b	1.85 × 10−3 b	3.62 × 10−3 b	0.73 × 10−3 b
4–8	2.66 × 10−3 b	2.67 × 10−3 b	5.30 × 10−3 b	1.71 × 10−3 a
8–12	2.47 × 10−3 b	2.48 × 10−3 b	4.73 × 10−3 b	1.01 × 10−3 ab
12–16	4.91 × 10−3 a	4.92 × 10−3 a	9.57 × 10−3 a	2.05 × 10−3 a

Different letters in the same column for each treatment indicate significant differences according to one-way ANOVA (LSD, p < 0.05) when comparing treatments within the same incubation time. Putative abundance is the proportion of each nitrification gene to the total number of genes in the sample. Data are mean value of three repetitions of the same treatment. Control: samples at Day 0 (pre-culture and pre-incubation). ^‡: Day 8, samples at 8 days after fertilization. ^§: Day 20, samples at 20 days after fertilization. ^†: Day 60, samples at 60 days after fertilization.

). The KEGG database returned 7992 functional genes, of which 4 regulated nitrification. The putative abundances of nitrification genes were affected by treatment and incubation time. For QD, the putative abundance of nitrification genes in FP and 0–4 treatments were significantly less than those in 8–12 and 12–16 treatments throughout the trial. The FP and 0–4 treatments reduced the putative abundance of amoA by 38.9–83.4% and 40.7–67.6%, amoB by 38.9–83.4% and 40.6–67.6%, amoC by 41.1–84.1% and 43.6–69.9%, and hao by 67.0–80.2% and 70.1–75.4%, respectively, compared with the control group. The 4–8 treatment reduced the putative abundance of amoA, amoB, and amoC at days 8 and 20 while increasing the putative abundance of these genes at day 60. The 8–12 and 12–16 treatments increased the putative abundance of nitrification genes throughout the trial. Under the same treatment, the putative abundance of nitrification genes at day 60 were higher than those at days 8 and 20. For DT, the FP and 0–4 treatments reduced the putative abundance of amoA by 85.0–87.3% and 28.9–82.6%, amoB by 84.6–87.2% and 29.1–82.5%, and amoC by 81.9–87.1% and 27.5–82.7%, respectively, compared with the control group. The 12–16 treatment increased the putative abundance of nitrification genes at day 60 compared with the control group. In addition, the putative abundance of nitrification genes in 4–8, 8–12, and 12–16 treatments for QD were higher than those for DT at day 60.

3.4. Relationships among Soil Bacteria and Soil Chemical Properties

Mantel analysis (

Table 5. The correlations (r) between soil chemical properties and bacterial communities determined by Mantel test.

Treatments	QD		DT
	r	p	r	p
Day 8 ‡
NH4+-N	0.82	0.001 ***	0.67	0.001 ***
NO3−-N	0.01	0.14	0.12	0.098
pH	0.83	0.001 ***	0.74	0.001 ***
Day 20 §
NH4+-N	0.49	0.001 ***	0.38	0.002 **
NO3−-N	0.22	0.023 *	0.49	0.001 ***
pH	0.66	0.001 ***	0.32	0.004 **
Day 60 †
NH4+-N	0.86	0.001 ***	0.67	0.001 ***
NO3−-N	0.74	0.001 ***	0.68	0.001 ***
pH	0.82	0.001 ***	0.75	0.001 ***

*, **, and *** indicate statistical significance at p < 0.05, <0.01, and <0.001 levels, respectively, tested by the Mantel test. QD, Qidong soil; DT, Dangtu soil. ^‡: Day 8, samples at 8 days after fertilization. ^§: Day 20, samples at 20 days after fertilization. ^†: Day 60, samples at 60 days after fertilization.

) revealed the relationships among the soil bacterial community and the soil chemical properties. For the both soils, the bacterial community was significantly positively correlated with soil NH₄⁺ content and pH throughout the entire incubation period. The soil bacterial community was significantly positively correlated with soil NO₃⁻ content at days 20 and 60.

We used a Spearman’s correlation analysis to study the relationships among the nitrification gene abundances and the soil chemical properties (

Table 6. The Spearman’s correlations (r) between soil properties and nitrification genes at different incubation times for Qidong (QD) soil.

Treatments	amoA	amoB	amoC	hao
Day 8 ‡
NH4+-N	−0.89 **	−0.89 **	−0.89 **	−0.79 **
NO3−-N	0.56 *	0.56 *	0.56 *	0.76 **
pH	−0.92 **	−0.92 **	−0.93 **	−0.77 **
Day 20 §
NH4+-N	−0.51 *	−0.51 *	−0.53 *	−0.66 *
NO3−-N	0.78 **	0.78 **	0.78 **	0.81 **
pH	−0.66 *	−0.66 *	−0.65 *	−0.79 **
Day 60 †
NH4+-N	−0.83 **	−0.83 **	−0.83 **	−0.74 **
NO3−-N	0.85 **	0.85 **	0.85 **	0.88 **
pH	−0.85 **	−0.85 **	−0.85 **	−0.80 **

* and ** indicate statistical significance at p < 0.05 and <0.01 levels, respectively. ^‡: Day 8, samples at 8 days after fertilization. ^§: Day 20, samples at 20 days after fertilization. ^†: Day 60, samples at 60 days after fertilization.

and

Table 7. The Spearman’s correlations (r) between soil properties and nitrification genes at different incubation times for Dangtu (DT) soil.

Treatments	amoA	amoB	amoC	hao
Day 8 ‡
NH4+-N	−0.82 **	−0.79 **	−0.79 **	−0.76 **
NO3−-N	−0.36	−0.33	−0.33	−0.41
pH	−0.88 **	−0.86 **	−0.86 **	−0.75 **
Day 20 §
NH4+-N	−0.86 **	−0.86 **	−0.84 **	−0.69 **
NO3−-N	0.64 *	0.64 *	0.64 *	0.57 *
pH	−0.89 **	−0.89 **	−0.90 **	−0.62 *
Day 60 †
NH4+-N	−0.69 **	−0.69 **	−0.69 **	−0.64 *
NO3−-N	0.71 **	0.71 **	0.71 **	0.63 *
pH	−0.64 *	−0.64 *	−0.64 *	−0.64 *

* and ** indicate statistical significance at p < 0.05 and <0.01 levels, respectively. ^‡: Day 8, samples at 8 days after fertilization. ^§: Day 20, samples at 20 days after fertilization. ^†: Day 60, samples at 60 days after fertilization.

). For QD, there were significant negative correlations between soil NH₄⁺ concentration and the putative abundances of amoA, amoB, amoC, and hao and between soil pH and the abundances of these genes throughout the entire incubation period. Soil NO₃⁻ content was positively correlated with the relative abundances of amoA, amoB, amoC, and hao at days 20 and 60. The relationships among the nitrification gene abundances and the soil chemical properties were similar for both the DT and QD soils.

The results of the putative abundances of genes encoding the outer membrane protein regulator family are shown in Tables S1 and S2. For the both soils, the FP and 0–4 treatments reduced the putative abundances of genes encoding osmotic stress response and tricarboxylic acid transport while enhancing the putative abundance of gene encoding secret stress response (K07770) throughout the trial.

4. Discussion

4.1. Migration Characteristics of the Hole Application of Urea

The distribution characteristics of NH₄⁺-N in the soil disclosed that a nitrogen concentration gradient forms in the direction that the urea migrates (Figure 1a and Figure 2a). This observation is consistent with those of previous investigations [1,22]. Only a small amount of urea was detected after 4 d, and no urea was detected after 8 d in the FP and 0–4 treatments (data not shown). After the hole application, then, the urea rapidly and thoroughly hydrolyzed in the soil. Therefore, soil NH₄⁺-N became the main factor affecting soil microorganisms and N transformation in the subsequent cultivation process. We also found that the change trends for the soil pH and NH₄⁺ concentration were mutually consistent. High soil pH persisted in the FP and 0–4 treatments, especially at 8 d after fertilization. Urea hydrolysis generates ammonium bicarbonate, which is alkaline and rapidly increases the soil pH. As NH₄⁺-N migrates and forms a concentration gradient, the soil pH gradually decreases in the direction of NH₄⁺-N migration. Here, the FP and 0–4 treatments maintained high NH₄⁺ concentrations for >60 d. This finding was consistent with those of previous studies. Liu et al. [1] reported >400 mg kg⁻¹ soil NH₄⁺ near the fertilization zone at 60 d after hole application of urea.

We observed that there was strong nitrification inhibition in the FP and 0–4 treatments during the incubation trial. This might be attributed to the toxic effect of high NH₄⁺ concentration on soil-nitrifying microorganisms [40]. A strong nitrification inhibition was also observed in the 4–8 treatment at days 8 and 20; however, the inhibitory effects of high NH₄⁺ concentrations on nitrification weakened as the incubation time increased. This might be related to the decrease of NH₄⁺ concentration in the 4–8 treatment over time. In addition, the NH₄⁺ concentrations of QD soil were less than those of DT soil under the same treatments (except for 12–16 treatment on day 8), which might be related to the difference in soil texture. The finer-textured DT soil (silty loam) generally had more negative charges compared with the QD soil (sandy loam), so it had a stronger NH₄⁺-fixation capacity. This might be an important factor affecting the inhibitory effect of hole application of urea.

4.2. Effects of N Migration on the Soil Bacterial Community

Proteobacteria, Bacteroidetes, Acidobacteria, and Actinobacteria are widely distributed in farmland soil. In the present study, we also found that Proteobacteria, Bacteroidetes, Acidobacteria, and Actinobacteria were the dominant soil bacterial phyla in the control. Interestingly, N migration following hole application of urea significantly affected the soil bacterial community composition. The soil NH₄⁺ concentration was the key factor affecting the abundance of dominant soil bacterial phyla. High NH₄⁺ concentration inhibited the growth of Proteobacteria, Bacteroidetes, and Acidobacteria while promoting the growth of Firmicutes. Under the FP, 0–4, and 4–8 treatments, the relative abundance of Firmicutes sharply increased, and this bacterial phylum predominated. For the FP treatment, its abundances even reached 80.4% and 89.0% on d 20 for QD and DT soil, respectively. This is because Firmicutes are eutrophic [47], grow rapidly, and tolerate better the high pH and N concentrations [38,48]. However, high NH₄⁺ concentrations in the FP, 0–4, and 4–8 treatments increased osmotic pressure and pH, which are regulators of life movements. The results of correlation analysis also showed that the soil NH₄⁺ concentration was significantly positively correlated with the relative abundance of Firmicutes while significantly negatively correlated with the relative abundances of Proteobacteria, Bacteroidetes, and Acidobacteria (except for the Bacteroidetes at 60 d in DT soil) (Tables S3 and S4). Autotrophic ammonia-oxidizing bacteria, which belong to Proteobacteria, play an important role in nitrification [49]. Therefore, such a high NH₄⁺ concentration around the fertilizer point for a long time (>60 d) maybe not be conducive to the growth of ammonia-oxidizing bacteria.

The Shannon and PD diversity indices reflect community and phylogenetic diversity, respectively. Prior research showed that fertilization strongly influences soil bacterial diversity [50,51]. We found that the PD index was significantly negatively correlated with soil NH₄⁺ concentration (Table S5). The high NH₄⁺ concentrations in the FP and 0–4 treatments reduced the Shannon and PD indices for >60 d. Decreases in the diversity indices reflect small evolutionary differences between species and low functional diversity of bacterial communities [45]. The effect of NH₄⁺-N on microorganisms has two sides. The NH₄⁺-N can provide nutrients and energy for the growth of microorganisms. However, excessive NH₄⁺ has negative effects on the growth of microorganisms [40]. The high soil NH₄⁺ concentration alters the osmotic potential on both sides of the bacterial cell membrane, urging the water in the bacteria to flow to the side with high osmotic potential, leading to the water loss of the bacterial cytoplasm. The cell membrane plays a very important role in maintaining the dynamic balance of the cell. When bacteria are subjected to osmotic stress, the cell membrane produces a variety of responses to maintain cell balance and protect it from damage. The outer membrane protein regulator plays an important role in bacterial response to osmotic stress [33]. In this work, the FP and 0–4 treatments reduced the putative abundance of genes that regulate osmotic stress response and tricarboxylic acid transport (Tables S1 and S2). This might be because the bacteria resisted the rapid increase of osmotic stress by reducing the synthesis of transmembrane proteins. For this reason, numerous soil bacterial species cannot survive at high NH₄⁺ concentrations, and those that actually can tolerate this condition will predominate in the soil [25]. In addition, soil pH is a key factor affecting the survival of microorganisms. The sharp increase of soil pH caused by high NH₄⁺ concentration will change the enzyme activities of soil microorganisms.

Soil bacterial community structure and function may adapt to changes in the soil environment [52,53]. Here, NMDS plots (Figure 6) indicated that the soil bacterial community structure significantly and rapidly changed along the diffusion direction of hole-applied urea. The Mantel test confirmed that soil NH₄⁺ concentration and pH were significantly correlated with soil bacterial community throughout the entire trial (Table 5). Previous studies showed that soil pH is the major factor affecting the soil bacterial community [50,54]. In the present study, however, the distribution of the soil NH₄⁺ concentration determined the variation in soil pH. Overall, the NH₄⁺ gradient caused by hole application of urea modulated soil bacterial communities through direct or indirect effects. At 8 d after fertilization, the soil NH₄⁺ concentrations of the 12–16 treatments were 207 and 115 kg mg⁻¹ for the QD and DT soils, respectively. These values are close to the soil NH₄⁺ concentration during the early stage of traditional fertilization. However, the bacterial community structure of the 12–16 treatment was close to that of the control, which indicated that only extremely high NH₄⁺ concentration due to hole application of urea could significantly change the soil bacterial community structure in a short time.

4.3. Relationships among the Changes in the Soil Bacterial Community and Nitrification

The genes amoA, amoB, and amoC regulate the oxidation of ammonium to nitrate. An earlier study showed that appropriate increases in the nitrogen input level will increase the relative abundances of amoA, amoB, and amoC [28,55]. The gene hao converts hydroxylamine to nitrite during nitrification [29]. Here, the FP and 0–4 treatments lowered the relative abundances of the nitrification genes, whereas the 8–12 and 12–16 treatments had no such effect. By contrast, the 8–12 and 12–16 treatments increased the relative abundances of the nitrification genes with increasing incubation time. We also found that the 4–8 treatment reduced the abundances of the nitrification genes at <20 d. With increasing incubation time, however, the 4–8 treatment had the opposite effect. Therefore, high soil NH₄⁺ concentrations reduce the abundances of the nitrification genes [25]. Appropriate soil NH₄⁺ concentrations may be conducive to the growth of bacterial soil nitrifiers [28]. We found that the abundances of the nitrification genes in the FP and 0–4 treatments generally increased with incubation time possibly because the soil NH₄⁺ concentration decreased with increasing incubation time, thereby reducing nitrifier inhibition.

NH₄⁺-N has poor soil mobility because the soil readily immobilizes it. In contrast, NO₃⁻-N has relatively good soil mobility because the soil cannot easily immobilize it [18]. If plant roots fail to absorb NO₃⁻-N, it may leach and contaminate the groundwater [6]. Inhibiting nitrification improves NUE and reduces N loss. We observed strong inhibition of nitrification in the FP and 0–4 treatments throughout the entire incubation period. The Mantel and correlation analyses revealed that the soil bacterial community and nitrification gene abundances were significantly positively correlated with soil NO₃⁻ concentration at days 20 and 60. Overall, high NH₄⁺ concentrations inhibit nitrification by altering the soil bacterial community and reducing the abundances of nitrification genes. We also found that the inhibition of nitrification was stronger under the 4–8, 8–12, and 12–16 treatments in sandy loam (QD) than it was under those in silty loam (DT) after 60 d fertilization. This might be related to the soil texture, crop-rotation system, and expression of nitrification genes. To our knowledge, QD had a soybean-rapeseed (dryland) rotation, while DT had a rice-rapeseed (paddy-upland) rotation. The dryland rotation and sandy loam texture resulted in the better aeration of QD, which was conducive to the activities of nitrifying microorganisms and nitrification. In this study, the putative abundances of ammonia-oxidizing microorganisms in the 4–8, 8–12, and 12–16 treatments for QD soil were higher than those for DT soil. In addition, owing to the stronger NH₄⁺ retention capacity of silty loam, compared with QD, DT soil can maintain a higher NH₄⁺ concentration for a longer time, thereby inhibiting the activities of nitrifying microorganisms. These results are consistent with our previous findings for oilseed rape cultivation, where the N loss rate was significantly higher in QD than DT soil [21].

Following the hole application of urea, the high NH₄⁺ concentrations in the fertilizer core region (FP and 0–4 treatments) will inhibit the growth of most bacterial species and plant roots. As NH₄⁺-N diffuses to the periphery of the fertilizer core region, high NH₄⁺ concentrations stimulate plant roots and soil bacterial growth. Thus, roots and soil microorganisms will grow around the fertilizer core region, plants effectively absorb soil nutrients, and N fertilizer loss is reduced. Liu et al. [1] reported that application of urea to the rhizosphere promotes root proliferation. Furthermore, the soil adsorbs the NH₄⁺-N near the fertilizer core region, the relative differences in NH₄⁺ concentration among treatments are reduced, NH₄⁺-N migration is impeded, and high NH₄⁺ concentrations are maintained in the fertilizer core region. Hence, the fertilizer core region slowly releases NH₄⁺ and inhibits nitrification for a long time. These factors can explain the high yield and efficiency of hole application of urea. Although the migration characteristics of hole application of urea were investigated by one direction of N migration, we understood the relationships among soil bacteria and soil chemical properties under hole-applied urea by this work. High soil NH₄⁺ concentrations ensuing from hole application of urea decreased the putative abundance of nitrification genes, thereby inhibiting nitrification. In practical crop production, the distribution characteristics of NH₄⁺-N and the timing and extent of nitrification inhibition in the soil may be controlled by adjusting the fertilizer dosage and form. Moreover, owing to the inhibition of nitrification, the risk of soil acidification is also reduced. In this study, the disturbance of hole application of urea to soil microorganisms occurred mainly in the fertilizer core region. The ecological effect of hole application of urea deserves further study. In addition, four grams of urea were hole-applied in this trial, which was just an example to achieve the purpose of this study. In the agricultural practice, the nutrient demand of a plant (i.e., fertilizer dosage per hole) was generally determined according to the yield, nutrient uptake, and plant density.

5. Conclusions

The incubation trial performed herein elucidated the mechanisms by which hole application of urea functions as a slow-release N fertilizer and a nitrification inhibitor. High NH₄⁺ concentrations altered the soil bacterial community structure, lowered the bacterial community diversity index, and decreased the relative abundances of nitrification genes. The NH₄⁺ concentration gradient created by urea diffusion strongly inhibited nitrification in the fertilizer core region for >60 d. However, this trial was conducted without crops, and the absorption of N fertilizer by crop roots may affect the soil NH₄⁺ concentration and duration of nitrification inhibition. There is a need to further investigate the relationship between the root extension of different crops and fertilizer diffusion. In addition, only urea was considered in the present study. The combined effects of hole application of nitrogen, phosphorus, potassium, and other nutrients on their soil migration characteristics and effect on soil microorganisms merit further investigation. Overall, the present work provides theoretical and technical support for long-acting hole application of urea. The hole application of urea combines the merits of slow-release fertilizer and nitrification inhibitor; thus, it has broad application prospects.