The Effects of Localized Plant–Soil–Microbe Interactions on Soil Nitrogen Cycle in Maize Rhizosphere Soil under Long-Term Fertilizers

: Long-term fertilization can result in changes in the nitrogen (N) cycle in maize rhizosphere soil. However, there have been few reports on the impacts of plant–soil–microbe regulatory mechanisms on the N cycle in soil. In this study, soil samples were collected from a long-term experimental site located at Jilin Agricultural University, Changchun City, Jilin Province, Northeast China. We then analyzed the changes in the functional genes related to the N cycle, soil enzyme activity, and maize root exudates under long-term fertilizer application using metagenomics and liquid chromatography analysis. We aimed to investigate the response of the N cycle to long-term fertilizers, the interaction among plant, soil, and microbes, and the effect of the plant–soil–microbe system on the N cycle. Long-term fertilization had a signiﬁcant effect on soil N contents, N 2 O emissions, and enzyme activity related to the N cycle in maize rhizosphere soil. The functional genes of the N cycle were mainly enriched in the N degradation pathway in maize rhizosphere soil. N fertilizer application decreased the abundance of functional genes related to N ﬁxation and degradation, denitriﬁcation, and assimilatory nitrate reduction (ANRA) and increased the abundance of functional genes participating in dissimilatory nitrate reduction (DNRA) and anaerobic ammonia oxidation/hydroxylamine oxidation. The soil environment was positively related to soil enzyme activity and negatively related to the microbial community composition and amino acids in root exudates. The contribution rate of microorganisms to the N cycle was the highest (r 2 = 0.900), followed by amino acids (r 2 = 0.836) and the soil environment (r 2 = 0.832). Therefore, we concluded that N fertilizer is the main factor limiting the soil N cycle and that microorganisms are the main factor regulating the N cycle in the plant–soil–microorganism system.


Introduction
The nitrogen (N) cycle is the most basic ecological process in agricultural ecosystems. Soil N balance is very important for maintaining soil quality and crop growth and is directly related to the stability and productivity of agricultural ecosystems [1,2]. N is an important component controlling the stability and function of global ecosystems, and an increase in its content plays an important role in the stability of the soil nitrogen pool (reducing nitrogen leaching or gaseous loss) [3]. Fertilizers play an irreplaceable role in ensuring food security in China and have a significant impact on changes in soil nitrogen content and forms [4,5]. Note: the N fertilizer is urea (46%), the P fertilizer is heavy superphosphate (45%), and the K fertilizer is potassium sulfate (52%).

Method for Determination of Soil Physical and Chemical Properties, Enzyme Activity, and Root Exudates
Soil pH was measured by a pH meter. Total N (TN) was measured by the Kjeldahl method, and total phosphorus (TP) was measured by alkali fusion molybdenum antimony anti-spectrophotometry. Total potassium (TK) was measured by the NaOH melting-fire photometer method. Ammonium-N (NH 4 + −N) and nitrate-N (NO 3 − −N) were measured by a continuous flow analyzer. Available phosphorus (AP) and available potassium (AK) were measured by the 0.5 mol/L NaHCO 3 method and NH 4 CO 3 method, respectively. Readily oxidizable organic carbon (ROC) was determined by the potassium permanganate oxidation method, and soil organic carbon (SOC) and dissolved organic carbon (DOC) were determined by the TOC analyzer [37]. Determination of microbial biomass carbon (MBC) was performed by chloroform fumigation extraction [38]. The measured soil properties are listed in Appendix A, Table A1. Extraction was carried out according to the instructions of the soil enzyme activity kit, and soil enzyme activity, including N fixation enzymes, protease enzymes, nitrification enzymes, denitrification enzymes, and ureases enzymes, was determined by a microplate reader. In determining root exudates, the retrieved plants were washed away from the soil on the root surface and cultured in 1000 mL of distilled water for 48 h at 25 • C. A rotary evaporator (100 mL to 4 mL) was concentrated on the root exudates, the amino acid and organic acid contents were determined by a liquid-phase analyzer, and the soluble sugar content was extracted by a reagent kit and determined by a microplate analyzer [39].

Metagenomics Sequencing, Contig Assembly, and Annotation
Microbial DNA was extracted according to the manufacturer's instructions, and metagenomic sequencing was conducted at Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China), following the detailed steps listed in previous research [40]. The nonredundant contig set was obtained by using MMseqs 2 software. The (merged) contig sequence set was deredundant according to the similarity degree of 95% and the coverage degree of alignment area of 90%. Then, the contig sequence set was further filtered to remove nontarget fragment sequences and sequences with an actual depth of 0. Subsequently, contigs were used for the prediction of open reading frames (ORFs) using MetaGene-Mark (http://exon.gatech.edu/GeneMark/ accessed on 1 March 2021) and then translated into amino acid sequences. The nonredundant gene catalog was searched against the Kyoto Encyclopedia of Genes and Genomes (http://www.genome.jp/kegg/ accessed on 15 March 2023) for functional annotation.

Data Analysis
We used analysis of variance (ANOVA) to analyze the differences among means in soil physical-chemical properties, followed by the least significant difference (LSD) test if the difference was significant, and Spearman's rank correlation coefficients to examine the associations between the soil enzyme activity and soil properties and bacteria. The influencing factors were divided into four groups, namely, plant variables (root exudates), soil environmental variables (soil physical and chemical properties), microbial variables (microbial diversity, microbial composition ( Figure A1)), and soil enzyme activity. The key factors of each influencing factor were obtained by principal component analysis, and then the significant relationship among plants, soil, and microorganisms was analyzed by the key factors using Spearman's analysis and output by a pie network [41]. To analyze the influencing factors of the soil N cycle, redundancy analysis was used to select the factors that contributed the most to the functional genes of the N cycle in each group of variables, and then the contribution rate of each factor to the soil N cycle was calculated [42].

Effects of Long-Term Fertilization on Soil Nitrogen Content and N 2 O Emissions
Long-term fertilization significantly increased the TN content in the soil, and the TN content in the straw treatment was significantly higher than that in the chemical fertilizer treatment. The TN content in the NPKS treatment was the highest, which was 0.34 g/kg higher than that in the CK treatment. The NH 4 + −N content was also the highest under the NPKS treatment. The addition of straw increased the soil alkali-hydrolyzable nitrogen (ALN) content, and the NO 3 − −N content in the treatment with the chemical fertilizer was significantly higher than that in the treatment with straw addition. Fertilization also affected the percentage of nitrogen content to TN in the soil, and the ALN content was lower than that of CK except in the NPKS treatment, while fertilizer application increased the proportion of NH 4 + −N and NO 3 − −N to TN. During the whole growth period of maize, the N 2 O emissions were the highest at the large bell-mouth stage (long-term fertilization significantly affected the N 2 O emissions at the large bell-mouth stage). The N 2 O emissions of the N and NPK treatments were significantly higher than those of the other treatments, which were 236.93 and 230.45 mg·m −2 h −1 higher than those of CK, respectively ( Figure 1). significantly affected the N2O emissions at the large bell-mouth stage). The N2O emissions of the N and NPK treatments were significantly higher than those of the other treatments, which were 236.93 and 230.45 mg·m −2 h −1 higher than those of CK, respectively ( Figure 1). The activities of N-related enzymes in the soil were analyzed. The results showed that the highest activity of protease was 0.30~1.78 U/g in the rhizosphere soil of maize, and its activity increased significantly under long-term fertilization. The highest activity of protease was under the NPKS treatment. Urease had the second-highest activity (0.63~1.18 U/g). Compared with the CK treatment, the treatment with straw significantly The activities of N-related enzymes in the soil were analyzed. The results showed that the highest activity of protease was 0.30~1.78 U/g in the rhizosphere soil of maize, and its activity increased significantly under long-term fertilization. The highest activity of protease was under the NPKS treatment. Urease had the second-highest activity (0.63~1.18 U/g). Compared with the CK treatment, the treatment with straw significantly increased urease activity, while the treatment with N fertilizer significantly decreased urease activity. The activities of nitrification (20.89~43.11 IU/mg) and denitrification enzymes (26.43~37.17 IU/mg) were relatively low. Long-term fertilization significantly increased the activities of nitrification enzymes and decreased the activities of denitrification enzymes ( Table 2). Twenty-seven functional genes related to the N cycle in maize rhizosphere soil were analyzed. The results showed that these genes were involved in nitrogen fixation, nitrification, denitrification, the assimilative nitrate reduction pathway (ANRA), the dissimilatory nitrate reduction pathway (DNRA), anaerobic ammonia oxidation/hydroxylamine oxidation, and the N degradation pathway. The functional genes related to the N cycle were mainly concentrated in the N degradation pathway, and the relative abundance of genes in this pathway accounted for 27.36~32.20% of the N cycle. The N fixation pathway, denitrification pathway, and ANRA accounted for 23.95-25.99%, 17.79-23.85%, and 15.38-17.16% of the N cycle, respectively (Table 3). Some functional genes involved in the N cycle changed significantly under long-term fertilization. Compared with that of CK, the abundance of the nifA gene decreased by 1.26, 1.52, and 0.21% under the N, NPK, and NPKS treatments, respectively, while the abundance of the nifH gene increased by 0.08, 0.08, and 0.02% under the N, NPK, and NPKS treatments, respectively. Long-term fertilization also had a significant effect on N decomposition (p < 0.001). Compared with that of the CK treatment, the abundance of functional genes related to N degradation decreased significantly under N fertilizer application but increased significantly after straw addition. The abundance of the ureC and GDH2 genes was the highest under the S treatment, at 9.21% and 20.97%, respectively. The abundance of the ureC gene was the lowest under the NPK treatment, which was 1.83% lower than that under the CK treatment. The abundance of the GDH2 gene was the lowest under the N treatment, which was 3.09% lower than that under the CK treatment. The nitrification functional genes amoA, amoB, and amoC decreased significantly under long-term fertilization, especially under the N and NPK treatments. The response of denitrification functional genes to fertilization was opposite to that of nitrification functional genes. Denitrification functional genes had the lowest abundance in the CK treatments, while the N and NPK treatments significantly increased the abundance of the narG, narH, nirK, narB, and norZ genes. The abundance of the functional gene nrfA related to DNRA in the N and NPK treatments was significantly higher than that in the other treatments, while the abundance of the functional gene nasD related to ANRA in the N and NPK treatments was significantly lower than that in the other treatments. In addition, the N and NPK treatments also significantly increased the abundance of the functional genes hao/hzo related to anaerobic ammonia oxidation/hydroxylamine oxidation ( Figure 2). was the highest under the S treatment, at 9.21% and 20.97%, respectively. The abundance of the ureC gene was the lowest under the NPK treatment, which was 1.83% lower than that under the CK treatment. The abundance of the GDH2 gene was the lowest under the N treatment, which was 3.09% lower than that under the CK treatment. The nitrification functional genes amoA, amoB, and amoC decreased significantly under long-term fertilization, especially under the N and NPK treatments. The response of denitrification functional genes to fertilization was opposite to that of nitrification functional genes. Denitrification functional genes had the lowest abundance in the CK treatments, while the N and NPK treatments significantly increased the abundance of the narG, narH, nirK, narB, and norZ genes. The abundance of the functional gene nrfA related to DNRA in the N and NPK treatments was significantly higher than that in the other treatments, while the abundance of the functional gene nasD related to ANRA in the N and NPK treatments was significantly lower than that in the other treatments. In addition, the N and NPK treatments also significantly increased the abundance of the functional genes hao/hzo related to anaerobic ammonia oxidation/hydroxylamine oxidation ( Figure 2).

Effect of Long-Term Fertilization on Root Exudates of Maize
Thirteen kinds of amino acids were found in the root exudates of maize, among which the tyrosine content was the highest, accounting for 50.41-91.10% of the total amino acids, and its content was the highest under the S treatment, which was 11.42 µg/g. Long-term fertilization significantly reduced the histidine content, especially in the S and PKS treatments, which was lower than 0.31 µg/g. The methionine content in the N, NPK, and NPKS treatments was significantly higher than that in the S and PKS treatments. In addition, the contents of serine, threonine, and proline decreased significantly under long-term fertilization (p < 0.05) (Figure 3). mine oxidation (hao/hzo). The lower case letters 'a', 'b', and 'c' indicate a significant difference (p < 0.05) among different samples for each treatment (n = 18) using the ANOVA method.

Effect of Long-Term Fertilization on Root Exudates of Maize
Thirteen kinds of amino acids were found in the root exudates of maize, among which the tyrosine content was the highest, accounting for 50.41-91.10% of the total amino acids, and its content was the highest under the S treatment, which was 11.42 µg/g. Longterm fertilization significantly reduced the histidine content, especially in the S and PKS treatments, which was lower than 0.31 µg/g. The methionine content in the N, NPK, and NPKS treatments was significantly higher than that in the S and PKS treatments. In addition, the contents of serine, threonine, and proline decreased significantly under long-term fertilization (p < 0.05) (Figure 3). Nine organic acids were found in the root exudates of maize in this experiment, among which citric acid was the highest, accounting for 79.77-92.32% of the total organic acids. Long-term fertilization increased the citric acid content, especially in the treatment containing N fertilizer. The content of succinic acid increased significantly in the NPKS treatment. Long-term fertilization increased the content of formic acid and acetic acid, while the content of lactic acid decreased significantly under the N and NPK treatments and increased significantly under the S, PKS, and NPKS treatments (p < 0.05) (Figure 3).
The sugar content in the root exudates was sucrose > fructose > glucose, and the content of the three sugars changed significantly under long-term fertilization (p < 0.001). N, Nine organic acids were found in the root exudates of maize in this experiment, among which citric acid was the highest, accounting for 79.77-92.32% of the total organic acids. Long-term fertilization increased the citric acid content, especially in the treatment containing N fertilizer. The content of succinic acid increased significantly in the NPKS treatment. Long-term fertilization increased the content of formic acid and acetic acid, while the content of lactic acid decreased significantly under the N and NPK treatments and increased significantly under the S, PKS, and NPKS treatments (p < 0.05) (Figure 3).
The sugar content in the root exudates was sucrose > fructose > glucose, and the content of the three sugars changed significantly under long-term fertilization (p < 0.001). N, NPK, and NPKS significantly increased the sucrose content; compared with that of CK, the sucrose content increased by 0.97, 0.78, and 0.78 mg/kg, respectively. The fructose content also increased significantly under the N and NPK treatments but decreased significantly under the S, PKS, and NPKS treatments. The glucose content increased significantly under long-term fertilization (Figure 3).

Effect of Plant-Soil-Microorganism Interaction on the N Cycle
The principal component analysis was used to determine the principal component factors in each variable. There are four principal component factors in the soil's physical and chemical properties: SOC, DOC, ROC, and TN. There are two main components in soil enzyme activity, namely, nitrogenases and proteases. The Simpson index was the main factor in the microbial community diversity, and Elusimicrobia, Nitrospirae, and Firmicutes were the main factors in the community composition. A principal component analysis of sugar, organic acid, and amino acid in plant variables was carried out, and the main factors of phenylalanine, threonine, histidine, proline, arginine, citric acid, oxalic acid, and sucrose were finally determined (Table A3).
From the relationship diagram of significant correlations among plants, soil, and microorganisms, it can be seen that there is an extremely complex correlation among plants, soil, and microorganisms. There was a negative interaction between the soil environmental variables and the microbial community composition, and there was a significant negative correlation between DOC, ROC, TN, and SOC and Elusimicrobia and Nitrospirae (p < 0.001). There was a positive interaction between the soil environmental variables and soil enzyme activities, and there was a significant positive correlation between NITs and proteases and soil DOC, ROC, TN, and SOC (p < 0.05). Phenylalanine, threonine, and histidine in maize root exudates were negatively correlated with the soil environmental variables, especially with TN (p < 0.05). In addition, there was a significant negative correlation between DOC and threonine, histidine, and citric acid. Microbial diversity was significantly negatively correlated with the levels of sucrose and oxalic acid in plant root exudates, while it was highly significantly positively correlated with the levels of citric acid. There was a highly significant positive correlation between Elusimicrobia and Nitrospirae and histidine. There was a significant negative correlation between threonine and NITs and proteases ( Figure 4, Table A4). under long-term fertilization (Figure 3).

Effect of Plant-Soil-Microorganism Interaction on the N Cycle
The principal component analysis was used to determine the principal component factors in each variable. There are four principal component factors in the soil's physical and chemical properties: SOC, DOC, ROC, and TN. There are two main components in soil enzyme activity, namely, nitrogenases and proteases. The Simpson index was the main factor in the microbial community diversity, and Elusimicrobia, Nitrospirae, and Firmicutes were the main factors in the community composition. A principal component analysis of sugar, organic acid, and amino acid in plant variables was carried out, and the main factors of phenylalanine, threonine, histidine, proline, arginine, citric acid, oxalic acid, and sucrose were finally determined (Table A3).
From the relationship diagram of significant correlations among plants, soil, and microorganisms, it can be seen that there is an extremely complex correlation among plants, soil, and microorganisms. There was a negative interaction between the soil environmental variables and the microbial community composition, and there was a significant negative correlation between DOC, ROC, TN, and SOC and Elusimicrobia and Nitrospirae (p < 0.001). There was a positive interaction between the soil environmental variables and soil enzyme activities, and there was a significant positive correlation between NITs and proteases and soil DOC, ROC, TN, and SOC (p < 0.05). Phenylalanine, threonine, and histidine in maize root exudates were negatively correlated with the soil environmental variables, especially with TN (p < 0.05). In addition, there was a significant negative correlation between DOC and threonine, histidine, and citric acid. Microbial diversity was significantly negatively correlated with the levels of sucrose and oxalic acid in plant root exudates, while it was highly significantly positively correlated with the levels of citric acid. There was a highly significant positive correlation between Elusimicrobia and Nitrospirae and histidine. There was a significant negative correlation between threonine and NITs and proteases ( Figure 4, Table A4). The soil's physical and chemical properties, soil enzyme activity, root exudates, microbial diversity, and community composition have significant effects on the soil N cycle. The results showed that SOC in soil components; the Shannon index and Actinobacteria The soil's physical and chemical properties, soil enzyme activity, root exudates, microbial diversity, and community composition have significant effects on the soil N cycle. The results showed that SOC in soil components; the Shannon index and Actinobacteria in microbial components; and urine enzymes, methionine, lactic acid, and sucrose in root exudates had the highest influence on the soil N cycle functional genes (Table A5). The soil environmental variables, microbial variables, and root exudates also had significant effects on soil N cycling (p = 0.001). Among them, the contribution rate of microorganisms was the highest, the r 2 of community composition and the N cycle was 0.900, and the r 2 of diversity and N-cycle genes was 0.841. The contribution rate of soil environmental variables to the N cycle was 0.832. The lowest contribution rate of soil enzyme activity was 0.640. The contribution rate of root exudates to the N cycle was amino acid > organic acid > sugar ( Figure 5, Table A6). soil environmental variables, microbial variables, and root exudates also had significant effects on soil N cycling (p = 0.001). Among them, the contribution rate of microorganisms was the highest, the r 2 of community composition and the N cycle was 0.900, and the r 2 of diversity and N-cycle genes was 0.841. The contribution rate of soil environmental variables to the N cycle was 0.832. The lowest contribution rate of soil enzyme activity was 0.640. The contribution rate of root exudates to the N cycle was amino acid > organic acid > sugar ( Figure 5, Table A6).

Effects of Long-Term Fertilization on the Soil N Cycle
Fertilization plays a very important role in the soil N cycle [43]. The addition of fertilizers can increase the TN content within the soil, and the addition of straw will promote this effect. This is due to the application of chemical fertilizers, which increase the source of N in the soil. Additionally, the decomposition of straw releases the N contained in the straw into the soil, thus improving its fertility [44]. In this experiment, the N and NPK treatments significantly increased the NO3 − −N content and promoted the NH4 + −N and NO3 − −N percentages to TN, but fertilization reduced the ALN/TN percentage ( Figure 1, Table A2). This result is the same as that of Sun Ruijuan et al. [45]. N2O emissions reached their highest value (Figure 1) at the big bell-mouth stage of maize, which may be due to the enhanced respiration ability of roots to grow in the big bell-mouth stage of maize. Furthermore, the soil temperature at this stage was suitable for the growth and development of microorganisms, which promoted their respiration, thus making N2O emissions significantly higher than those at other growth stages [46,47].
Soil enzyme activity plays an important role in soil nutrient cycling and is a good index to measure the decomposition of soil organic matter [48,49]. Soil enzyme activity is very sensitive to environmental changes. Past research has noted that fertilization has a significant impact on soil enzyme activity, and the application of organic fertilizer can improve soil enzyme activity [50]. This study also reached a similar conclusion. The addition of straw promoted the activities of protease and urease. The activity of protease was highest in maize rhizosphere soil, which may be related to the increase in soil carbon and

Effects of Long-Term Fertilization on the Soil N Cycle
Fertilization plays a very important role in the soil N cycle [43]. The addition of fertilizers can increase the TN content within the soil, and the addition of straw will promote this effect. This is due to the application of chemical fertilizers, which increase the source of N in the soil. Additionally, the decomposition of straw releases the N contained in the straw into the soil, thus improving its fertility [44]. In this experiment, the N and NPK treatments significantly increased the NO 3 − −N content and promoted the NH 4 + −N and NO 3 − −N percentages to TN, but fertilization reduced the ALN/TN percentage ( Figure 1, Table A2). This result is the same as that of Sun Ruijuan et al. [45]. N 2 O emissions reached their highest value (Figure 1) at the big bell-mouth stage of maize, which may be due to the enhanced respiration ability of roots to grow in the big bell-mouth stage of maize. Furthermore, the soil temperature at this stage was suitable for the growth and development of microorganisms, which promoted their respiration, thus making N 2 O emissions significantly higher than those at other growth stages [46,47].
Soil enzyme activity plays an important role in soil nutrient cycling and is a good index to measure the decomposition of soil organic matter [48,49]. Soil enzyme activity is very sensitive to environmental changes. Past research has noted that fertilization has a significant impact on soil enzyme activity, and the application of organic fertilizer can improve soil enzyme activity [50]. This study also reached a similar conclusion. The addition of straw promoted the activities of protease and urease. The activity of protease was highest in maize rhizosphere soil, which may be related to the increase in soil carbon and nitrogen content by long-term fertilization. Through correlation analysis, it was found that the activity of protease was significantly correlated with the soil TN, SOC, DOC, and ROC (Table 2, Figure A2). The addition of straw promoted an increase in soil nutrient content and nitrogenase activity, especially in the treatment with straw addition. In addition, the long-term application of chemical fertilizer decreased the activity of urease, which may be related to the change in soil microbial community diversity [51,52]. The long-term application of chemical fertilizer in this experiment decreased the relative abundance of Actinobacteria and Candidatus Tectomicrobia, which were significantly positively correlated with soil urease. Overall, long-term chemical fertilizer application also increased the relative abundance of Verrucomicrobia, Gemmatimonadetes, Armatimonadetes, and Elusimicrobia, which were significantly negatively correlated with soil urease activity, thus reducing urease activity (Figures A1 and A2). The nitrifying enzyme activity increased significantly and the denitrifying enzyme activity decreased significantly under long-term fertilization, which may be related to the change in soil nutrient content [53]. SOC, DOC, and TN were positively correlated with the nitrifying enzyme activity and negatively correlated with the denitrifying enzyme activity ( Figure A2).
The N cycle is an important biogeochemical cycle that is almost completely completed by the redox reactions of microorganisms [54,55]. There are many factors affecting soil N cycling, among which fertilization plays a very important role [43]. Because black soil has high fertility and can transform inherent soil nitrogen into NH 4 + −N [56], in this study, the proportion of the N degradation pathway in the N cycle was the highest, at 27.36~32.20%. In addition, the ureC and GDH2 genes decreased significantly under N fertilizer application but increased significantly under straw application. This may be due to the decrease in soil pH caused by the long-term application of chemical fertilizer, which leads to a decrease in microbial diversity and a decrease in the ability of microbial decomposition of the soil's intrinsic N [19]. Moreover, the straw return increased the nutrients needed for microbial activity, increased microbial diversity, and then promoted the decomposition of the soil's inherent N [57]. Five nitrogen-fixing genes (nifA, nifS, nifD, nifH, and nifK) were identified in this study. Among them, the relative abundance of nifS was the highest, accounting for 18.87-20.21% of the N cycle, but long-term fertilization had no significant effect on its abundance. The gene abundance of nifA and nifH is strongly affected by N fertilizers. Some studies have shown that soil organic matter is the key factor driving nitrogen fixation in nifH [58]. In this study, the SOC content in the N and NPK treatments was lower than that in the other treatments, which led to a decrease in nitrogen fixation gene abundance. In addition, the functional genes in different N transformation processes changed significantly under long-term fertilization, which indicated that fertilization changed the transformation among various forms of N. Numerous studies have shown that soil nitrification and denitrification are important processes in the biogeochemical cycle of farmland soil and have a strong response to nitrogen fertilizers [59][60][61]. In this experiment, the application of nitrogen fertilizer reduced soil microbial nitrification and ANRA and promoted denitrification and DNRA. amoA and amoB abundance decreased significantly in the N, NPK, and S treatments, but there was no significant difference between PKS and NPKS. amoC abundance was the highest in the CK treatment. There are seven functional genes involved in denitrification in maize rhizosphere soil, among which the abundance of narG, narH, narB, nirK, and norZ significantly increased under long-term N fertilizer application, which promoted the emission of NO, N 2 O, and N 2 . However, the addition of straw had no significant effect on it. Although the abundance of hao/hzo genes involved in anaerobic ammoxidation/hydroxylamine oxidation is significantly lower than that of other functional genes, it is actually a chemical autotrophic pathway without organic carbon, so it also has a crucial impact on soil nitrogen loss [62]. The abundance of hao/hzo decreased significantly under the N and NPK treatments but increased significantly under the S treatment. ANRA and DNRA can transform N from the highest oxidized state (NO 3 − −N) to the lowest reduced state (NH 4 + −N), obtain energy from organic carbon, and enhance soil-available N retention for crop uptake and utilization [63]. In this study, the N and NPK treatments decreased nasD abundance and increased nrfA abundance. The nasD gene was also affected by straw, while the S treatment increased its abundance. In addition, compared with the application of chemical fertilizer, straw return increased the abundance of N fixation and decomposition genes, promoted nitrification, and inhibited denitrification. The above results indicate that straw return to the field may reduce the loss of N in black soil, and the application of a N fertilizer will increase the emission of N 2 O.

Role of Root Exudates in the N Cycle of Maize
Tyrosine and phenylalanine aromatic exudates were detected in maize root exudates, and the content of tyrosine was significantly increased in the treatment without N fertilizer, which also indicated that straw return could improve the ability of plants to produce pigments, defense compounds, and lignin (cell wall components) [64]. In this study, the methionine content increased significantly under long-term fertilization, especially in the treatment containing N fertilizer, and N fertilizer also increased the methionine content. All the above results indicate that N fertilizer was the main factor regulating amino acid secretion from maize roots. Citric acid had the highest content of organic acid in the root exudates of maize, which was significantly increased under long-term fertilization, especially in the treatment containing N fertilizer. A previous study has shown that there is a significant correlation between citric acid content and soil pH [65]. The increase in citric acid content promoted a decrease in soil pH (Table A1). In this study, the total amount of organic acids increased significantly under long-term fertilization, resulting in soil acidification. In addition, formic acid and acetic acid increased significantly under longterm fertilization, which was similar to the research results of Hu et al. [62]. The addition of straw promoted the lactic acid content secreted by corn and accelerated the degradation of corn straw. The content of sucrose in the root exudates of maize is the highest, which may be because sucrose is the main product of plant photosynthesis and carbohydrate transport is mainly in the form of sucrose in plants [66]. The contents of sucrose, fructose, and grape increased significantly after the long-term application of N fertilizer, which may be due to the increase in photosynthesis of plants and the increase in sugar content transmitted to the soil [67]. The addition of straw promoted glucose content and decreased fructose content, which indicated that the content of sucrose converted into glucose increased after the addition of straw.

Role of Plant-Soil-Microorganism Systems in N Cycle
Given that plants, soil, and microbes are tightly linked in the N cycle, long-term fertilizer application changes in the soil's physicochemical properties can be regarded as potential mechanisms driving the dynamics of plants, soil, and microbes and their interaction [41]. In this paper, through the correlation analysis of the main components in plants, soil, and microorganisms, similar conclusions are drawn. In maize root soil, there was a negative interaction between the soil environment and the microbial community composition, and there was a significant negative correlation between soil SOC, DOC, ROC, and TN and microbial Elusimicrobia and Nitrospirae. That is, under long-term fertilization, the contents of soil carbon and nitrogen nutrients increased significantly, which reduced the stability of the microbial community. Studies have shown that changes in the soil environment also change soil enzyme activity [68,69]. In this paper, soil enzyme activity was significantly related to the soil environment. Fertilization not only enhances soil nutrient content but also enhances the substrate needed by microbial activity, which promotes the enhancement of enzyme activity related to the N cycle. Previous studies have shown that root exudates can affect the microbial community and diversity [70,71]. In addition, Maurer et al. [72] confirmed that root secretion is significantly related to soil denitrification. It was found that amino acids and organic acids in maize root exudates negatively interacted with the soil environment, which showed that DOC, ROC, SOC, and TN in the soil were significantly negatively correlated with histidine under long-term fertilization conditions. In addition, TN was negatively correlated with phenylalanine, threonine, histidine, and citric acid.
The correlation between plants, soil, and microorganisms has a significant impact on the N cycle. The soil nitrogen cycle can be controlled by both biotic and abiotic factors. Studies have shown that soil nitrification and denitrification are affected by soil pH, soil water content, and environmental factors [73,74]. The changes in root exudates, enzymes, and microorganisms had significant effects on the soil N content [75]. The contribution rate of each component to the N cycle was obtained by redundancy analysis, and it was found that the contribution rate of microorganisms was the highest (composition r 2 = 0.900, diversity r 2 = 0.841), which may be because each reaction of the N cycle was driven by microorganisms; for example, nitrogen fixation was dominated by AOA and AOB, and nitrification and denitrification were dominated by nitrifying bacteria and denitrifying bacteria, so microorganisms were the main factor regulating the N cycle. Second, amino acids in root exudates and the soil environment contributed 0.836 and 0.832 to the N cycle, respectively. Amino acids and soil SOC are the main nutrients for soil microbial activity. Amino acids can be directly utilized by microorganisms through cell membranes, while SOC needs to be decomposed by microorganisms before being utilized. Therefore, the contribution of amino acids to the N cycle is slightly higher than that of soil. The lowest contribution rate of soil enzyme activity was 0.640. The main function of urease is to promote the decomposition of urea; it decomposes urea into ammonium carbonate and then dissociates into ammonia (NH 3 ) and carbonic acid with the participation of water [76]. Therefore, the effect on the N cycle is related to the amount of urea applied. In the CK and straw treatments, the amount of urea was low or no urea was applied, so the effect on the N cycle was relatively low.

Conclusions
In this paper, long-term fertilization increased the TN content in the rhizosphere soil of maize, especially in the NPKS treatment, compared with that of CK, which increased by 0.34 g/kg. Fertilizer application increased the percentage of NH 4 + −N and NO 3 − −N in TN and the emission of N 2 O. Long-term fertilization had a significant effect on the soil enzyme activities involved in the N cycle. Protease and nitrifying enzymes increased significantly under long-term fertilization, while denitrifying enzymes decreased under long-term fertilization. Urease activity increased significantly in the straw addition treatment but decreased significantly in the N fertilizer application treatment. The functional genes of the N cycle in maize rhizosphere soil were mainly concentrated in the N degradation pathway, followed by the nitrogen fixation, denitrification, and ANRA pathways. The abundances of nifA, ureC, and GDH2; amoA, amoB, and amoC; and nasD of ANRA were decreased by applying N fertilizer, and the abundances of nrfA (DNRA) and hao/hzo (anaerobic ammonia oxidation and hydroxylamine oxidation) were increased by applying N fertilizer, which indicated that N fertilizer was the main factor regulating the N cycle. In addition, the contents of amino acids, organic acids, tyrosine, citric acid, and sucrose in the root exudates of maize were the highest, and the contents of the root exudates also changed significantly under long-term fertilization. The soil environment has a positive interaction with soil enzyme activity and a negative interaction with the microbial community composition and amino acids in root exudates in maize rhizosphere soil. Soil N cycling was influenced by soil, microorganisms, root exudates, and enzyme activity. The contribution rate of microorganisms to N cycling was the highest (r 2 = 0.900), followed by amino acids (r 2 = 0.836) and the soil environment (r 2 = 0.832). The contribution rate of soil enzyme activity was the lowest (r 2 = 0.640). Overall, our results highlight the importance of long-term fertilization in the N cycle and the interaction among plant, soil, and microbes in the plant-soil-microbe system that regulates the N cycle in the maize rhizosphere. However, this paper analyzed only the effects of plant-soil-microorganism interactions on the whole N cycle, and the effects on specific reaction processes need further analysis.

Data Availability Statement:
The data presented in this study are available in the articles.    Figure A1. Relative abundance of bacterial community composition under long-term fertilization. Figure A1. Relative abundance of bacterial community composition under long-term fertilization. Figure A1. Relative abundance of bacterial community composition under long-term fertilization. Figure A2. Correlation heatmap between soil enzyme activity and soil properties and bacteria. '*' indicates a significant difference (p < 0.05) and '**' indicates a significant difference (p < 0.01) between two variables for each treatment (n = 18) using Spearman's rank correlation coefficients.