Effects of different maize residue managements on soil organic nitrogen cycling in different soil layers in northeast China

A field experiment was conducted in northeast China to examine the response of nitrogen cycling enzymes, that is, protease, N‐acetyl‐β‐D‐glucosaminidase (NAG), amidase, urease, and peptidase, as well soil organic nitrogen (SON) fractions and their relationships to RT (no maize residue application), NT (no tillage with maize residues placed on the surface), TT (plow maize residues into the soil at 0–35 cm depth in the first year, 0–20 cm in the second year, and 0–15 cm in the third year), and PT (plow maize residues into soil at 0–35 cm depth). The results have shown that NT significantly enhanced the activities of protease and NAG at 0–10 cm soil depth in comparison with other treatments. NT and TT significantly enhanced the activities of protease compared to RT and PT at 10–20 cm soil depth. TT significantly enhanced the activities of NAG in comparison with RT at 10–20 cm soil depth. TT and PT significantly enhanced the activities of NAG and peptidase compared to RT and NT at 20–35 cm soil depth. PT significantly increased the activities of protease in comparison with RT at 20–35 cm soil depth. NT, TT, and PT significantly enhanced the activities of peptidase compared to RT at 10–20 cm soil depth. NT significantly increased the concentration of hydrolyzable NH4+‐N$$ {\mathrm{NH}}_4^{+}\hbox{-} \mathrm{N} $$ in comparison with other treatments at 0–10 cm soil depth. PT significantly enhanced the concentration of hydrolyzable NH4+‐N$$ {\mathrm{NH}}_4^{+}\hbox{-} \mathrm{N} $$ and amino acid N compared to other treatments at 20–35 cm soil depth. Redundancy analysis showed that protease played a crucial role in the cycling of SON under RT and NT, whereas peptidase and NAG played a significant role in the cycling of SON under TT and PT, respectively. This study provided a comprehensive understanding of crop residue return methods for regulating soil N cycling.


| INTRODUCTION
Approximately 74 Tg of dry crop residues are produced in the world (Medina et al., 2015), which contain large amounts of organic matter and other nutrients (Li et al., 2018).Therefore, incorporating crop residues into field is an effective way to enhance soil health and fertility in agricultural ecosystems (Liu et al., 2021;Tian et al., 2020;Zhang et al., 2015).Nitrogen is a limiting factor that affects plant growth and agricultural production (Kandeler et al., 2011).Previous studies have found that return crop residues into field can improve soil N storage (Akhtar et al., 2018;Wu et al., 2021).Organic nitrogen comprises up to 90% of soil total nitrogen (Stevenson, 1982).According to Stevenson (1982), soil organic nitrogen (SON) can be segregated into total hydrolyzable N and acid-insoluble N. Total hydrolyzable N is composed of hydrolyzable NH + 4 -N, amino acid N, amino sugar N, and hydrolyzable unknown N. Based on their availability to plants and microorganisms, these organic nitrogen fractions can be categorized into active organic nitrogen fractions, that is, hydrolyzable NH + 4 -N, amino acid N, and amino sugar N, as well as stable organic nitrogen fractions, namely hydrolyzable unknown N and acid-insoluble N. Straw return can significantly affect organic nitrogen fractions.Several studies have reported that straw return or no tillage can enhance the content of active organic nitrogen fractions (Arshad et al., 1990;Guggenberger et al., 1999;Lu et al., 2018;Lü et al., 2013;Qiu et al., 2012;Wu et al., 2021;Xu et al., 2003), which can be attributed to an increase in microbial biomass, conversion from stable organic nitrogen fractions to active organic nitrogen fractions, and transformation of nitrogen from fertilizer to amino acids or amino sugars.However, some researches demonstrated that straw return decreases the concentration of active organic nitrogen fractions (Lü et al., 2013;Praveen et al., 2002), which might be due to the conversion of active organic nitrogen fractions into stable organic nitrogen fractions and crop uptake.For stable organic nitrogen fractions, straw return can increase the concentration of acid-insoluble N (Wu et al., 2021;Xu et al., 2003), because straw N is the major component in acid insoluble-N (Qiu et al., 2012); while another study demonstrated that straw return decreased the content of stable organic nitrogen fractions, because straw return favors the conversion from stable organic nitrogen fractions to active organic nitrogen fractions (Wu et al., 2021).Above all, the responses of different organic nitrogen fractions to straw return were different, which could be driven by differences in soil type, experiment duration, crop residue management practices, residue quality, and sampling depth (Arshad et al., 1990;Guggenberger et al., 1999;Lu et al., 2018;Lü et al., 2013;Praveen et al., 2002;Qiu et al., 2012;Wu et al., 2021;Xu et al., 2003).
Soil more organic nitrogen needs to be hydrolyzed into inorganic nitrogen, which can be taken up directly by plants and microorganisms (Geisseler et al., 2010).Enzymes involved in the nitrogen cycle play a vital role in the hydrolysis of organic nitrogen (Kandeler et al., 2011).Enzymes that participate in nitrogen cycling include protease, urease, amidase, N-acetyl-β-D-glucosaminidase (NAG), and peptidase (Kandeler et al., 2011;Landi et al., 2011).Straw return also significantly affects the activity of nitrogen cycle enzymes (Acosta-Martínez & Tabatabai, 2001;Akhtar et al., 2018;Deng & Tabatabai, 1996;Jin et al., 2009;Li, Wang, et al., 2022;Muruganandam et al., 2009;Tian et al., 2020;Wu et al., 2021;Zhao et al., 2016).It has long been known that straw return can increase the activities of nitrogen cycle enzymes (Acosta-Martínez & Tabatabai, 2001;Deng & Tabatabai, 1996;Muruganandam et al., 2009;Wu et al., 2021), because straw return can provide substrate for microbe consumption, which can produce more enzymes.However, a previous study found that subsoiling with mulch did not enhance urease activity in some seasons in comparison with conventional tillage, because the effect of different moisture and temperature under different seasons was higher than the effect of crop residue incorporation (Jin et al., 2009).Besides, Muruganandam et al. (2009) found that mold-board plow treatment decreased the activities of NAG compared with no-till.In addition, nitrogen cycling enzymes played a significant role in the hydrolysis of organic nitrogen under different crop residue managements: A 38 year field experiment found that decomposition of amino acids by leucine aminopeptidase may occur prior to decomposition of amino sugars by NAG to meet the N demand of microorganisms (Li et al., 2019), moreover, a 10 year study in northeast China found that protease and NAG played a significant role in SON cycling at 10-20 cm soil depth (Wu et al., 2021).However, these studies predominantly explored the relationship between nitrogen cycling enzymes and organic nitrogen under no tillage or plowing crop residue into soils of 0-20 cm in depth.Plowing crop residue into the soil at greater depths is also an effective way to enhance soil nitrogen storage (Liu et al., 2021;Tian et al., 2020), besides, plowing crop residue into different soil depths each year was a popular cultivation method used in northeast China.However, to date, the effects on SON and nitrogen cycling enzymes from annually plowing crop residue into soils at greater depths and plowing crop residue into soil at different depths each year remains unclear.At present, there is still relatively little data on the relationship between SON fractions and nitrogen cycling enzymes at different soil depths.Therefore, to deepen our understanding of soil nitrogen cycling under different crop residue managements in northeast China, it is important to explore the correlations between nitrogen fractions and nitrogen cycling enzymes at different soil depths.The aims of this study were to: (1) investigate the response of different organic nitrogen fractions and nitrogen cycling enzymes to different crop residue management practices and soil depths; and (2) explore the correlations between nitrogen cycling enzymes and different organic nitrogen fractions at different soil depths.

| Site description
The experiment was conducted in 2017 at a long-term conservation tillage experimental site in Lishu County, Jilin Province, northeastern China.The study site is located at 43°19′ N, 124°14′ E, and has a mean annual temperature of 6.9°C and mean annual precipitation of 614 mm.This area has a temperate, sub-humid, continental monsoon climate.The soil is classified as Phaeozem.In 2018, the soil was found to contain 28.3 g kg −1 soil organic matter (SOM) and pH (soil: water = 1: 2.5) of 7.8 at 0-20 cm in depth.The total C and N content of the maize residues was 448.2 and 7.1 g kg −1 , respectively.This area was planted with corn only, and no additional crops were planted following the autumn harvest.

| Experimental design and soil sampling
This trial used a randomized block design with four replicates.Each experimental plot measured 400 m 2 (40 m × 10 m).The treatments included: (1) RT: ridge tillage with no maize residue; (2) NT: no tillage with maize residues placed on the soil surface; (3) TT: 3 years as a cycle; in the first year, maize residues were plowed to 0-35 cm soil depth with a kind of plough (moldboard plow); in the second year, maize residues were plowed to 0-20 cm soil depth; in the third year, maize residues were plowed to 0-15 cm soil depth with a rotary tiller; and (4) PT: maize residues were plowed to 0-35 cm soil depth every year.Under TT and PT treatments, maize residue was crushed by a straw crusher before plowing into soil layer.Maize residue was applied at a rate of 7.5 t ha −1 year −1 each plot.All the plots were fertilized annually with 240 kg N ha −1 , 90 kg P 2 O 5 ha −1 , and 90 kg K 2 O ha −1 and received natural rainfall without artificial irrigation.In September 2020, five soil cores of 3 cm in diameter were randomly sampled from each plot using an auger at depths of 0-10 cm, 10-20 cm, and 20-35 cm from each plot.After removing visible root and plant residues, the cores were combined for each plot, homogenized, and passed through a 2-mm sieve.A portion of each soil sample was stored at 4°C for up to 2 weeks for analysis of nitrogen cycling activities and dehydrogenase.The remaining samples were air-dried for analyzing SON fraction and other soil properties within 3 months of sampling.

| SON fractions
Organic nitrogen fractions were determined by hydrolyzing the soil sample with 6 M HCl in an autoclave at 121°C for 6 h (Stevenson, 1982).Total hydrolyzable N was measured using steam distillation with 10 M NaOH after digestion with sulfuric acid and a K 2 SO 4 -catalyst mixture.Amino acid N was determined using steam distillation with a phosphate-borate buffer after treatment with NaOH at 100°C and ninhydrin powder.Hydrolyzable NH + 4 -N was measured via steam distillation using MgO.Amino sugar N was calculated as the difference between the amount of N evaporated by steam distillation with phosphate-borate at pH 11.2 and hydrolysable NH + 4 -N .Hydrolysable unknown-N was calculated as the difference between total hydrolyzable-N and N accounted for as hydrolyzable NH + 4 + amino acid + amino sugar − N. The amount of acid-insoluble N was calculated as the difference between soil TN and total hydrolyzable N.

| Nitrogen cycling enzyme activities
Protease activity was measured as described by Ladd and Butler (1972).Field moist soil samples were incubated with 0.05 M Tris buffer (pH 8.10) and 2% sodium caseinate (w/v) for 2 h at 50°C.The tyrosine produced was determined using the colorimetrical method at 700 nm, and the protease activity was expressed as μg tyrosine g −1 soil 2 h −1 .Amidase activity was assayed using the method described by Frankenberger and Tabatabai (1980).Field moist soil samples were incubated with 0.1 M sodium borate buffer (pH 8.50) and 0.5 M formamide for 2 h at 37°C.The ammonium released was determined using the colorimetrical method at 660 nm, and amidase activity was expressed as μg NH + 4 g −1 soil 2 h −1 .Soil NAG activity was measured using field moist soil incubated with 0.1 M acetate buffer (pH 5.50) and N-acetyl-β-D-glucosaminide solution at 37°C for 1 h.The nitrophenol produced was measured using the colorimetrical method at 405 nm, and the activity of NAG was expressed as mg ρnitrophenol kg −1 soil h −1 (Parham & Deng, 2000).Soil urease activity was assayed by determining the amount of urea remaining after incubation (Tabatabai, 1994).Field moist soil was incubated with 2 mg mL −1 of urea solution at 37°C for 5 h.The residual urea was determined using a colorimetric method at 527 nm.Soil urease activity was expressed as mg urea kg −1 soil 5 h −1 (Douglas & Bremner, 1970).Soil peptidase activity was measured using a fluorescent substrate (7-amino-4-methylcoumarin).A total of 2.00 g of field-moist soil was weighed and mixed with 200 mL NaN 3 to produce soil suspension, 50 μL acetate buffer (pH 5.0), 100 μL substrates, and 50 μL soil suspension was then added to a 96-well flat-black-bottomed microplate (NUNC, Denmark) before being incubated for 3 h at 30°C.In this study, soil peptidase activity was the sum of the activity of leucine aminopeptidase, glutamate aminopeptidase, and aspartate aminopeptidase.The substrates of these enzymes were L-leucine-AMC, L-glutamic acidγ-AMC, and aspartic acid-AMC, respectively.The soil peptidase activity was expressed as nmol AMC g −1 soil h −1 (Frossard et al., 2012;Marx et al., 2001).

| Statistical analysis
All values were expressed based on the weight of soil that had been oven-dried at 105°C.The soil data represent the mean values of four replicates with standard errors.One-way ANOVA with Duncan's test at p = 0.05 level was performed separately for the 0-10 cm, 10-20 cm, and 20-35 cm soil depths to analyze the responses of different organic nitrogen fractions and nitrogen cycling enzymes to different crop residue management.All the statistical analyses were conducted using SPSS software (version 21.0; SPSS, Chicago, IL, USA).Redundancy analysis (RDA) was performed using CANOCO software (Version 5.0, Microcomputer Power, Ithaca, NY) to explore the correlations between SON fractions and soil environmental factors.The Monte Carlo permutation test was used to test the statistical significance of variance factors.

| Soil nitrogen cycling enzyme activities
Compared to the other three treatments, NT significantly enhanced the activities of protease and NAG (Figure 1a,b, p < 0.01) at 0-10 cm soil depth, whereas TT significantly decreased the activity of peptidase (Figure 1c, p < 0.01) in comparison with RT and NT; PT significantly decreased the activities of NAG, peptidase and amidase in comparison with RT and NT (Figure 1b-d, p < 0.01).At soil of 10-20 cm in depth, NT significantly enhanced the activities of protease and peptidase compared to RT (Figure 1a,c, p < 0.01); TT significantly increased the activities of protease, peptidase and NAG compared to RT (Figure 1a-c, p < 0.01); PT significantly increased the activities of peptidase compared to RT (Figure 1c, p < 0.01); no significant differences were found among NT, TT and PT in NAG, protease and peptidase.At a soil depth of 20-35 cm, TT significantly enhanced the activities of NAG and peptidase (Figure 1b,c, p < 0.01) compared with RT and NT; PT significantly increased the activities of protease, peptidase, and NAG compared with RT and NT (Figure 1a-c, p < 0.01).

| SON fractions
At a soil depth of 0-10 cm, NT significantly enhanced the concentration and proportion of hydrolyzable NH + 4 -N (Figures 2a and 3A, p < 0.05) compared to the other three treatments, however, it did not show a significant difference with RT in other organic nitrogen fractions; TT and PT significantly decreased the concentration of hydrolyzable NH + 4 -N, amino acid N and amino sugar N (Figure 2a-c, p < 0.05) compared to RT and NT.At a soil depth of 10-20 cm, the concentration of hydrolyzable NH + 4 -N under the RT and NT treatments was significantly higher than those under the TT and PT treatments (Figure 2a, p < 0.05); the concentration and proportion of amino acid N under the NT, TT and PT treatments were significantly lower than those under the RT treatment (Figures 2b and 3B, p < 0.05); the concentration and proportion of amino sugar N under the NT and PT treatments were significantly lower than RT (Figures 2c and 3B, p < 0.05); the concentration and proportion of hydrolyzable unknown N under NT, TT and PT were significantly higher than under the RT treatment (Figures 2e and 3B, p < 0.05).At a soil depth of 20-35 cm, PT significantly increased the concentration and proportion of hydrolyzable NH + 4 -N and amino acid N in comparison with the other three treatments (Figures 2a,b and 3C, p < 0.05).
We divided the five SON fractions into two parts (Figures 4 and 5), that is, active SON including hydrolyzable NH + 4 -N, amino acid N, and amino sugar N and stable SON, comprising hydrolyzable unknown N and acid-insoluble N. At a soil depth of 0-10 cm, proportion of active SON under RT and NT was significantly higher than that under TT and PT (Figure 4A, p < 0.01), moreover, the concentration of active SON under NT was significantly higher than other three treatments (Figure 5A, p < 0.05).At a soil depth of 10-20 cm, the concentration and proportion of active SON under RT was significantly higher than that under the other three treatments (Figures 4B and 5B, p < 0.05).At a soil depth of 20-35 cm, the concentration and percentage of active SON under PT was significantly higher than that under the other three treatments (Figures 4C and 5C, p < 0.05).

| Relationships between SON fractions and soil nitrogen enzymes
The RDA results showed that the first two axes explained 83.93%, 80.88%, 57.67% and 79.15% of the total variation in SON fractions under RT, NT, TT, and PT, respectively (Figure 6).Protease explained significant proportions of the variations in SON fractions under RT and explained 73.3% of the overall variation in SON fractions (Table 1, p < 0.01).Protease was the most important enzyme that influenced SON fractions under NT and explained 66.0% of the overall variation in SON fractions ( Soil protease was significantly and positively correlated with amino acid N under RT and NT (Tables 2 and 3, p < 0.01).There was a significantly negative relationship between peptidase and amino acid N under TT (Table 4, p < 0.01), whereas there was a significantly negative relationship between NAG and amino sugar N under TT (Table 5, p < 0.01).

| Soil nitrogen cycling enzyme activities
Compared to other treatments, NT enhanced the activities of nitrogen cycling enzymes in soil at 0-10 cm in depth, TT increased the activities of nitrogen cycling enzymes in soil at 10-20 cm and 20-35 cm in depth, PT increased the activities of nitrogen cycling enzymes in soil at 20-35 cm in depth.It is generally accepted that nitrogen cycling enzymes primarily originate from microorganisms (Kandeler et al., 2011;Landi et al., 2011).Our study has shown that NT, TT, and PT significantly enhanced the concentrations of MBC and MBN at soil depths of 0-10 cm, 10-20 cm and 20-35 cm, respectively (Figures S1 and S2).This indicated that the incorporation of maize straw increased the microbial biomass, which increased the production of nitrogen cycling enzymes.Moreover, NT significantly enhanced the content of SOC at a soil depth of 0-10 cm; TT and PT significantly increased the content of SOC at a soil depth of 20-35 cm (Figure S3), which indicated that maize straw return could supply more substrate for microorganism consumption, thus enabling them to secrete more nitrogen cycling enzymes (Muruganandam et al., 2009;Wu et al., 2021;Zhao et al., 2016).However, NT did not the activities of peptidase, amidase and urease in soil at 0-10 cm in depth and TT did not increase the activities of amidase and urease in soil at 10-20 cm in depth compared to RT, because NT and TT significantly increased the concentration of NH + 4 at the 0-10 cm and 10-20 cm soil depths in comparison with RT, respectively (Figure S4), previous studies have demonstrated that high concentrations of NH + 4 can suppress the production of peptidase, amidase, and urease (Dodor & Tabatabai, 2003;McCarty & Bremner, 1992;Sinsabaugh & Follstad Shah, 2012).When comparing different soil depths, peptidase activity was different from that of other nitrogen cycling enzymes: In soil at depths of 10-20 cm and 20-35 cm, peptidase activity was significantly higher than in soil at 0-10 cm in depth under TT and PT.This is likely because the nitrogen availability was relatively low (the concentration of hydrolysable NH + 4 -N in 10-20 cm and 20-35 cm soil depth) at soil depths of 10-20 cm and 20-35 cm, which forced microorganism to produce extracellular enzyme (peptidase) to meet their nitrogen requirements at deeper soil depth (Norman et al., 2020).| SON fractions

| SON
Compared to RT, NT could only enhance the accumulation of hydrolyzable NH + 4 -N at the 0-10 cm soil depth, which was similar to the findings of Wu et al. (2021), because low molecular weight organic materials originating from the decomposition of maize residue could hinder the NH + 4 entering into the interlayers of clay minerals (Porter & Stewart, 1970); moreover, our study showed that NT significantly enhanced the concentration of MBN at the 0-10 cm soil depth (Figure S2), previous studies have demonstrated that the higher MBN caused by crop residue incorporation can compete for NH + 4 with clay minerals and reduce the fixation of NH + 4 (Qiu et al., 2012).However, NT did not enhance the accumulation of amino acid N over the entire soil profile, which is inconsistent with the results of Arshad et al. (1990), Qiu et al. (2012) and Wu et al. (2021).On one hand, the significantly lower soil pH under NT in comparison with RT in the 0-10 cm and 10-20 cm soil depths (Figure S5) enhanced the turnover rate of amino acids, because previous studies have confirmed that the a 0.5 unit of decline in soil pH enhances the turnover rate of amino acids four times (Jones & Kielland, 2002), which might offset the increase in input from maize residue incorporation at the 0-10 cm soil depth and decreased the concentration of amino acid N at the 10-20 cm soil depth; on the other hand, hydrolysis by nitrogen cycling enzymes may also have led to this result.In line with the findings for amino acid N, NT did not increase the concentration of amino sugar N in the entire soil profile, which is in contrast with the findings previous studies (Ding et al., 2010;Liang et al., 2007;Liu et al., 2019;Wu et al., 2021), because the significantly higher activities of NAG under NT may hydrolyze the amino sugar, moreover, the significantly higher hydrolyzable NH + 4 -N and NH + 4 -N under NT in comparison with RT at the 0-10 cm soil depth indicated that the decomposition of amino sugar N may partially transform into hydrolysable NH + 4 -N and NH + 4 -N.In contrast with NT, after a cycle of maize residue incorporation, TT did not have the potential to accumulate active SON in the entire soil profile in comparison with RT and NT.At the 0-10 cm soil depth, when maize residue was plowed into the soil profile, the soil surface was disturbed, which transferred the nutrients on the soil surface to greater soil depths (Hou et al., 2012;Tian et al., 2020).In soil at 10-20 cm in depth, the reason that TT did not have the potential to accumulate active SON could be partially attributed to the significantly lower pH compared to RT (Figure S5); the decline of soil pH under TT might be also due to the significantly higher NH + 4 compared with RT, because the hydrolysis of NH + 4 can produce H + ; besides, our results showed that the nitrogen content of maize residue under TT in soil at 10-20 cm in depth was significantly higher than that under RT (Figure S6), which indicated that more organic nitrogen may have been taken up by maize roots under TT.At the 20-35 cm soil depth, although maize straw was plowed into this soil depth in the first year (2018), this soil layer was not disturbed in the following 2 years, which did not carry more nutrients to greater soil depths; besides, the duration of the current study was too short to release more nutrients from the maize residue.to PT did not have the potential to enhance active SON storage in comparison with RT and NT at 10-20 cm soil depths, because previous studies have demonstrated that when available N becomes scarce, microorganisms can decompose their cell walls (amino sugars) to meet their nitrogen demand (Amelung et al., 2001;Li et al., 2019;Wu et al., 2021).PT increased the accumulation of hydrolyzable NH + 4 -N and amino acid N at 20-35 cm soil depth compared to the other three treatments, and the proportion of hydrolyzable NH + 4 -N and amino acid N in total N under PT was significantly higher than the other three treatments at 20-35 cm soil depth, and the concentration and proportion of active SON in total N under PT were significantly higher than those in the other three treatments (Figures 4 and 5), which indicates that the significantly higher hydrolyzable NH + 4 -N and amino acid N may have originated from the decrease in acid-insoluble N (Kandeler et al., 1999).Our study has shown that PT enhanced the concentration of MBN and DHA activity at the 20-35 cm soil depth compared to RT and NT (Figures S2 and S7), which also confirms this.Moreover, plowing maize residue to a depth of 20-35 cm can carry more nutrients and maize roots to this soil depth (Hou et al., 2012;Tian et al., 2020), and nutrient decomposed from the maize residue may also enhance the concentration of SOM (Figure S8), thus maintaining more active SON (amino acid N) (Li, Ye, et al., 2022).
F I G U R E 6 Redundancy analysis between soil organic N fractions and nitrogen cycling enzymes under RT, NT, TT, and PT (AAN, amino acid N; AIN, acid insoluble N; ASN, amino sugar N; HNN, hydrolysable NH + 4 -N; HUN, hydrolysable unknown N; pep, peptidase; a: RT, b: NT, c: TT, d: PT).NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year.

| Stable SON
hydrolyzable N is a stable organic nitrogen fraction, which is in the form of-amino-N, and its origin is largely unknown (Li et al., 2014), and its content has remained unclear owing to the complexity of its structure (Schulten & Schnitzer, 1997).In our study, at 10-20 cm soil depth, NT, TT and PT significantly increased the concentration and proportion of hydrolyzable unknown N compared to RT, which was in line with the results of Qiu et al. (2012) and Wu et al. (2021), moreover, NT, TT and PT significantly decreased the concentration and proportion of amino acid N compared to RT (Figures 2 and 3), NT and PT decreased the concentration and proportion of amino sugar N compared to RT (Figures 2 and 3), which was similar to the results of Praveen et al. (2002), because as discussed above, the decreased soil pH under NT, TT and PT might enhance the turnover rate of amino acid N (Jones & Kielland, 2002), and the decrease of amino acid N might convert to hydrolysable unknown N, which Abbreviations: NAG, N-acetyl-β-D-glucosaminidase; NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year.
T A B L E 1 Statistic results of redundancy analysis between soil organic N fractions and nitrogen cycling enzymes under different maize residue placements.T A B L E 2 Correlations between nitrogen cycling enzymes and soil organic nitrogen fractions under ridge tillage with no maize residue (n = 12).a component of stable SON; Similar to amino acid N, amino sugars might partially convert to hydrolyzable unknown N (Li et al., 2014).

NAG
Acid-insoluble N is partially regarded as a heterocyclic compound and may contain proteinaceous materials (Li et al., 2014;Piper & Posner, 1972;Wu et al., 2021).It can originate from insoluble soil residues, the condensation of amino acids and sugars, and N compounds that are associated with soil minerals or enter the mineral lattice (Anderson, 1961).In our study, maize residue return did not enhance the concentration of acid-insoluble N among three soil depths.However, the concentration of acid-insoluble N was significantly lower at the 10-20 cm soil depth under NT than under RT, which was inconsistent with previous studies that straw return increased the content of acid-insoluble N, because acid-insoluble N might be converted to hydrolyzable unknown N (Schulten & Schnitzer, 1997).It has been demonstrated that acid-insoluble N is not totally inert, and some of it can be converted to other N forms (Schulten & Schnitzer, 1997).Therefore, deep decomposition of maize residue to humic substances may increase acid-insoluble N content along with the experimental duration, which need further investigations in future.

| Relationships between SON fractions and nitrogen cycling enzymes
To explore the mechanisms of nitrogen cycling enzymes in SON turnover under different treatments, RDA was

T A B L E 4
Correlations between nitrogen cycling enzymes and soil organic nitrogen fractions under maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year (n = 12).T A B L E 5 Correlations between nitrogen cycling enzymes and soil organic nitrogen fractions under maize residues were plowed into 0-35 cm soil layer (n = 12).

NAG
used to analyze their correlations.cycling enzymes were as environmental variables and SON fractions as samples.The discussions are as follows: TT, based on the result of RDA and Pearson correlation analysis, we speculated that peptidase might play a significant role in the cycling of organic nitrogen (amino acid N), It is generally accepted that peptidase hydrolyze peptide bonds in proteins and produce amino acids (Norman et al., 2020).The concentration of amino acid N at the 0-10 cm soil depth was significantly higher than that of 20-35 cm soil depth, whereas the trend of peptidase activity was reversed, which indicated that low levels of end products (amino acid N) de-repressed enzyme production (peptidase) at the 20-35 cm soil depth, high levels of end products (amino acid N) repressed enzyme production (peptidase) at the 0-10 cm soil depth.Unlike TT, NAG might play a significant role in the cycling of organic nitrogen (amino sugar N) under PT.Our results showed that the concentration of amino sugar N at the 10-20 cm soil depth was significantly lower than that at the other two soil depths, whereas the trend of NAG was reversed.Previous studies have demonstrated that when available C or N is depleted, microorganisms tend to produce NAG to decompose amino sugar N to meet their demand (Amelung et al., 2001;Li et al., 2019), which leads to relatively lower amino sugar N at the 10-20 cm soil depth.Therefore, negative feedback may regulate the production of nitrogen cycling enzymes under the TT and PT conditions.
Unlike TT and PT, protease plays a significant role in the cycling of organic nitrogen under both RT and NT conditions.Similar to peptidase, protease can also hydrolyze the peptide bonds of proteins to produce amino acids (Landi et al., 2011).In contrast to TT and PT, there was a significant positive relationship between protease activity and amino acid N under RT and NT (Tables 2 and  3, p < 0.01).Therefore, negative feedback may not be an enzyme production mechanism regulating the cycling of SON, the absence of negative feedback may be caused by the lack of enough substrate for enzymes.At the 20-35 cm soil depth under RT and NT, although nutrient availability was significantly lower than that at the upper two soil depths, no more enzymes were produced at the 20-35 cm soil depth.Unlike TT and PT, RT and NT did not disturb the deeper soil layers, which could not transfer more nutrients into deeper soil layers, moreover, the absence of disturbance of the soil layer made the activity of microorganisms under RT and NT lower than that under TT and PT at the 20-35 cm soil depth.Previous studies have demonstrated that enzyme production is enhanced when there is a lack of simple and complex nutrients.However, low nutrient availability can also constrain enzyme production (Allison & Vitousek, 2005).Therefore, lower nutrient availability and microorganism activity under RT and NT may not produce more enzyme at 20-35 cm soil depth.The production of protease under RT and NT may depend on the different amounts of substrate at different soil depths.
Therefore, in our study, both the different distributions of substrate or end product and varied microbial activities among three soil layers under different treatments could lead to the different mechanisms that produce nitrogen cycling enzymes under different maize residue placements.

| CONCLUSIONS
Our results indicated that different maize straw placements can affect nitrogen cycling by affecting nitrogen cycling enzymes and SON fractions at the three soil depths after a 3 year long cycle of maize straw return.Among the three maize straw treatments, NT, TT and PT all had the potential to regulate the cycling of SON, but the capacity at different soil depths may be different.NT had the potential to enhance nitrogen cycling enzymes and supply active SON at the 0-10 cm soil depth.PT had the potential to enhance nitrogen cycling enzymes and supply active SON at the 20-35 cm soil depth, while having the potential to conserve stable SON at the 10-20 cm soil depths.TT had the potential to enhance nitrogen cycling enzymes at the 10-20 cm soil depth and had a potential to conserve stable SON at depths of 10-20 cm.Therefore, TT and PT are likely agronomic strategies for potentially maintaining the stable soil N pool in the shallow soil layer, whereas PT is likely an agronomic strategy for potentially increasing the supply of soil N in the deep soil layer in northeast China.RDA indicated that protease played a crucial role in the cycling of SON under RT and NT, whereas peptidase and NAG played a significant role in the cycling of SON under TT and PT, respectively.

F
Nitrogen enzyme activities under different maize straw placements.The error bars represent standard deviation.Bars followed by different uppercase letters indicate differences (p < 0.05) among different maize straw placements.Bars followed by different lowercase letters indicate differences (p < 0.05) among different soil depths.NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year, respectively (a: protease; b: NAG; c: peptidase; d: amidase; e: urease).

F
Soil organic nitrogen fractions under different maize straw placements.The error bars represent standard deviation.Bars followed by different uppercase letters indicate differences (p < 0.05) among different maize straw placements.Bars followed by different lowercase letters indicate differences (p < 0.05) among different soil depths.NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year, respectively (a: Hydrolysable NH + 4 -N; b: Amino acid N; c: Amino sugar N; d: Acid insoluble N; e: Hydrolyzable Unknown N).

F
Distribution (%) of soil organic N under different maize straw placements.The error bars represent standard deviation.Bars followed by different lowercase letters indicate differences (p < 0.05) among different maize straw placements.NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year, respectively (A: 0-10 cm; B: 10-20 cm; C: 20-35 cm; AAN, amino acid N; AIN, acid insoluble N; ASN, amino sugar N; HNN, hydrolysable NH + 4 -N; HUN, hydrolyzable unknown N).F I G U R E 4 Distribution (%) of soil active soil organic nitrogen (SON) and stable SON under different maize straw placements.The error bars represent standard deviation.Bars followed by different lowercase letters indicate differences (p < 0.05) among different maize straw placements.NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year, respectively (A: 0-10 cm; B: 10-20 cm; C: 20-35 cm).

F
Concentrations of soil active soil organic nitrogen (SON) and stable SON under different maize straw placements.The error bars represent standard deviation.Bars followed by different lowercase letters indicate differences (p < 0.05) among different maize straw placements.NT, no tillage with maize residues placed on the surface; PT, maize residues were plowed into 0-35 cm soil layer; RT, ridge tillage with no maize residue; TT, maize straw was plowed into 0-35 cm, 0-20 cm and 0-15 cm soil depth in the first, second and third year, respectively (A: 0-10 cm; B: 10-20 cm; C: 20-35 cm).

Table 1 ,
p < 0.01).Peptidase explained significant proportions of the variations in SON fractions under TT and explained 46.2% of the overall variation in SON fractions (Table 1, p < 0.01).NAG was the most important enzyme that influenced SON fractions under PT and explained 52.2% of the overall variations in SON fractions (Table 1, p < 0.01).
Bold numbers indicate statistically significant correlations between soil parameters at p < 0.05. Note: Correlations between nitrogen cycling enzymes and soil organic nitrogen fractions under no tillage with maize residues placed on the surface (n = 12).
Correlation is significant at the 0.05 level.*Correlation is significant at the 0.01 level.