A long-term field trial was established in 2009 at the Dahe Experimental Station, located in Shijiazhuang City, Hebei Province, North Central China (38°07′ N and 114°29′ E). This region is characterized by fluvoaquic soil and exhibits a typical temperate and subhumid continental monsoon climate, with an average annual temperature of 14.3°C and an annual precipitation of 400 mm. Another field trial was also initiated in 2009 in Liufangzi County, Gongzhuling City, Jilin Province, situated in northeastern China (43° 34.86 N and 124° 53.92 E). The study area in Liufangzi County features black soils known as haplic phaeozems (FAO classification) and mollisols (USDA classification). It experiences a temperate and semi-humid continental monsoon climate, with an average annual temperature ranging from 4 to 5°C and an annual precipitation between 500 and 900 mm.

The experimental design employed in both locations was a randomized complete block with four replicates. Each plot size in the fluvo-aquic soil site was 45 m² (5 × 9 m), while in the black soil site, it was 40 m² (10 × 4 m). In fluvo-aquic soil, three treatments were implemented, namely the control, moderate-N182 kg ha⁻¹, and high-N225 kg ha⁻¹. Similarly, in black soils, three treatments were implemented, including a control, moderate-N200 kg ha⁻¹, and high-N251 kg ha⁻¹. A total of 24 experimental plots were used, resulting from combining 3 treatments, 4 replications, and 2 sites. In both sites, the moderate-N treatments represented lower N fertilization compared to the general farmer’s fertilization rate. On the other hand, the high-N treatments represented the typical fertilization rate used by farmers in north-central (fluvo-aquic soil) and northeast (black soil) China. In treatments, all the N, P, and potassium (K) were applied as a basal application. Urea was used as the source of N fertilizer in this study, and no organic fertilizer was used. Other agricultural practices, such as tillage and pesticide application, remained the same for both the moderate-N fertilization and high-N fertilization treatments. In north-central China, the crop rotation system consisted of winter wheat followed by summer maize, while in northeast China, a summer maize monocropping system was practiced. In September 2018, we sampled soil from both study sites. Typically, five soil cores with a diameter of 2 cm were taken from each experimental plot, reaching a depth of 0–20 cm. These cores were combined to form a composite sample. A portion of the composite sample was preserved at a temperature of −80°C for subsequent molecular analysis.

The concentrations of soil organic carbon (SOC) and total nitrogen (TN) were determined through dichromate oxidation (Kalembasa and Jenkinson, 1973) and Kjeldahl digestion (Bottomley et al., 2020), respectively. The concentration of soil inorganic N, specifically NO₃⁻N and NH₄⁺N, was determined by analyzing a 12 g fresh soil sample mixed with a 0.01 mol L^–1 CaCl₂ solution at a 1:10 ratio. The analysis was conducted using continuous flow analysis with the Foss FlAstar 5,000 instrument from Sweden. The pH of the soil was determined by employing a compound electrode (PE-10, Sartorius, Germany) and utilizing a soil-to-water ratio of 1:2.5. The fumigation extraction method using 0.5 M K₂SO₄ was employed to conduct the microbial biomass C (MBC) and microbial biomass N (MBN) analyses (Brookes et al., 1985), and the results were determined using a total organic C/N analyzer (Multi N/C 3100/HT1300, Analytik Jena AG).

Comprehensive details regarding the EEA that are associated with P, N and C cycling are presented in Table 1. The fluorescence-based protocols as described by Ai et al. (2012) were utilized to determine the activities of extracellular enzymes associated with nitrogen and carbon. The units of measurement used were expressed as nmol h⁻¹ g⁻¹. In summary, a quantity of 1 gram of freshly collected soil was subjected to homogenization in 100 milliliters of sterilized water with the aid of a polytron homogenizer. Subsequently, a magnetic stirrer was employed to ensure the homogeneity of the suspension. The experimental procedure involved the introduction of a suspension sample, sterilized water, 200 μM of 4-methylumbelliferyl-linked substrates, and 10 μM of references into the wells of a black 96-well microplate. The microplates were incubated in darkness at a temperature of 25°C for a duration of 4 h while being covered. Following the incubation period, a 10 μl aliquot of a 1 M NaOH solution was promptly introduced into every well of the microplate to stop the enzymatic reaction. The quantification of fluorescence was conducted utilizing a microplate fluorometer (Scientific Fluoroskan Ascent FL, Thermo Fisher Scientific, Waltham, MA, USA), with the implementation of 365 nm excitation and 450 nm emission filters.

The soil microbial community and microbial biomass were assessed using phospholipid fatty acid analysis (PLFA), following the method described by Wu et al. (2009). In brief, 3 g of freeze-dried soil samples were utilized for PLFA extraction. A chloroform/methanol/citric acid buffer (volume ratio of 1:2:0.8, pH 4.0) was used for the extraction process. The polar lipids were separated from glycolipids and neutral lipids by eluting with acetone and chloroform, respectively, on a silica-bonded phase column. To serve as an internal standard, nonadecanoic acid methy ester (19:0) was added. The polar lipids were further converted to fatty acid methyl esters (FAMEs) through mild alkaline methanolysis. The resulting dried FAMEs were dissolved in n-hexane and subsequently analyzed and identified using the MIDI Sherlock microbial identification system version 4.5 (MIDI Inc., Newark, DE, USA) and gas chromatography (Agilent N6890). The concentrations of PLFAs were expressed as nmol g⁻¹ of dry soil. The total microbial biomass was determined by calculating the sum of all PLFA concentrations (nmol g⁻¹), while the relative abundance of each PLFA was determined by its percentage mole abundance in each sample. Various taxonomic groups of microorganisms were identified using specific PLFAs as biomarkers, including bacteria, gram-positive (G+) and gram-negative (G–) bacteria, fungi, arbuscular mycorrhizal fungi (AMF), saprophytic fungi, and actinomycetes. Table 2 shows the PLFAs that were used as signatures for each taxonomic group of microorganisms based on fatty acid biomarker data that has already been published (Ai et al., 2012; Willers et al., 2015; Dai et al., 2019).

To assess the significance of each measured variable, such as soil chemical properties, microbial populations, and enzyme activities a one-way analysis of variance (ANOVA) was conducted. Fisher’s least significant difference (LSD) test was used to compare the means of different treatments, with a significance level set at p ≤ 0.05. The statistical analysis was carried out using SPSS version 24. To differentiate the individual contributions of soil types and N fertilization on soil enzyme activities and microbial populations, a permutational multivariate analysis of variance (PERMANOVA) was performed using Primer software version 6.0. We utilized principal component analysis (PCA) to detect alterations in enzyme activities and microbial community structure following nitrogen (N) addition in two distinct soil types using Conoco 5. To examine the interrelationships between soil properties, enzyme activities, and microbial community composition, we employed a statistical technique called partial least squares path modeling (PLSPM). This approach allows us to visualize and analyze the cause-and-effect relationships among observed and latent variables. In PLSPM, path coefficients represent the direct effects, indicating the direction and strength of the linear relationship between variables. Additionally, indirect effects are considered, which involve the multiplied path coefficients between a predictor and a response variable, considering all possible paths except the direct effects. To validate the estimates of the coefficients of determination (R²) and the path coefficients in our model, we utilized the “plspm” package in R software version 3.4.4. The validation process involved performing 1,000 bootstraps, which helped assess the stability and reliability of the model estimates.

The impact of N fertilization on soil chemical properties and maize yield was investigated in two different soil types: fluvo-aquic soil and black soil (Table 3). In the fluvo-aquic soil, pH values did not show significant differences among the treatments. However, in the black soil, both moderate-N and high-N treatments resulted in a significant decrease in pH compared to the control. Total N content was significantly higher in the fluvo-aquic soil with both moderate-N and high-N treatments. In the black soil, nitrate nitrogen (NO₃⁻N) levels were significantly increased by both moderate-N and high-N treatments, while Total N, SOC, and MBC content showed slight differences among the treatments. MBN content was significantly impacted by N application in fluvo-aquic soil but not in black soil. Regarding maize yield, the moderate-N treatment resulted in the highest yield in both soil types, followed by the high-N treatment. However, the crop yield was more responsive to N addition in black soil.

The impact of nitrogen fertilization on soil microbial communities was assessed in both fluvo-aquic and black soils (Table 4). In the fluvo-aquic soil, there were no significant differences in the abundance of bacteria, fungi, saprophytic fungi, and actinomycetes among the control, moderate-N, and high-N treatments. However, the abundance of AMF showed a significant decrease with an increasing nitrogen fertilization rate. Similarly, in the black soil, the abundance of bacteria did not significantly differ among the control, moderate-N, and high-N treatments. However, the abundance of fungi and AMF significantly decreased with increasing nitrogen fertilization rates. The ratio of Fungi:bacteria (F:B) decreased while the GP:GN ratio increased with N fertilization. PCA was utilized to evaluate the distinct differentiation of microbial community composition across two soil types. PCA demonstrated a clear separation of microbial communities in both fluvo-aquic and black soil (Figure 1A). To separate the individual impacts of N fertilization and soil types on microbial community composition, we employed PERMANOVA. The outcomes of PERMANOVA revealed that soil type accounted for most of the variation, specifically 42%, in microbial community composition (Figure 2A). Conversely, N fertilization contributed 14% of the variation.

The impact of nitrogen fertilization on extracellular enzyme activities was assessed in two soil types. The results suggest that nitrogen fertilization did not have a significant impact on soil enzyme activities except Phosphatase and LAP in fluvo-aquic soil. In contrast, in the black soil, significant differences were observed in several enzyme activities with nitrogen fertilization (Figures 3 4; Table 5); for example, BG, BX, and BC activities were significantly impacted by N application. Phosphatase and NAG activities did not show significant differences among the treatments. PCA was utilized to evaluate the distinct changes in EEA across the two soil types. PCA demonstrated clear separation of EEA in both fluvo-aquic soil and black soil (Figure 1B). The results of PERMANOVA revealed that soil type accounted for the majority of the variation, specifically 56%, in EEA (Figure 2B). Conversely, N fertilization contributed 11% of the variation. Overall, the results indicate that the influence of nitrogen fertilization on soil enzyme activities varied depending on the soil type.

To investigate the relationships between N fertilization, maize yield, soil properties, soil bacterial and fungal communities, and EEA, a PLSPM analysis was conducted. The findings in fluvo-aquic soils revealed that the addition of N exhibited a significant direct relationship with soil N contents and pH (Figure 5A). Moreover, soil pH significantly influenced both fungal and bacterial communities. It was observed that soil N contents had a significant direct association with crop yield. Additionally, SOC and soil fungi exhibited a significant relationship with C-cycling enzymes, while no significant relationship was found between soil bacteria and C-cycling enzymes. Notably, C-cycling enzymes displayed a significant direct relationship with N-cycling enzymes. In the context of black soil, the N application exhibited a significant direct association with soil pH and SOC (Figure 5B). Soil N contents exerted a significant influence on crop yield. Furthermore, soil pH demonstrated a significant direct correlation with soil fungal communities and EEA involved in C cycling, whereas no significant relationship was observed with soil bacterial communities. Importantly, a significant direct relationship was observed between C-cycling enzymes and N-cycling enzymes.

The application of N fertilizer impacted soil chemical properties and increased nutrient contents across both soil types (Table 3). Earlier research has indicated that extended application of nitrogen fertilizers resulted in an enhanced level of SOC, overall N content, and particulate organic matter (Dai et al., 2019; Yu et al., 2019). It has been established that N fertilization can directly affect soil physiochemical properties, which in turn can have a significant impact on crop production. In particular, PLSPM indicated that the soil chemical properties were influenced by the addition of N, which in turn had a direct or indirect effect on the yield of maize. The levels of available N, TN, and SOC were identified as significant factors in regulating maize grain yield, either through individual or interactive effects (Figure 5). A study by Yu et al. (2019), found that SOC played a significant role in enhancing crop yields through mediating microbial biomass and soil enzyme activities. The findings of Luo et al. (2018) in a comprehensive meta-analysis indicate that N fertilization can enhance crop production by augmenting SOC and TN levels, as well as biological soil enzyme activity across diverse cropping systems worldwide.

The N fertilization induced changes can have an impact on the soil microbial community, leading to significant modifications in soil microbial characteristics such as total microbial biomass, MBC, and MBN contents. In the present study, divergent responses of MBC were observed in the two soils (Table 2). The divergent responses of MBC to N fertilization in the two soils could be related to differences in their initial microbial communities and nutrient utilization strategies (Sun et al., 2021). Some microbial groups are specialized in efficiently utilizing N and other nutrients, while others may rely on different carbon sources (Xu et al., 2021). Additionally, the initial nutrient status of the soils can also influence microbial responses to N fertilization. If one soil already has a relatively high nutrient content, including nitrogen, the additional nitrogen from the fertilization might not have as pronounced an effect on microbial growth compared to a soil with low initial nutrient levels (Chen et al., 2020). It has been established that the application of N fertilizer results in a significant increase in soil MBN and MBC contents, regardless of soil type (Yu et al., 2019). Conversely, in a meta-analysis conducted by Jian et al. (2016), it was found that the application of N fertilization resulted in a decrease in MBC content. Results in Table 3 showed that the application of N fertilization had a greater impact on reducing the pH of black soil compared to fluvo-aquic soil. One potential reason for the greater impact of N fertilization on reducing the pH of black soil compared to fluvo-aquic soil could be related to differences in the buffering capacity of the two soil types. Buffering capacity refers to a soil’s ability to resist changes in pH when acidic or alkaline substances are added. Soils with higher buffering capacity can maintain a more stable pH even when external factors, such as fertilizer application, attempt to alter it (Lopes et al., 2015). In addition, distinct soil types exhibit distinct initial characteristics (Staley et al., 2018; Yu et al., 2019). Consequently, the impact of N fertilization on different soil properties was found to be diverse (Yu et al., 2019).

In the present study, the higher N rate resulted in substantial soil acidification in fluvo-aquic soil and black soil. The application of additional N treatments resulted in a reduction in pH levels, irrespective of the soil type, which consequently led to soil acidification (Guo et al., 2010). The phenomenon of soil acidification has been observed to have significant impacts on the availability of nutrients for plants and the decomposition of soil organic matter (Treseder, 2008). In order to ensure the preservation of soil quality and mitigate the environmental consequences associated with agriculture, it is imperative to prioritize the reduction of fertilizer rates, specifically N fertilizer. This approach is crucial for maintaining optimal crop productivity in a sustainable manner. In pursuit of this objective, the Chinese government has implemented a comprehensive initiative known as the Zero Growth of Chemical Fertilizer Use by 2020 (Shuqin and Fang, 2018). The implementation of such initiatives is anticipated to have multiple benefits, including the maintenance of elevated crop productivity levels, improved efficiency in nutrient utilization, and the safeguarding of the environment against the detrimental impacts associated with agricultural practices.

The potential impact of N enrichment on soil ecosystem stability is a topic of concern in ecological research. It is well established that sustaining of soil productivity and ecological equilibrium in agroecosystems is heavily reliant on the abundance and diversity of soil microorganisms (Zhou et al., 2017). Specifically, the reduction in abundance and diversity resulting from N enrichment may have negative consequences for the interactions between above- and below-ground ecosystems (Tilman et al., 2006; Zhou et al., 2017).

In the present study significant reduction of fungal abundance in response to N addition was witnessed in black soil while no significant shifts were observed in fluvo-aquic soils. The specific shifts in fungal microbial community composition are attributed to ecosystem N status, where N additions have little effect the abundance of fungal communities in fluvo-aquic soil (Wang et al., 2018). The response of fungal microbial communities to N addition can also be influenced by changes in soil pH. N fertilizers can lead to increased soil acidity due to the nitrification process. Fungi generally prefer slightly acidic to neutral pH conditions, and the reduction of pH in black soil might have negatively impacted fungal growth, contributing to the observed decline in their abundance. In addition, Fungi and bacteria often compete for limited resources in the soil. Nitrogen addition could have intensified this competition, affecting fungal communities more severely in black (Bluhm et al., 2019).

Despite of earlier reports that bacterial growth is more susceptible to low pH compared to fungi (Pennanen et al., 1999 Rousk et al., 2010), our observations did not reveal a significant reduction in bacterial abundance or increase in F:B ratio in plots with lower pH levels (Table 4). The findings of our study indicated the possibility that certain bacterial groups have developed significant adaptations to environmental stressors induced by nitrogen in the study area. Our results revealed that N input consistently decreased F:B ratio across both the soils but particularly in black soil. The study conducted by Strickland and Rousk (2010) found that the observed responses between the F:B and microbial C:N to N addition were primarily due to the higher biomass C:N of the fungal communities compared to the bacterial communities. This result holds significant implications. Firstly, the microbial C:N ratio can serve as an indicator to determine the relative abundance of fungi over bacteria in a given environment (Nannipieri et al., 2003). Secondly, extracellular enzymes serve as the immediate agents for the decomposition of soil organic matter. These enzymes’ activities can serve as indicators for soil nutrient cycling, and microbial soil respiration (Jian et al., 2016). Therefore, the reduction in microbial biomass does not necessarily imply that the addition of N prevents the growth of microorganisms (Zhou et al., 2017).

The application of N fertilizer resulted in a shift in the GP:GN ratio of bacteria (Table 4). Schimel et al. (2007) demonstrated that within the domain of microorganisms, GP bacteria possess a strong and intricately connected peptidoglycan cell wall, whereas GN bacteria exhibit a single-layer cell wall and an outer membrane. However, earlier research has identified ecological and physiological differences between GP and GN bacteria. The predominant microorganisms that facilitate the decomposition of litter and soil organic matter are GP bacteria and fungi. A significant number of microorganisms that perform specific or limited functions in soil are Gram-negative (GN) bacteria (Schimel, 1995). According to earlier research, it is believed that GP bacteria possess a greater inherent resistance to environmental stress when compared to GN bacteria (Schimel, 1995; Schimel et al., 2007). The observed increase in GP:GN appears to be consistent with the increase in soil available N resulting from the N application, as indicated in Table 4. The study conducted by Fanin et al. (2013) highlights the significance of P-driven stoichiometric control in understanding the spatiotemporal heterogeneity of the GP:GN. In particular, the positive response of the GP:GN ratio has been observed to potentially facilitate the breakdown of recalcitrant SOM and enhance C acquisition by soil microbiota (Zhou et al., 2017).

Soil EEA serves as significant biological indicators of soil biodiversity, ecosystem function, and soil fertility levels by reflecting nutrient cycling activities (Sinsabaugh et al., 2009; Wallenstein and Burns, 2011). We observed that glycosidases activities (β-cellobiosidase, β-glucosidase, β-xylosidase, and α-glucosidase) were enhanced following the addition of N (Figures 3, 4). A meta-analysis conducted on the effects of N additions on glycosidase activities across various ecosystems showed that anthropogenic N input stimulated the activities of glycosidases (Jian et al., 2016). In present study, N fertilization plots had enhanced activities of Phosphatase. The production of AP requires elevated levels of N, as noted by Treseder and Vitousek (2001). Consequently, the application of nitrogen amendments may have potentially stimulated microbial activity, resulting in increased AP production. Furthermore, the acidification of soil induced by N application could potentially contribute to the mineral soil’s phosphorus binding capacity and subsequently decrease the accessibility of phosphorus to microorganisms. In order to fulfill the necessary nutrient requirements for phosphorus, microorganisms have the ability to synthesize a greater quantity of phosphatase (Sinsabaugh et al., 2009). We further observed that N-treatment plots exhibited an increase in specific enzyme activities while some enzymes activities stay neutral. This suggests that microbial populations carried out a more pronounced enzyme activity in N treated plots (Francioli et al., 2016). Moreover, it is has been observed that specific extracellular enzyme activities are better indicator of the metabolic status of a microbial community in acquiring C and nutrients, as compared to absolute enzyme activities which only reflect gross microbial activity (Raiesi and Beheshti, 2014).

The findings of our study indicated that the microbial community structure and enzyme activities of soil were primarily influenced by soil type, rather than N fertilization (Figures 2A, B). Earlier research claimed that the primary factor influencing the significant variation in the microbial community composition and enzyme activities on a large scale is geographic location/soil type rather than fertilization (Zhang et al., 2021). Additionally, it has been suggested that environmental variables resulting from N fertilization may have a lesser impact on this variation (Chen et al., 2016). Based on soil sampled across 03 different soil types in China, Zhang et al. (2021), reported that geographical distance explained 61.60% of the variance in microbial community composition, but N fertilization only accounted for 1.52%.

Nitrogen fertilization impacts soil microbial activity, nitrogen, and carbon dynamics, indicating that soil microbial communities can regulate nitrogen-use efficiency and carbon dynamics efficiency to mitigate resource imbalances (Mooshammer et al., 2014). The maintenance of ecosystem sustainability and productivity is heavily influenced by the presence of a suitable community structure and ample diversity of soil microbes, which play a crucial role in regulating various soil biogeochemical processes (Schimel, 1995; Strickland and Rousk, 2010; Chaparro et al., 2012). In the present study, soil bacterial and fungal communities were significantly impacted by soil pH (Figure 5), which indicated that soil acidification rather than N availability shaped soil fungal and bacterial communities. Earlier studies have indicated a strong correlation between shifts in microbial community composition and soil pH (Francioli et al., 2016; Zhou et al., 2017; Zhang et al., 2021). Furthermore, soil acidification has been found to have a direct impact on the composition of soil microbial communities (Dai et al., 2019). However, on the other hand, it has been reported that N addition-induced changes in soil properties could exert positive (Zhou et al., 2017) and neutral (Turner and Joseph Wright, 2014) effects on soil microbial communities.

Interestingly, we found that there was a significant positive correlation between C-cycling enzyme activities and N-cycling enzyme activities (Figures 5A,B). This result is consistent with the research conducted by Dai et al. (2019), which demonstrated that C-cycling enzyme activity has significant relationship with the N-cycling enzymes. The observed phenomenon can be attributed to the impact of low soil C:N ratios on the activities of C-cycling enzymes. This leads to an increase in the degradation of cellulose, lignin, and other carbon sources that are typically challenging to decompose. As a result, there is a subsequent increase in the availability of C and the activities of N-related enzymes. Our study revealed that there was a significant positive relationship between soil pH and the activities of C cycling enzymes (Figure 5B). Enzymes are sensitive to changes in environmental conditions, including pH. Many C-cycling enzymes have an optimal pH range at which they function most efficiently. When the soil pH falls within the optimal range for these enzymes, their activity is enhanced, leading to increased decomposition and cycling of organic carbon compounds (Puissant et al., 2019). Soil pH can affect the solubility of minerals, which, in turn, can influence enzyme activity. Some minerals act as cofactors for C-cycling enzymes, and their availability might change with soil pH (Sheng et al., 2022). The results of this study are similar to those of Dai et al. (2019), who found a significant correlation between the activity of extracellular c-cycling enzymes and the pH of the soil.

It was further observed that across both soil types, fungal communities were significantly linked to C-cycling enzymes but not N-cycling enzymes (Figures 5A,B). Studies have claimed that soil fungal communities dominate SOC turnover and that fungi have a greater impact on soil carbon cycling than bacterial microbes in both acidic and basic soils (Xiao et al., 2018). The relationship between soil carbon cycling enzyme activities and soil fungi is critical for understanding carbon dynamics in terrestrial ecosystems. Soil carbon cycling enzyme activities are closely linked to the decomposition and transformation of organic matter, which affect the storage and release of carbon in the soil. Soil fungi play a key role in this process, as they are major contributors to organic matter decomposition and nutrient cycling in soils (Xiao et al., 2018). We further witnessed that C-cycling enzymes were highly linked to SOC (Figure 5A). SOC represents a reservoir of organic carbon in the soil. As the SOC content increases, it provides a continuous supply of carbon-rich substrates for microbial activity. Increased microbial activity, in turn, leads to higher enzyme production, facilitating further decomposition of organic matter and contributing to the maintenance of higher SOC levels. This creates a positive feedback loop where higher SOC leads to increased enzyme activity, which, in turn, supports the preservation and accumulation of more SOC (Wang et al., 2022).