Abstract
Background:
Atmospheric nitrogen (N) deposition can strongly alter soil organic carbon (SOC) stabilization by regulating microbial-derived carbon (C) formation. Microbial necromass carbon (MNC) and glomalin-related soil protein (GRSP) are important contributors to SOC persistence, yet how long-term N enrichment affects these microbial-derived C pools and their microbial drivers across soil depths remains unclear in alpine forest ecosystems.
Methods:
We investigated the effects of six-year N addition on bacterial necromass C (BNC), fungal necromass C (FNC), MNC, total GRSP (T-GRSP), their SOC-normalized proportions, and associated soil and microbial mechanisms in an alpine coniferous forest on the southeastern Xizang Plateau. Four N addition treatments—control, low N, medium N, and high N addition—were established, and soils were collected from the 0–20 and 20–40 cm layers. Random forest analysis, bacterial–fungal co-occurrence networks, and microbial module analysis were used to identify variables associated with microbial-derived C accumulation.
Results:
Low and medium N addition increased BNC, FNC, MNC, and T-GRSP, whereas these positive effects were weakened under high N addition. Although microbial-derived C contents were consistently higher in surface soil, the SOC-normalized contribution of microbial necromass responded more strongly to N addition in the 20–40 cm layer, while T-GRSP:SOC remained relatively stable across treatments and depths. Random forest analysis suggested that soil pH, nutrient availability, and C:N:P stoichiometry were closely associated with MNC and T-GRSP variation, with additional associations involving enzyme activities, microbial diversity, and community composition, particularly in deeper soil. Exploratory depth-specific bacterial–fungal co-occurrence networks indicated fewer associations and a higher proportion of negative links in deeper soil. Module analysis further identified specific bacterial and fungal modules associated with SOC fractions, microbial necromass, and GRSP-related indicators.
Conclusion:
Moderate N enrichment was associated with greater microbial-derived C accumulation, whereas high ammonium-sulfate N addition showed weaker positive effects accompanied by nutrient imbalance, soil acidification, and altered microbial attributes. These findings highlight the importance of integrating microbial residues, GRSP, soil nutrient stoichiometry, and microbial co-occurrence patterns when assessing SOC stabilization under increasing N deposition.
1 Introduction
Atmospheric nitrogen (N) deposition has increased substantially due to intensified anthropogenic activities and is expected to continue altering terrestrial carbon (C) cycling under global environmental change (Galloway et al., 2008; Fowler et al., 2013; Liu et al., 2013). Forest soils store a large proportion of terrestrial organic C (Pan et al., 2011), and even small changes in soil organic carbon (SOC) formation and stabilization can strongly influence ecosystem C balance and climate feedbacks (Bradford et al., 2016; Hicks Pries et al., 2023). In N-limited forest ecosystems, exogenous N inputs may stimulate plant productivity, litter inputs, and microbial growth, thereby promoting SOC accumulation (Wang et al., 2024). However, excessive N enrichment can also induce soil acidification, nutrient imbalance, phosphorus (P) limitation, and shifts in microbial metabolism, ultimately weakening microbial growth and SOC stabilization (Liu et al., 2025; Tang et al., 2023). Therefore, understanding how long-term N addition regulates biologically mediated SOC formation is critical for predicting forest soil C dynamics under increasing atmospheric N deposition.
Emerging concepts emphasize that microbial anabolism is central to SOC formation and persistence (Whalen et al., 2024). Soil microorganisms transform plant-derived C into microbial biomass, extracellular products, and necromass, which can be subsequently stabilized through mineral association, aggregation, and chemical protection (Wang et al., 2021; Kallenbach et al., 2016). Microbial necromass carbon (MNC), commonly estimated using amino sugar biomarkers, has been recognized as a major component of SOC and an important indicator of microbial-derived C accumulation (Hu et al., 2024; Joergensen, 2018). Bacterial necromass carbon (BNC) and fungal necromass carbon (FNC) represent different microbial sources of necromass, and their relative contributions to SOC may vary with nutrient availability, microbial community structure, and soil depth (Takele et al., 2025; Liao et al., 2022; Zhu X. et al., 2024). In parallel, glomalin-related soil protein (GRSP), particularly total GRSP (T-GRSP), is widely used as an operational indicator of mycorrhizal-associated proteinaceous compounds involved in SOC stabilization through aggregate formation and physicochemical protection (Rillig, 2004; Lu et al., 2024). GRSP can contribute to soil aggregation, organo-mineral protection, and the persistence of organic C (Ao et al., 2025). Thus, MNC and T-GRSP together provide complementary perspectives for evaluating microbial contributions to SOC stabilization. Nitrogen enrichment may strongly affect microbial-derived C pools, but the magnitude and direction of these effects can vary with N addition rate, soil nutrient status, microbial community attributes, and ecosystem type (Hu et al., 2022). Moderate N addition can enhance microbial biomass and turnover by alleviating N limitation and increasing labile substrate availability, thereby increasing microbial necromass accumulation (Wang et al., 2024; Hu et al., 2022). In contrast, high N inputs may suppress microbial growth or alter microbial resource allocation by intensifying possible relative P limitation and disrupting soil C:N:P stoichiometry (Xu et al., 2022; Luo et al., 2022; Liu et al., 2025). Similarly, GRSP accumulation may respond to N addition through changes in mycorrhizal activity, microbial community composition, soil nutrient status, and aggregate-associated C protection (Zhang J. et al., 2015; Huang et al., 2022; Wang et al., 2023). However, previous studies have reported inconsistent responses of microbial necromass and GRSP to N enrichment, suggesting that their responses depend on N addition rate, soil nutrient status, microbial community attributes, and ecosystem type (Hu et al., 2023). A better mechanistic understanding is therefore needed to clarify how microbial-derived C pools respond to long-term N enrichment in forest soils.
Soil depth is another key factor regulating microbial-derived C accumulation and its contribution to SOC (Takele et al., 2025). Surface soils generally contain greater plant-derived C inputs, higher microbial biomass, and more active microbial processes, whereas deeper soils are often characterized by lower substrate availability, stronger mineral protection, and slower C turnover (Hicks Pries et al., 2023; Naylor et al., 2022). These vertical gradients may lead to distinct responses of MNC and GRSP to N addition. Importantly, changes in the absolute contents of microbial-derived C pools do not necessarily reflect their proportional contributions to SOC (Ni et al., 2020), because background SOC pools differ strongly between surface and deeper soils. Therefore, evaluating both microbial-derived C contents and their SOC-normalized contributions across soil depths is essential for understanding the role of microbial residues and GRSP in SOC stabilization under N enrichment. Microbial community structure and co-occurrence patterns may further be associated with the response of microbial-derived C to N addition (Ma et al., 2022; Zhu X. et al., 2024; Wang et al., 2026). Bacterial and fungal communities differ in nutrient acquisition strategies, substrate use, turnover rates, and necromass formation potential (Wang and Kuzyakov, 2024; Buckeridge et al., 2022). N addition can shift bacterial and fungal diversity and community composition (Ma et al., 2022), thereby altering the biological sources of microbial residues and GRSP-related compounds (Zhu X. et al., 2024). In addition, taxa with similar environmental preferences or functional roles may form tightly connected modules within microbial co-occurrence networks (Zhang et al., 2025). These modules may respond coherently to N addition and soil depth and may be more closely linked to C stabilization than whole-community diversity metrics (Banerjee et al., 2021; Zhang M. et al., 2023). However, most previous studies have focused on microbial diversity or community composition alone, while less attention has been paid to whether bacterial–fungal network reorganization and key microbial modules are associated with MNC, GRSP, and SOC fractions under long-term N addition. Compared with whole-community diversity metrics, network modules may better capture coordinated microbial assemblages that respond to environmental filtering and are associated with microbial residue formation, GRSP accumulation, and SOC stabilization (Yang et al., 2023).
Alpine forest ecosystems are important components of regional C cycling and ecosystem functioning because their soils store substantial amounts of organic C (Pan et al., 2011). At the same time, these high-elevation ecosystems are highly sensitive to environmental change, including increasing atmospheric N deposition (Fowler et al., 2013). However, the mechanisms by which long-term N enrichment regulates microbial-derived C accumulation and SOC stabilization in alpine forest soils remain poorly understood. In this study, we conducted a long-term N addition experiment in an alpine coniferous forest and examined BNC, FNC, MNC, T-GRSP, their contributions to SOC, bacterial and fungal communities, soil nutrient stoichiometry, enzyme activities, microbial biomass, and bacterial–fungal co-occurrence networks in the 0–20 and 20–40 cm soil layers. We aimed to determine how N addition affects microbial-derived C pools and to identify the soil and microbial variables associated with these responses. Bacterial–fungal co-occurrence networks and microbial modules were further used as exploratory tools to examine whether depth-dependent microbial assemblages were associated with SOC fractions, microbial necromass, and GRSP-related indicators. Based on the expectation that moderate N addition may alleviate N limitation and enhance microbial biomass, whereas excessive N input may induce soil acidification and nutrient imbalance, we proposed three testable hypotheses. First, low and medium N addition would increase MNC and T-GRSP by improving soil nutrient availability, labile C and N supply, and microbial biomass, whereas high N addition would result in weaker positive responses. Second, soil depth would regulate the response of microbial-derived C pools to N addition; specifically, absolute MNC and T-GRSP contents would be higher in the 0–20 cm layer, whereas SOC-normalized microbial necromass responses would be stronger in the 20–40 cm layer because of lower background SOC and microbial biomass. Third, soil nutrient availability and C:N:P stoichiometry would be the primary predictors associated with MNC and T-GRSP variation, while enzyme activities, microbial diversity, and community composition would show stronger associations in deeper soil.
2 Material and methods
2.1 Site description, experimental design, and soil sampling
The experiment was conducted at the Sejila Mountain Forest Ecosystem National Long-Term Observation and Research Station in Nyingchi, Xizang Autonomous Region, China 29°39′ N, 94°43′ E; 3850 m a.s.l. (Figure 1). The study area has an alpine temperate semi-humid climate, with a mean annual temperature of −0.73 °C and a mean annual precipitation of 1,134 mm, approximately 85% of which occurs from June to September. According to the USDA soil taxonomy, the soil is classified as Fluvic Cambisol, with a sandy clay loam texture comprising 68.8% sand, 0.4% silt, and 30.8% clay. The study site is a typical alpine coniferous forest dominated by Abies georgei var. smithii, with an average canopy height of 28.2 m and a mean diameter at breast height of 43.9 cm. The vegetation is vertically structured into tree, shrub, and herb layers, with coniferous trees dominating the overstory.
Figure 1
The N addition experiment was established in June 2019 using a randomized block design. Four N addition treatments were established: control without N addition (CK, 0 kg N ha−1 yr−1), low N addition (LN, 10 kg N ha−1 yr−1), medium N addition (MN, 15 kg N ha−1 yr−1), and high N addition (HN, 20 kg N ha−1 yr−1). Each treatment was replicated three times, resulting in 12 plots in total. Each plot measured 20 m × 20 m, and adjacent plots were separated by at least 25 m to minimize edge effects and treatment interference. Nitrogen was applied as ammonium sulfate (NH4)2SO4. In mid-August of each year, the required amount of fertilizer for each treatment was dissolved in 16 L of water, equivalent to approximately 2 mm of precipitation, and evenly sprayed onto the corresponding plot using a backpack sprayer. Accordingly, 1.886, 2.829, and 3.772 kg of (NH4)2SO4 were applied to the LN, MN, and HN plots, respectively. Control plots received the same amount of water without fertilizer to avoid potential effects of unequal water addition. The N addition rates were selected to represent a gradient of low, moderate, and relatively high N enrichment rather than conventional fertilization. These rates were designed to simulate possible increases in atmospheric N inputs and are comparable to N addition levels commonly used in forest N-enrichment experiments (Zhu L. et al., 2024). Therefore, the treatments allowed us to evaluate how alpine coniferous forest soils respond to progressively increasing N availability. We acknowledge that atmospheric N deposition occurs continuously through wet and dry deposition events, whereas our experimental N was applied once per year during the growing season. Thus, this approach simulated the cumulative annual N input rather than the exact temporal pattern of natural atmospheric deposition. The single annual application was adopted to maintain consistency among years and to minimize field disturbance in this remote alpine forest site.
Soil samples were collected in early August 2025 from the 0–20 and 20–40 cm soil layers, before the scheduled N addition in 2025. Therefore, the samples were collected approximately one year after the previous N application in mid-August 2024 and before any new N input in the sampling year. Thus, the observed soil and microbial responses mainly reflected the cumulative effects of six previous annual N additions rather than an immediate short-term N pulse following fertilization. Before sampling, the surface organic layer, visible litter, and debris were carefully removed. In each plot, five soil cores were collected from each soil depth using a hand auger with a diameter of 5 cm and then thoroughly mixed to form one composite sample per depth. In total, 24 composite soil samples were obtained from four N addition treatments, three replicates, and two soil depths. All samples were placed on ice and immediately transported to the laboratory. Visible roots, stones, and plant residues were removed, and the soils were passed through a 2-mm sieve. Each sample was then divided into four subsamples: one was stored at −80 °C for DNA extraction, one was stored at 4 °C for enzyme activity and microbial biomass analyses within one week, one was freeze-dried for microbial necromass biomarker analysis, and the remaining soil was air-dried or stored at room temperature for soil physicochemical properties and glomalin-related soil protein analyses.
2.2 Soil physicochemical properties, microbial biomass and enzyme activities
Soil physicochemical properties were determined according to standard soil agrochemical analysis methods, with minor modifications (Bao et al., 1988). Soil pH was measured in a soil-to-distilled water suspension at a ratio of 1:2.5 using a glass electrode pH meter. Soil organic carbon (SOC) and total nitrogen (TN) concentrations were determined using an elemental analyzer (Vario MACRO cube, Elementar, Germany). Soil total phosphorus (TP) was determined after H2SO4–HclO4 digestion using the molybdenum–antimony colorimetric method. Soil available phosphorus (AP) was extracted with 0.05 mol L−1 HCl and 0.025 mol L−1 H2SO4 and then measured colorimetrically using the molybdenum blue method. The SOC:TN, SOC:TP, and TN:TP ratios were calculated and used as soil stoichiometric indicators. Particulate organic carbon (POC) and mineral-associated organic carbon (MAOC) were also determined to characterize SOC fractions, and detailed procedures are provided in Supplementary methods.
Dissolved organic carbon (DOC) and dissolved organic nitrogen (DON) were extracted from fresh soil with 0.5 mol L−1 K2SO4 at a soil-to-extractant ratio of 1:4. After shaking and filtration, DOC and DON concentrations in the extracts were measured using a total organic carbon analyzer. Microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) were determined using the chloroform fumigation–extraction method. Briefly, paired fumigated and non-fumigated fresh soil samples were extracted with 0.5 mol L−1 K2SO4, and the differences in extractable C and N between fumigated and non-fumigated samples were used to calculate MBC and MBN using conversion coefficients of 0.45 and 0.54 (Brookes et al., 1982; Vance et al., 1987), respectively.
Potential extracellular enzyme activities related to C, N, and P acquisition were measured using a 96-well microplate method following previously established protocols with minor modifications (Jing et al., 2016). Five hydrolytic enzymes were measured, including β-1,4-glucosidase (BG) and cellobiohydrolase (CBH) for C acquisition, β-1,4-N-acetylglucosaminidase (NAG) and L-leucine aminopeptidase (LAP) for N acquisition, and alkaline phosphatase (ALP) for P acquisition.
Briefly, fresh soil suspensions were prepared and incubated with the corresponding enzyme substrates in 96-well microplates under controlled conditions. After incubation, enzyme activities were determined using a microplate reader. Enzyme activities were expressed as U g−1. To represent the overall resource-acquisition enzyme activities, C-, N-, and P-acquiring extracellular enzyme activities and microbial necromass carbon were calculated using Equations 1–6, as detailed below.
Where CAE, NAE, and PAE represent C-, N-, and P-acquiring extracellular enzyme activities, respectively. These composite enzyme variables were used in subsequent correlation and random forest analyses.
2.3 Amino sugar analysis and calculation of microbial necromass carbon
Microbial necromass carbon was estimated using amino sugar biomarkers. Freeze-dried soil samples were used for amino sugar analysis following previously established methods with minor modifications. Briefly, amino sugars, including glucosamine (GluN), galactosamine (GalN), and muramic acid (MurA), were extracted and quantified to estimate microbial residue-derived carbon. Among these biomarkers, GluN was mainly used to estimate fungal residues, whereas MurA was used as a specific biomarker for bacterial residues. Bacterial necromass carbon (BNC) and fungal necromass carbon (FNC) were calculated from amino sugar contents using established equations (Appuhn and Joergensen, 2006; Joergensen, 2018):
where GluN and MurA represent the contents of glucosamine and muramic acid, respectively. The constants 179.17 and 251.23 are the molecular weights of glucosamine and muramic acid, respectively. The factor of 9 was used to convert fungal glucosamine to fungal necromass carbon, and the factor of 45 was used to convert muramic acid to bacterial necromass carbon. Total microbial necromass carbon (MNC) was calculated as the sum of BNC and FNC:
BNC, FNC, and MNC were expressed as g kg−1 soil. Their proportional contributions to SOC were calculated as BNC:SOC, FNC:SOC, and MNC:SOC ratios.
2.4 Determination of glomalin-related soil protein
Glomalin-related soil protein (GRSP) was extracted from air-dried soil using the sodium citrate extraction method with minor modifications (Wright et al., 1996). Briefly, soil samples were extracted with sodium citrate buffer under autoclaving conditions, and the supernatant obtained from the first extraction was defined as easily extractable GRSP (EE-GRSP). Sequential extractions were then repeated until the supernatant became light straw-colored, and total GRSP (T-GRSP) was calculated as the sum of GRSP obtained from all extracts. Protein concentrations were determined using the Bradford method with bovine serum albumin as the standard, and absorbance was measured at 595 nm using a microplate reader (TECAN, Austria). T-GRSP was expressed as g kg−1 dry soil, and the T-GRSP:SOC ratio was calculated to evaluate the SOC-normalized abundance of T-GRSP. Detailed extraction and calculation procedures are provided in Supplementary methods.
2.5 DNA extraction, amplicon sequencing, and microbial community analysis
Soil genomic DNA was extracted from frozen soil samples using the E.Z.N.A.® Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA) according to the manufacturer’s instructions. DNA quality was checked using 1% agarose gel electrophoresis, and DNA concentration and purity were determined using a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). For bacterial community analysis, the V3–V4 region of the 16S rRNA gene was amplified using primers 341F (5′-CCTAYGGRBGCASCAG-3′) and 806R (5′-GGACTACNVGGGTWTCTAAT-3′). For fungal community analysis, the ITS1 region was amplified using primers ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′). Purified amplicons were pooled in equimolar concentrations and sequenced on an Illumina NovaSeq 6000 platform using a paired-end 250 bp strategy. Sequencing and initial bioinformatic processing were performed by Sequencing and initial bioinformatic processing were performed by Shenzhen Wekemo Bioincloud (Shenzhen, China).1
Raw paired-end reads were quality-filtered, merged, and chimera-checked before downstream analysis. High-quality sequences were clustered into operational taxonomic units (OTUs) at 97% sequence similarity. Taxonomic annotation was performed using the SILVA Release 138 database for 16S rRNA gene sequences and the UNITE 10.0 database for fungal ITS sequences. Non-target sequences, including chloroplast, mitochondrial, archaeal, and non-fungal sequences, were removed before downstream analyses, whereas unclassified fungal sequences were retained. Amplicon data processing and visualization were performed using Wekemo Bioincloud, a cloud-based platform for meta-omics data analysis and visualization (Gao et al., 2024). Bacterial and fungal alpha diversity was evaluated using the Shannon index. Principal coordinate analysis (PCoA) based on Bray–Curtis dissimilarity was used to visualize differences in bacterial and fungal community composition among N addition treatments and soil depths. The relative abundances of dominant bacterial and fungal phyla were calculated and visualized using stacked bar plots.
2.6 Random forest, correlation, co-occurrence network, and module analyses
Random forest analysis was used as an exploratory approach to identify soil and microbial variables associated with MNC, T-GRSP, and their SOC-normalized ratios. To avoid circular or overly proximate interpretation, microbial biomass variables, including MBC and MBN, were excluded from the candidate predictor set before random forest model construction. The remaining candidate predictors included soil physicochemical properties, soil stoichiometric indicators, extracellular enzyme activities, microbial diversity, and microbial community composition. To prevent model overfitting caused by the limited sample size and large number of candidate predictors, a rigorous two-step predictor screening was performed before model construction. First, only variables significantly correlated with the response variable (Spearman P < 0.05) were retained. Second, we applied Variance Inflation Factor (VIF) analysis to iteratively eliminate highly collinear predictors (VIF > 10). Model performance was robustly evaluated using Leave-One-Out Cross-Validation (LOOCV) with the caret package. Predictor importance was estimated using permutation-based importance, and statistical significance was assessed using the rfPermute package with 999 permutations. To evaluate model uncertainty, bootstrap resampling was performed to examine the stability of variable importance rankings. Therefore, the random forest results were interpreted as associative and exploratory rather than as evidence of causal effects.
Bacterial–fungal co-occurrence networks were constructed as an exploratory analysis to examine depth-dependent microbial association patterns. Because each N addition treatment had only three biological replicates, treatment-specific networks were not constructed. Instead, samples from all four N addition treatments were pooled within each soil depth to construct two depth-specific networks. Thus, each network was based on 12 samples, corresponding to four N addition treatments × three biological replicates within the same soil depth. Therefore, the network analysis was used only to explore broad depth-related microbial association patterns and was not used to infer treatment-specific network responses to N addition. We acknowledge that pooling samples across N addition treatments may include treatment-related variation and may therefore confound treatment-level ecological interpretation. Accordingly, the resulting networks were interpreted cautiously as descriptive patterns of potential microbial co-occurrence rather than as evidence of N-induced changes in microbial interactions. Before network construction, rare and low-abundance OTUs were removed to reduce spurious correlations. Only OTUs occurring in at least 50% of samples within a given soil depth and with a mean relative abundance ≥ 0.0001 were retained. If more than 400 OTUs remained after filtering, the top 400 OTUs with the highest mean relative abundance were retained for network construction. The retained OTU relative abundance matrix was log10-transformed before correlation analysis. Spearman correlations were calculated among the retained bacterial and fungal OTUs, and only strong and statistically significant correlations, defined as |Spearman’s ρ| ≥ 0.60 and FDR-adjusted P < 0.05, were retained as network edges. Network topological properties, including node number, edge number, network density, average degree, and the proportions of positive and negative edges, were calculated using the igraph package in R. Because each soil depth generated only one pooled network, these network-level properties were presented as descriptive values and were not treated as statistically replicated estimates. Network robustness was evaluated descriptively by randomly removing 50% of nodes for 200 iterations and calculating the relative size of the largest remaining connected component. These iterations represent simulation replicates rather than biological replicates. The negative/positive cohesion ratio was calculated separately for each of the 12 samples within each soil depth using the corresponding pooled network. Therefore, the violin plots represent the distribution of sample-level cohesion values rather than replicated network topologies. These cohesion values were used to visualize within-depth sample-level variation, but they were not interpreted as independent replicated networks. Because correlation-based networks do not directly demonstrate true ecological interactions, positive and negative edges were interpreted as potential co-occurrence associations rather than direct cooperative or competitive interactions.
To reduce noise, prevent overfitting, and improve the stability of module detection, stringent variable selection criteria were applied: low-abundance and rare OTUs (occurring in less than 50% of samples) were removed before WGCNA. Modules were detected using a signed network based on biweight midcorrelation, and module eigengenes were used to represent the overall abundance patterns of individual modules. To assess the statistical robustness of module-trait relationships and control for model uncertainty, Pearson correlations between module eigengenes and soil C stabilization indicators were evaluated using 999 permutation tests. Furthermore, all resulting P-values were strictly adjusted using the Benjamini-Hochberg False Discovery Rate (FDR) method to control for false positives. Therefore, the significant module-trait relationships were interpreted cautiously as robust statistical associations rather than direct causal evidence.
2.7 Statistical analyses
Statistical analyses were conducted using SPSS software version 25.0 and R software version 4.5.0. Before statistical analysis, data normality and homogeneity of variances were evaluated using the Shapiro–Wilk test and Levene’s test, respectively. Two-way analysis of variance (ANOVA) was used to test the effects of N addition, soil depth, and their interaction on soil physicochemical properties, microbial biomass, extracellular enzyme activities, microbial necromass carbon, T-GRSP, SOC-normalized ratios, and bacterial and fungal Shannon diversity. When significant effects were detected, post hoc comparisons among treatment means were performed using Fisher’s least significant difference (LSD) test at P < 0.05. Principal coordinate analysis (PCoA) based on Bray–Curtis dissimilarity was performed to visualize differences in bacterial and fungal community composition among N addition treatments and soil depths. Permutational multivariate analysis of variance (PERMANOVA) was conducted using the adonis2 function in the vegan package with 999 permutations to test the effects of N addition, soil depth, and their interaction on bacterial and fungal community composition. Homogeneity of multivariate dispersion was assessed using the betadisper function before interpreting the PERMANOVA results. Random forest analysis was conducted using the randomForest and rfPermute packages in R to evaluate the relative importance of soil and microbial predictors for MNC, T-GRSP, and their SOC-normalized ratios. Pearson correlation analysis was used to examine relationships among soil properties, microbial variables, microbial modules, and soil C stabilization indicators. Co-occurrence network analysis and WGCNA-based module analysis were performed using the igraph and WGCNA packages, respectively. Data visualization was performed using Origin 2024 and the ggplot2 package in R.
All values are presented as means ± standard errors (SE). Statistical significance was defined at P < 0.05. In figures and tables, different lowercase letters indicate significant differences among N addition treatments within the same soil depth, and ns, *, **, and *** indicate not significant, P < 0.05, P < 0.01, and P < 0.001, respectively. Given the limited number of biological replicates, random forest, co-occurrence network, and WGCNA analyses were used as exploratory tools to identify potential associations among soil properties, microbial attributes, and microbial-derived C pools. These analyses were not interpreted as direct evidence of causality.
3 Results
3.1 Effects of nitrogen addition on soil pH, MNC, T-GRSP, and their contributions to SOC
Soil pH was significantly affected by N addition, soil depth, and their interaction (Supplementary Table S3). Across both soil depths, N addition decreased soil pH, indicating soil acidification after six years of ammonium sulfate addition. In the 0–20 cm layer, soil pH decreased by 5.35, 7.92, and 17.69% under LN, MN, and HN, respectively, compared with CK. In the 20–40 cm layer, soil pH decreased by 4.69, 7.64, and 8.10% under LN, MN, and HN, respectively. The stronger decline in the surface soil suggests that N-induced acidification was more pronounced in the 0–20 cm layer. Therefore, soil pH was included in subsequent random forest and correlation analyses. The contents of bacterial necromass carbon (BNC), fungal necromass carbon (FNC), microbial necromass carbon (MNC), and total glomalin-related soil protein (T-GRSP) were significantly affected by N addition and soil depth, whereas their interaction was not significant (Figure 2). Across all N addition treatments, BNC, FNC, MNC, and T-GRSP were higher in the 0–20 cm soil layer than in the 20–40 cm soil layer. Compared with CK, low and medium N addition increased BNC, FNC, and MNC by 14–48% in the 0–20 cm layer and by 96–143% in the 20–40 cm layer. Similarly, low and medium N addition increased T-GRSP by 27–33% in the 0–20 cm layer and by 41–59% in the 20–40 cm layer. In contrast, high N addition showed weaker effects, with most microbial necromass carbon fractions remaining comparable to CK.
Figure 2
The contributions of microbial necromass carbon and T-GRSP to SOC showed more depth-dependent responses than their absolute contents (Figure 3). The BNC:SOC ratio was significantly affected by soil depth and the interaction between N addition and soil depth, but not by N addition alone. The FNC:SOC and MNC:SOC ratios were significantly affected by N addition, soil depth, and their interaction. In the 0–20 cm layer, FNC:SOC and MNC:SOC did not differ significantly among N treatments, whereas in the 20–40 cm layer, low and medium N addition increased BNC:SOC, FNC:SOC, and MNC:SOC by 46–71% relative to CK. By contrast, the T-GRSP:SOC ratio was not significantly affected by N addition, soil depth, or their interaction.
Figure 3
3.2 Responses of bacterial and fungal diversity and community composition to nitrogen addition
N addition significantly affected bacterial Shannon diversity, whereas soil depth and the N addition × soil depth interaction were not significant (Figure 4a). Compared with CK, medium N addition reduced bacterial Shannon diversity by 26.4% in the 0–20 cm layer and by 23.2% in the 20–40 cm layer. Low N addition also significantly decreased bacterial Shannon diversity in the 20–40 cm layer, while high N addition showed no significant difference from CK in either soil layer. Fungal Shannon diversity was significantly affected by N addition and the N addition × soil depth interaction, but not by soil depth alone (Figure 4b). In the 0–20 cm layer, fungal Shannon diversity was lowest under low N addition, whereas in the 20–40 cm layer, high N addition reduced fungal Shannon diversity by 40.4% compared with CK.
Figure 4
PCoA based on Bray–Curtis dissimilarity showed distinct shifts in bacterial and fungal community composition across N addition treatments and soil depths (Figure 5). PERMANOVA further confirmed that bacterial community composition was significantly affected by N addition, soil depth, and their interaction, explaining 42.46, 14.74, and 15.80% of the variation, respectively (Supplementary Table S5). For bacterial communities, the first two PCoA axes explained 29.44 and 20.07% of the total variation, respectively, and medium N addition samples were clearly separated from CK, LN, and HN samples along PCoA1 (Figure 5a). Fungal community composition was also significantly affected by N addition, soil depth, and their interaction, explaining 53.53, 7.41, and 18.32% of the variation, respectively (Supplementary Table S5). For fungal communities, PCoA1 and PCoA2 explained 25.20 and 17.24% of the total variation, respectively, and samples formed more distinct treatment-dependent clusters than bacterial communities (Figure 5b). Homogeneity tests showed no significant differences in multivariate dispersion among N addition treatments or between soil depths for either bacterial or fungal communities, supporting the reliability of the PERMANOVA results.
Figure 5
At the phylum level, bacterial communities were mainly dominated by Acidobacteriota, Pseudomonadota, and Bacillota, which accounted for 38.3, 25.3, and 15.9% of the bacterial community on average, respectively (Supplementary Figure S1). Medium N addition markedly increased the relative abundance of Bacillota, while reducing Acidobacteriota and Pseudomonadota. Fungal communities were dominated by Basidiomycota, followed by Ascomycota, and their relative abundances varied among N addition treatments and soil depths (Supplementary Figure S2).
3.3 Soil and microbial predictors associated with MNC and T-GRSP
Random forest models suggested that soil and microbial variables were associated with variation in MNC and T-GRSP after excluding microbial biomass variables from the candidate predictor set (Figure 6a). Based on the LOOCV evaluation, the explained variation of MNC was 74.6 and 88.2% in the 0–20 and 20–40 cm soil layers, respectively, whereas that of T-GRSP was 40.1% and 32.3%, respectively. The MNC:SOC ratio was well explained in the 20–40 cm layer, with an explained variation of 73.7%, while the explained variation of T-GRSP:SOC was relatively low, particularly in the 0–20 cm layer.
Figure 6
Grouped variable contribution analysis further indicated that soil nutrients and stoichiometric indicators showed the highest relative importance among the screened predictor groups associated with MNC and T-GRSP (Figures 6b,c). In contrast, the relative importance of biotic predictors became more evident in the 20–40 cm layer. For MNC in the deeper soil, enzyme activities and microbial diversity contributed more strongly than in the surface soil. Similarly, microbial diversity represented an important biological association for T-GRSP in the 20–40 cm layer. These results suggest that MNC and T-GRSP variation was primarily associated with soil resource status and stoichiometric balance, with additional contributions from enzyme activities, microbial diversity, and community composition, particularly in deeper soil.
3.4 Depth-dependent bacterial–fungal co-occurrence patterns and microbial modules associated with soil carbon stabilization
The exploratory depth-specific bacterial–fungal co-occurrence networks suggested different potential microbial association patterns between the 0–20 and 20–40 cm soil layers after applying the stringent 50% prevalence filtering and false discovery rate (FDR) correction (Figure 7). Because samples from all N addition treatments were pooled within each soil depth, these networks were used only to describe broad depth-related co-occurrence patterns and were not interpreted as treatment-specific responses to N addition. The 0–20 cm network contained 395 nodes and 5,893 edges, whereas the 20–40 cm network contained 378 nodes and 4,351 edges. Compared with the surface-soil pooled network, the deeper-soil pooled network showed lower network complexity, as indicated by fewer nodes and edges. Positive co-occurrence correlations dominated both networks, but their proportion decreased from 86.0% in the 0–20 cm layer to 80.2% in the 20–40 cm layer, whereas negative co-occurrence correlations increased from 14.0% to 19.8%. Edge summary analysis further showed that bacterial intra-kingdom associations were the dominant link type in both soil layers. In the 0–20 cm network, bacterial–bacterial associations comprised 4,311 positive and 211 negative links, fungal–fungal associations comprised 64 positive and 59 negative links, and bacterial–fungal associations comprised 691 positive and 557 negative links. In the 20–40 cm network, bacterial–bacterial associations comprised 2,718 positive and 391 negative links, fungal–fungal associations comprised 116 positive and 54 negative links, and bacterial–fungal associations comprised 654 positive and 418 negative links (Figures 7b,d). Network robustness and cohesion analyses were used only as descriptive indicators (Figures 7e,f). The 0–20 cm network showed slightly higher robustness against random node removal than the 20–40 cm network. The negative/positive cohesion ratio was summarized separately for the 12 samples within each soil depth based on the corresponding pooled network. Therefore, the cohesion distributions should not be interpreted as replicated network topologies but rather as variation among individual samples within each pooled depth-specific network. Because treatment-specific networks were not constructed, these robustness and cohesion results were interpreted descriptively rather than as comparisons among independently replicated networks.
Figure 7
Following stringent permutation tests and FDR correction, key microbial modules demonstrated highly robust associations with soil carbon fractions, microbial necromass, and GRSP-related indicators (Figure 8a). Among the bacterial modules, BM1 was strongly and negatively correlated with all measured SOC fractions, microbial necromass C, and GRSP-related indicators. Conversely, BM8 showed robust positive correlations across almost all carbon stabilization indicators, and BM3 exhibited specifically strong positive associations with particulate organic carbon (POC), BNC, FNC, and MNC. Among the fungal modules, FM5 was negatively correlated with these carbon indicators, while FM1 showed widespread positive correlations (Figure 8a). The standardized relative abundances of these key modules shifted significantly in response to N addition, soil depth, and their interactions (Figure 8b). These results indicate that depth-dependent differences in bacterial–fungal co-occurrence patterns were accompanied by shifts in specific, statistically robust microbial assemblages associated with soil carbon stabilization.
Figure 8
4 Discussion
4.1 Nitrogen addition regulated microbial necromass carbon through changes in microbial growth, turnover, and nutrient constraints
Low and medium N addition increased bacterial necromass carbon, fungal necromass carbon, and total microbial necromass carbon, whereas these positive effects were weakened under high N addition (Figure 2). This pattern suggests that microbial necromass accumulation in this alpine coniferous forest was sensitive to the level of N enrichment. Microbial necromass C is determined by the balance between microbial biomass production, microbial turnover, residue decomposition, and residue stabilization. Therefore, the increase in MNC under LN and MN may reflect enhanced microbial growth and turnover under improved nutrient and substrate availability. In our study, LN and MN were accompanied by higher SOC, TN, DOC, DON, MBC, and MBN, especially in the surface soil (Supplementary Tables S1, S2). These changes may have alleviated microbial N limitation and increased the availability of labile C and N substrates, thereby supporting microbial biomass production and subsequent microbial residue formation (Kuzyakov and Xu, 2013; Treseder, 2008; Liang et al., 2017). Because both bacterial and fungal necromass C increased under LN and MN, moderate N enrichment may have promoted residue accumulation from both bacterial and fungal sources rather than shifting microbial necromass formation toward a single microbial group.
The weaker positive response of microbial necromass C under HN may be explained by the shift from nutrient alleviation to nutrient and physiological constraints. High ammonium-sulfate addition was accompanied by lower AP, altered SOC:TP and TN:TP ratios, and stronger soil acidification, especially in the 0–20 cm layer (Supplementary Tables S1, S2). Soil acidification can constrain microbial growth, alter community composition, and reduce enzyme functioning by increasing physiological stress and modifying nutrient availability (Sinsabaugh and Follstad Shah, 2012; Rousk et al., 2010). At the same time, reduced P availability and altered stoichiometric balance may increase microbial nutrient imbalance, forcing microorganisms to allocate more energy to nutrient acquisition rather than biomass production and residue formation (Peñuelas et al., 2013). Under such conditions, the production of microbial residues may not continue to increase despite greater N input. In addition, acidification and nutrient imbalance may affect the decomposition and stabilization of existing microbial residues, further weakening the accumulation of MNC. Therefore, the weaker MNC response under HN was likely associated with combined constraints from acidification, reduced P availability, and disrupted soil C:N:P balance rather than with N availability alone.
Although this section focuses primarily on microbial necromass C, T-GRSP showed a broadly similar response pattern, with higher values under LN and MN but weaker responses under HN. This may indicate that moderate N enrichment also favored GRSP-related C pools, possibly through changes in mycorrhizal-associated activity, proteinaceous compound preservation, or aggregate-related protection (Zhang J. et al., 2023; Sun et al., 2018). However, our data cannot distinguish whether the increase in T-GRSP resulted from enhanced mycorrhizal production, slower GRSP turnover, or stronger physicochemical protection. Therefore, the T-GRSP response should be interpreted as a treatment-related change in an operationally defined GRSP pool rather than direct evidence for a specific mycorrhizal mechanism.
4.2 Depth-dependent SOC-normalized responses of microbial-derived carbon
The depth-dependent patterns of SOC-normalized microbial necromass partly supported our expectation that soil depth would modulate the relative contribution of microbial-derived C to SOC. Although the absolute contents of BNC, FNC, MNC, and T-GRSP were consistently higher in the surface soil (Figure 2), their SOC-normalized proportions showed more pronounced depth-dependent responses (Figure 3). In particular, low and medium N addition significantly increased BNC:SOC, FNC:SOC, and MNC:SOC in the 20–40 cm layer (Figures 3a–c), whereas FNC:SOC and MNC:SOC remained relatively stable among N treatments in the 0–20 cm layer. This discrepancy indicates that changes in microbial-derived C contents do not necessarily translate proportionally into changes in their SOC-normalized contributions, especially across soil depths (Ni et al., 2020; Hicks Pries et al., 2023).
The weaker response of MNC:SOC in the surface soil may be partly related to the larger SOC pool and stronger plant-derived C inputs in the 0–20 cm layer (Sokol and Bradford, 2019; Villarino et al., 2021). Under such conditions, increases in microbial necromass C may be partly diluted by the larger background SOC pool, resulting in relatively stable SOC-normalized ratios (Liang et al., 2019; Wang et al., 2021). In contrast, the 20–40 cm layer had lower SOC concentrations and lower microbial biomass (Supplementary Table S2), which may make SOC-normalized microbial necromass more sensitive to N-induced changes. Therefore, the pronounced increase in MNC:SOC under low and medium N addition indicates that microbial necromass made a larger proportional contribution to SOC in the deeper soil under moderate N enrichment. However, microbial turnover and residue stabilization processes were not directly measured; thus, this result should not be interpreted as direct evidence for a specific SOC accumulation pathway. The response of T-GRSP:SOC differed from that of MNC:SOC. Unlike MNC:SOC, T-GRSP:SOC was not significantly affected by N addition, soil depth, or their interaction, despite clear changes in absolute T-GRSP content (Figure 3d). This relatively conservative pattern is consistent with previous studies suggesting that GRSP-associated C can represent a relatively stable microbial-related C pool, which may respond less sensitively to N addition and soil depth than microbial necromass C (Zhang et al., 2017; Ma et al., 2025). However, GRSP turnover, mycorrhizal production, and physicochemical protection were not directly assessed in this study.
4.3 Soil stoichiometry and microbial attributes were associated with MNC and T-GRSP variation
In line with our third hypothesis, random forest analysis identified soil nutrient and stoichiometric variables as the main predictor group associated with MNC and T-GRSP variation (Figure 6). TN, DOC, DON, SOC:TP, and TN:TP were closely associated with these microbial-derived C pools (Figure 6a), suggesting that variation in MNC and T-GRSP was more closely related to the overall soil resource environment and stoichiometric balance than to N input alone. These variables represent different but connected aspects of the soil resource environment. TN and DON reflect N supply, DOC represents labile C availability, whereas SOC:TP and TN:TP indicate the balance between C, N, and P resources. Together, these factors may regulate microbial C allocation, extracellular enzyme production, and residue formation, thereby influencing MNC and T-GRSP accumulation. This interpretation is also consistent with the observed changes in soil properties: LN and MN generally increased SOC, TN, DOC, DON, MBC, and MBN, whereas HN was accompanied by lower AP, stronger soil acidification, and weaker increases in microbial biomass (Supplementary Tables S1, S2). Therefore, the weaker positive response of HN on MNC and T-GRSP may be associated with nutrient imbalance and acidification-related changes rather than with increased N availability alone (Averill and Waring, 2018; Tian and Niu, 2015; Luo et al., 2025). However, these results should be interpreted cautiously. The importance of SOC:TP, TN:TP, and P-acquiring enzyme activity provides only indirect evidence for possible relative P limitation, because microbial C:N:P stoichiometry and ecoenzymatic vector analysis were not directly measured. Similarly, although soil pH was closely associated with microbial-derived C pools, this study did not experimentally separate the effects of ammonium input, sulfate input, and soil acidification. Thus, pH and soil stoichiometry should be interpreted as associated factors rather than direct causal drivers.
The relative contribution of biotic predictors increased in the 20–40 cm soil layer (Figures 6b,c). In particular, enzyme activities, microbial diversity, and community composition contributed more strongly to MNC or T-GRSP variation in deeper soil than in surface soil. This pattern may occur because deeper soil generally has lower substrate availability, lower microbial biomass, and stronger environmental constraints than surface soil (Fierer et al., 2003; Rumpel and Kögel-Knabner, 2011). Under such conditions, microbial-derived C accumulation may depend more strongly on the capacity of microbial communities to acquire limiting resources and process organic substrates. Extracellular enzyme activities can reflect microbial functional activity and resource acquisition demand (Burns et al., 2013; Sinsabaugh and Follstad Shah, 2012), and may therefore be closely associated with microbial residue formation and GRSP-related C pools in deeper soil. However, enzyme activity itself is also partly controlled by microbial biomass and community composition. Thus, the stronger contribution of enzyme activities should not be interpreted as an independent causal effect, but rather as an indicator of microbial functional potential and resource acquisition processes. From the perspective of microbial resource acquisition strategies, enzyme activity ratios or ecoenzymatic vector analysis may provide more direct information about relative C, N, and P limitation than individual or grouped enzyme activities (Moorhead et al., 2016). In the present study, we used grouped C-, N-, and P-acquiring enzyme activities as exploratory predictors, whereas enzyme activity ratios and vector-based indicators were not included in the random forest models because of the limited sample size and potential collinearity with individual enzyme variables. Therefore, the associations between enzyme activities and MNC or T-GRSP should be interpreted cautiously. Future studies combining enzyme activity ratios, ecoenzymatic vector analysis, microbial biomass normalization, and microbial C:N:P stoichiometry would be useful for testing whether shifts in microbial resource acquisition strategies directly regulate microbial necromass and GRSP accumulation in deeper soils.
Similarly, the stronger associations of microbial diversity and community composition with T-GRSP in deeper soil may reflect the greater importance of microbial community structure under resource-limited conditions. Because T-GRSP is an operationally defined pool that may be influenced by mycorrhizal-associated production, microbial community composition, and physicochemical protection (Driver et al., 2005; Rillig, 2004), changes in fungal or bacterial community attributes could be linked to its variation. Nevertheless, these results should be regarded as exploratory associations rather than direct evidence that microbial diversity, community composition, or enzyme activities caused changes in MNC or T-GRSP. Although path analysis, structural equation modeling, or variance partitioning could provide additional insights into the direct and indirect pathways linking N addition, soil stoichiometry, microbial attributes, and microbial-derived C pools, these approaches were not applied in the present study because of the limited sample size and potential collinearity among the measured variables. Future studies with larger sample sizes and independent validation datasets could use SEM or variance partitioning to further test the causal pathways suggested by our exploratory analyses.
4.4 Microbial co-occurrence patterns and modules provide exploratory links to soil C indicators
Beyond changes in microbial diversity and community composition, our exploratory co-occurrence and module analyses suggested that potential bacterial–fungal association patterns differed between soil depths, and that several microbial modules were statistically associated with soil C fractions, microbial necromass C, and GRSP-related indicators (Figures 7, 8). The 0–20 cm pooled network had more nodes, more edges, and slightly higher descriptive robustness than the 20–40 cm pooled network, whereas the deeper-soil pooled network showed fewer co-occurrence correlations and a higher proportion of negative correlations. These depth-related patterns may reflect differences in substrate availability, microbial biomass, and habitat conditions between surface and deeper soils, which is consistent with previous studies showing that soil depth strongly shapes microbial community composition, function, and co-occurrence patterns (Upton et al., 2020; Naylor et al., 2022; Jiang et al., 2025). However, because these networks were constructed from compositional amplicon data, pooled within each soil depth, and based on pairwise correlations, they should not be interpreted as direct microbial interactions or as evidence of cooperation or competition among taxa (Blanchet et al., 2020; Hirano and Takemoto, 2019; Röttjers and Faust, 2018). Positive and negative edges may arise not only from biotic interactions, but also from shared environmental preferences, niche differentiation, indirect associations, or unmeasured environmental variables. Therefore, we interpreted these networks only as descriptive patterns of potential co-occurrence and microbial community organization rather than as mechanistic evidence of microbial interactions. This cautious interpretation is particularly important because each soil depth generated one pooled network rather than independently replicated networks.
Module analysis further showed that specific bacterial and fungal modules were statistically associated with SOC fractions, microbial necromass C, and GRSP-related indicators (Figure 8). Positive modules associated with MNC, T-GRSP, and SOC fractions may reflect coordinated microbial assemblages that share similar resource preferences or ecological niches, whereas negatively associated modules may reflect taxa favored under conditions less conducive to microbial-derived C accumulation. However, without functional gene or isotope evidence, these module-level patterns should be interpreted as indicators of microbial community organization rather than direct mechanistic evidence. For example, BM8, BM3, and FM1 were positively associated with most C-related indicators, whereas BM1 and FM5 showed negative associations. These results are consistent with previous studies showing that microbial modules or network-based microbial assemblages can be linked to soil properties and ecosystem functions, suggesting that microbial community organization at the module level may provide useful indicators of microbial-derived C accumulation (De Menezes et al., 2015; Wagg et al., 2019; Kallenbach et al., 2016). This module-level pattern also helps explain why Shannon diversity alone may not fully capture the microbial associations related to MNC and T-GRSP, because diversity indices describe overall richness and evenness but do not identify specific microbial assemblages associated with C-related indicators.
Together, these findings indicate that microbial co-occurrence patterns and module eigengenes were statistically associated with microbial-derived C pools and SOC fractions under different N addition treatments and soil depths. However, these relationships should be regarded as exploratory associations rather than evidence that microbial networks or modules caused changes in MNC, T-GRSP, or SOC stabilization. The functional roles of these modules remain uncertain because metagenomic, transcriptomic, isotope-tracing, and microbial manipulation evidence was not available. In addition, the co-occurrence networks were constrained by the limited number of biological replicates and were constructed as depth-specific pooled networks rather than independently replicated treatment-specific networks. Therefore, future studies with larger sample sizes, replicated networks, functional gene analysis, stable isotope tracing, and controlled microbial or nutrient-manipulation experiments are needed to test whether and how specific microbial assemblages contribute to microbial necromass formation and GRSP accumulation.
4.5 Ecological relevance and limitations of the N addition approach
Although our results provide evidence that repeated ammonium-sulfate N enrichment was associated with changes in microbial-derived C pools, several methodological and mechanistic limitations should be considered when interpreting these findings. The ecological realism and potential confounding effects of the N addition approach should be carefully considered. In this study, N was added as ammonium sulfate once each year in mid-August. This design was intended to simulate cumulative annual N enrichment during the growing season, because anthropogenic reactive N inputs have substantially increased atmospheric N deposition at regional and global scales (Galloway et al., 2008; Liu et al., 2013). The selected N addition rates represented low, moderate, and relatively high N-enrichment scenarios rather than conventional fertilization. This gradient was also ecologically relevant for evaluating potential future N-enrichment effects in southeastern Xizang forest ecosystems, where current atmospheric Nr deposition remains relatively low but is dominated by reduced N and may be sensitive to additional N inputs (Wang et al., 2020). However, our treatment cannot fully reproduce the continuous and episodic nature of atmospheric wet and dry N deposition. Conventional understory or forest-floor N addition experiments have been widely used to assess forest responses to elevated N deposition, but they may not completely simulate natural atmospheric deposition processes, especially canopy-associated deposition and retention (Zhang W. et al., 2015). A single annual application may also generate a transient ammonium pulse, which could temporarily alter microbial N availability, nitrification, enzyme activities, and microbial community responses, and such responses may depend on both the form and level of added N (Li et al., 2021). Therefore, our results should be interpreted as ecosystem responses to repeated annual ammonium-sulfate N enrichment rather than as an exact simulation of continuous atmospheric N deposition.
In addition, ammonium sulfate addition may introduce potential confounding effects associated with sulfate input and soil acidification. The observed decline in soil pH, especially in the surface soil, suggests that acidification accompanied the N addition treatment. Nitrogen addition has been shown to significantly reduce soil pH at the global scale, and ammonium-based inputs may contribute to acidification through nitrification and associated proton production (Tian and Niu, 2015; Powlson and Dawson, 2022). Therefore, the weaker positive response under high N addition may not be attributed to increased N availability alone, but may also reflect ammonium-induced acidification, sulfate-associated changes, and nutrient imbalance, particularly reduced P availability and altered C:N:P stoichiometry. Previous studies have also shown that combined N and S inputs can alter soil microbial functions and extracellular enzyme activities, supporting the need to consider sulfate-related effects when ammonium sulfate is used as the N source (Hu et al., 2013). Because this experiment did not include separate nitrate, sulfate-only, or liming treatments, we cannot fully disentangle the effects of N form, sulfate input, and acidification. Future studies using multiple N forms, split applications, sulfate controls, and acidification-buffering treatments would help clarify the relative contribution of these mechanisms. In addition, microbial life-history traits, such as copiotrophic–oligotrophic or r/K-related strategies, were not directly assessed in this study. Future studies combining amplicon sequencing with metagenomic or trait-based approaches would provide deeper insight into how microbial ecological strategies regulate microbial necromass formation and GRSP accumulation under N enrichment. Nevertheless, by measuring soil pH, nutrient availability, microbial biomass, enzyme activities, and microbial community attributes, our study provides evidence that changes in microbial-derived C pools were accompanied by shifts in nutrient balance, soil acidification, and microbial community attributes under repeated ammonium-sulfate N enrichment.
5 Conclusion
This study showed that six-year repeated ammonium-sulfate N enrichment was associated with changes in microbial-derived C accumulation and its SOC-normalized contribution in a depth-dependent manner in an alpine coniferous forest. Low and medium N addition increased bacterial necromass C, fungal necromass C, total microbial necromass C, and T-GRSP, whereas these positive effects were weakened under high N addition. Although microbial-derived C contents were higher in surface soil, the SOC-normalized contribution of microbial necromass responded more strongly to N addition in the 20–40 cm layer, indicating that subsoil microbial residues may be an important component of SOC responses to moderate N enrichment. Random forest analysis further suggested that soil nutrient availability and C:N:P stoichiometry were closely associated with MNC and T-GRSP, while enzyme activities, microbial diversity, community composition, and key microbial modules showed additional associations. Moreover, depth-dependent bacterial–fungal co-occurrence patterns and carbon-associated modules suggest that microbial community organization may provide useful indicators of microbial-derived C accumulation and SOC fractions under N enrichment. Overall, moderate ammonium-sulfate N enrichment was associated with greater microbial-derived C accumulation, whereas high N addition showed weaker positive effects that were accompanied by nutrient imbalance, soil acidification, and changes in microbial community attributes. Because N was applied once annually as ammonium sulfate, these findings should be interpreted as ecosystem responses to repeated annual ammonium-sulfate N enrichment rather than as an exact simulation of continuous atmospheric N deposition.
Statements
Data availability statement
The raw sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive (SRA) under the BioProject accession number PRJNA1489377.
Author contributions
SZ: Writing – original draft. YH: Writing – review & editing. ZC: Conceptualization, Writing – review & editing. ZW: Conceptualization, Writing – review & editing. MH: Investigation, Writing – review & editing. YY: Funding acquisition, Project administration, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by National Natural Science Foundation of China (31860141 and 31360119); Xizang College of Agriculture and Animal Husbandry Graduate Education Innovation Program Project (YJS2024-31, YJS2024-28, and YJS2024-26); Doctoral Program in Forestry at Xizang Agricultural and Animal Husbandry University (Phase I) (533325001); Xizang College of Agriculture and Animal Husbandry Key Discipline Construction Project (XK2025-04); The Seventh Batch of Flexible Talent Project of Xizang College of Agriculture, Animal Husbandry and Herding (53013001804); National Student Innovation and Entrepreneurship Training Program (2024-02); and the 2024 and 2025 Special Funds for Central Financial Support for the Development and Reform of Local Universities: Construction of Small Agricultural and Animal Husbandry Science and Technology Institutes with Highland Characteristics and Enhancement of Comprehensive Service Capability (XK2024-04, XK2024-01, and YJSXK2025-22).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/ffgc.2026.1905473/full#supplementary-material
Footnotes
References
1
AoD.WangB.WangY.ChenY.LiangC.AnS. (2025). Arbuscular mycorrhizal fungi communities and glomalin mediate particulate and mineral-associated organic carbon formation in grassland patches. Commun. Earth Environ.6:553. doi: 10.1038/s43247-025-02492-x
2
AppuhnA.JoergensenR. G. (2006). Microbial colonisation of roots as a function of plant species. Soil Biol. Biochem.38, 1040–1051. doi: 10.1016/j.soilbio.2005.09.002
3
AverillC.WaringB. (2018). Nitrogen limitation of decomposition and decay: how can it occur?. Glob. Chang. Biol.24, 1417–1427. doi: 10.1111/gcb.13980,
4
BanerjeeS.ZhaoC.KirkbyC. A.CogginsS.ZhaoS.BissettA.et al. (2021). Microbial interkingdom associations across soil depths reveal network connectivity and keystone taxa linked to soil fine-fraction carbon content. Agr Ecosyst Environ320:107559. doi: 10.1016/j.agee.2021.107559
5
BaoS.QinH.LaoJ. (1988). Soil Agrochemical Analysis. Beijing, China: Nong ye chu ban she.
6
BlanchetF. G.CazellesK.GravelD. (2020). Co-occurrence is not evidence of ecological interactions. Ecol. Lett.23, 1050–1063. doi: 10.1111/ele.13525,
7
BradfordM. A.WiederW. R.BonanG. B.FiererN.RaymondP. A.CrowtherT. W. (2016). Managing uncertainty in soil carbon feedbacks to climate change. Nat. Clim. Chang.6, 751–758. doi: 10.1038/nclimate3071
8
BrookesP.PowlsonD.JenkinsonD. (1982). Measurement of microbial biomass phosphorus in soil. Soil Biol. Biochem.14, 319–329. doi: 10.1016/0038-0717(82)90001-3
9
BuckeridgeK. M.CreamerC.WhitakerJ. (2022). Deconstructing the microbial necromass continuum to inform soil carbon sequestration. Funct. Ecol.36, 1396–1410. doi: 10.1111/1365-2435.14014
10
BurnsR. G.DeforestJ. L.MarxsenJ.SinsabaughR. L.StrombergerM. E.WallensteinM. D.et al. (2013). Soil enzymes in a changing environment: current knowledge and future directions. Soil Biol. Biochem.58, 216–234. doi: 10.1016/j.soilbio.2012.11.009
11
De MenezesA. B.Prendergast-MillerM. T.RichardsonA. E.ToscasP.FarrellM.MacdonaldL. M.et al. (2015). Network analysis reveals that bacteria and fungi form modules that correlate independently with soil parameters. Environ. Microbiol.17, 2677–2689. doi: 10.1111/1462-2920.12559
12
DriverJ. D.HolbenW. E.RilligM. C. (2005). Characterization of glomalin as a hyphal wall component of arbuscular mycorrhizal fungi. Soil Biol. Biochem.37, 101–106. doi: 10.1016/j.soilbio.2004.06.011
13
FiererN.SchimelJ. P.HoldenP. A. (2003). Variations in microbial community composition through two soil depth profiles. Soil Biol. Biochem.35, 167–176. doi: 10.1016/S0038-0717(02)00251-1
14
FowlerD.CoyleM.SkibaU.SuttonM. A.CapeJ. N.ReisS.et al. (2013). The global nitrogen cycle in the twenty-first century. Philos. Trans. R. Soc., B368, 20130164. doi: 10.1098/rstb.2013.0164
15
GallowayJ. N.TownsendA. R.ErismanJ. W.BekundaM.CaiZ.FreneyJ. R.et al. (2008). Transformation of the nitrogen cycle: recent trends, questions, and potential solutions. Science320, 889–892. doi: 10.1126/science.1136674,
16
GaoY.ZhangG.JiangS.LiuY.‐X. (2024). Wekemo Bioincloud: A user-friendly platform for meta-omics data analyses. iMeta3:e175. doi: 10.1002/imt2.175,
17
Hicks PriesC. E.RyalsR.ZhuB.MinK.CooperA.GoldsmithS.et al. (2023). The deep soil organic carbon response to global change. Annu. Rev. Ecol. Evol. Syst.54, 375–401. doi: 10.1146/annurev-ecolsys-102320-085332
18
HiranoH.TakemotoK. (2019). Difficulty in inferring microbial community structure based on co-occurrence network approaches. BMC Bioinf.20:329. doi: 10.1186/s12859-019-2915-1,
19
HuJ.DuM.ChenJ.TieL.ZhouS.BuckeridgeK. M.et al. (2023). Microbial necromass under global change and implications for soil organic matter. Glob. Chang. Biol.29, 3503–3515. doi: 10.1111/gcb.16676,
20
HuJ.HuangC.ZhouS.LiuX.DijkstraF. A. (2022). Nitrogen addition increases microbial necromass in croplands and bacterial necromass in forests: a global meta-analysis. Soil Biol. Biochem.165:108500. doi: 10.1016/j.soilbio.2021.108500
21
HuY.-L.JungK.ZengD.-H.ChangS. X. (2013). Nitrogen-and sulfur-deposition-altered soil microbial community functions and enzyme activities in a boreal mixedwood forest in western Canada. Can. J. For. Res.43, 777–784. doi: 10.1139/cjfr-2013-0049
22
HuH.QianC.XueK.JörgensenR. G.KeiluweitM.LiangC.et al. (2024). Reducing the uncertainty in estimating soil microbial-derived carbon storage. Proc. Natl. Acad. Sci.121:e2401916121. doi: 10.1073/pnas.2401916121
23
HuangB.YanG.LiuG.SunX.WangX.XingY.et al. (2022). Effects of long-term nitrogen addition and precipitation reduction on glomalin-related soil protein and soil aggregate stability in a temperate forest. Catena214:106284. doi: 10.1016/j.catena.2022.106284
24
JiangP.WanX.CheM.LiL.LiuM. (2025). Soil depth exerts greater effect on bacterial community than spatial structure in Longmenshan fault zone. Appl. Environ. Microbiol.91, e01161–e01124. doi: 10.1128/aem.01161-24
25
JingX.YangX.RenF.ZhouH.ZhuB.HeJ.-S. (2016). Neutral effect of nitrogen addition and negative effect of phosphorus addition on topsoil extracellular enzymatic activities in an alpine grassland ecosystem. Appl. Soil Ecol.107, 205–213. doi: 10.1016/j.apsoil.2016.06.004
26
JoergensenR. G. (2018). Amino sugars as specific indices for fungal and bacterial residues in soil. Biol. Fertil. Soils54, 559–568. doi: 10.1007/s00374-018-1288-3
27
KallenbachC. M.FreyS. D.GrandyA. S. (2016). Direct evidence for microbial-derived soil organic matter formation and its ecophysiological controls. Nat. Commun.7:13630. doi: 10.1038/ncomms13630,
28
KuzyakovY.XuX. (2013). Competition between roots and microorganisms for nitrogen: mechanisms and ecological relevance. New Phytol.198, 656–669. doi: 10.1111/nph.12235,
29
LiY.WangC.GaoS.WangP.QiuJ.ShangS. (2021). Impacts of simulated nitrogen deposition on soil enzyme activity in a northern temperate forest ecosystem depend on the form and level of added nitrogen. Eur. J. Soil Biol.103:103287. doi: 10.1016/j.ejsobi.2021.103287
30
LiangC.AmelungW.LehmannJ.KästnerM. (2019). Quantitative assessment of microbial necromass contribution to soil organic matter. Glob. Chang. Biol.25, 3578–3590. doi: 10.1111/gcb.14781,
31
LiangC.SchimelJ. P.JastrowJ. D. (2017). The importance of anabolism in microbial control over soil carbon storage. Nat. Microbiol.2: 17105. doi: 10.1038/nmicrobiol.2017.105,
32
LiaoC.MenX.WangC.ChenR.ChengX. (2022). Nitrogen availability and mineral particles contributed fungal necromass to the newly formed stable carbon pool in the alpine areas of Southwest China. Soil Biol. Biochem.173:108788. doi: 10.1016/j.soilbio.2022.108788
33
LiuH.GaoX.RenT.ChenH. Y. H.ZouX.SunY.et al. (2025). Nitrogen addition induces microbial phosphorus limitations in bulk soil but not in rhizospheric soil: a global analysis. Soil Tillage Res.252:106609. doi: 10.1016/j.still.2025.106609
34
LiuX.ZhangY.HanW.TangA.ShenJ.CuiZ.et al. (2013). Enhanced nitrogen deposition over China. Nature494, 459–462. doi: 10.1038/nature11917,
35
LuH.ChengZ.YaoZ.XueA. (2024). Impacts of pilot carbon emission trading policies on urban environmental pollution: evidence from China. J. Environ. Manag.359:121016. doi: 10.1016/j.jenvman.2024.121016,
36
LuoR.KuzyakovY.ChenJ.QiangW.ZhangY.PangX. (2025). Phosphorus enrichment mediates the responses of plant lignin and microbial necromass accumulation to nitrogen addition in subalpine forest soil. Geoderma458:117317. doi: 10.1016/j.geoderma.2025.117317
37
LuoM.MoorheadD. L.Ochoa-HuesoR.MuellerC. W.YingS. C.ChenJ. (2022). Nitrogen loading enhances phosphorus limitation in terrestrial ecosystems with implications for soil carbon cycling. Funct. Ecol.36, 2845–2858. doi: 10.1111/1365-2435.14178
38
MaY.ChengH.NingQ.SunT. (2025). Mowing enhances the contribution of glomalin over amino sugars to soil organic carbon pool following nitrogen cessation in temperate grasslands. Plant Soil514, 1473–1488. doi: 10.1007/s11104-025-07470-y
39
MaX.WangT.ShiZ.ChiarielloN. R.DochertyK.FieldC. B.et al. (2022). Long-term nitrogen deposition enhances microbial capacities in soil carbon stabilization but reduces network complexity. Microbiome10:112. doi: 10.1186/s40168-022-01309-9
40
MoorheadD. L.SinsabaughR. L.HillB. H.WeintraubM. N. (2016). Vector analysis of ecoenzyme activities reveal constraints on coupled C, N and P dynamics. Soil Biol. Biochem.93, 1–7. doi: 10.1016/j.soilbio.2015.10.019
41
NaylorD.McclureR.JanssonJ. (2022). Trends in microbial community composition and function by soil depth. Microorganisms10:540. doi: 10.3390/microorganisms10030540,
42
NiX.LiaoS.TanS.WangD.PengY.YueK.et al. (2020). A quantitative assessment of amino sugars in soil profiles. Soil Biol. Biochem.143:107762. doi: 10.1016/j.soilbio.2020.107762
43
PanY.BirdseyR. A.FangJ.HoughtonR.KauppiP. E.KurzW. A.et al. (2011). A large and persistent carbon sink in the world’s forests. Science333, 988–993. doi: 10.1126/science.1201609,
44
PeñuelasJ.PoulterB.SardansJ.CiaisP.van der VeldeM.BoppL.et al. (2013). Human-induced nitrogen–phosphorus imbalances alter natural and managed ecosystems across the globe. Nat. Commun.4:2934. doi: 10.1038/ncomms3934
45
PowlsonD. S.DawsonC. J. (2022). Use of ammonium sulphate as a sulphur fertilizer: implications for ammonia volatilization. Soil Use Manag.38, 622–634. doi: 10.1111/sum.12733,
46
RilligM. C. (2004). Arbuscular mycorrhizae, glomalin, and soil aggregation. Can. J. Soil Sci.84, 355–363. doi: 10.4141/S04-003
47
RöttjersL.FaustK. (2018). From hairballs to hypotheses–biological insights from microbial networks. FEMS Microbiol. Rev.42, 761–780. doi: 10.1093/femsre/fuy030,
48
RouskJ.BååthE.BrookesP. C.LauberC. L.LozuponeC.CaporasoJ. G.et al. (2010). Soil bacterial and fungal communities across a pH gradient in an arable soil. ISME J.4, 1340–1351. doi: 10.1038/ismej.2010.58,
49
RumpelC.Kögel-KnabnerI. (2011). Deep soil organic matter—a key but poorly understood component of terrestrial C cycle. Plant Soil338, 143–158. doi: 10.1007/s11104-010-0391-5
50
SinsabaughR. L.Follstad ShahJ. J. (2012). Ecoenzymatic stoichiometry and ecological theory. Annu. Rev. Ecol. Evol. Syst.43, 313–343. doi: 10.1146/annurev-ecolsys-071112-124414
51
SokolN. W.BradfordM. A. (2019). Microbial formation of stable soil carbon is more efficient from belowground than aboveground input. Nat. Geosci.12, 46–53. doi: 10.1038/s41561-018-0258-6
52
SunL.JingH.WangG.LiuG. (2018). Nitrogen addition increases the contents of glomalin-related soil protein and soil organic carbon but retains aggregate stability in a Pinus tabulaeformis forest. PeerJ6:e5039. doi: 10.7717/peerj.5039,
53
TakeleL.YangS.ChenZ.YuanJ.DingW. (2025). Contribution of microbial necromass to soil organic carbon in profile depths exhibited opposite patterns across ecosystems: a global meta-analysis. Soil Biol. Biochem.207:109842. doi: 10.1016/j.soilbio.2025.109842
54
TangB.RocciK. S.LehmannA.RilligM. C. (2023). Nitrogen increases soil organic carbon accrual and alters its functionality. Glob. Chang. Biol.29, 1971–1983. doi: 10.1111/gcb.16588,
55
TianD.NiuS. (2015). A global analysis of soil acidification caused by nitrogen addition. Environ. Res. Lett.10:024019. doi: 10.1088/1748-9326/10/2/024019
56
TresederK. K. (2008). Nitrogen additions and microbial biomass: a meta-analysis of ecosystem studies. Ecol. Lett.11, 1111–1120. doi: 10.1111/j.1461-0248.2008.01230.x,
57
UptonR. N.Checinska SielaffA.HofmockelK. S.XuX.PolleyH. W.WilseyB. J. (2020). Soil depth and grassland origin cooperatively shape microbial community co-occurrence and function. Ecosphere11:e02973. doi: 10.1002/ecs2.2973
58
VanceE. D.BrookesP. C.JenkinsonD. S. (1987). An extraction method for measuring soil microbial biomass C. Soil Biol. Biochem.19, 703–707. doi: 10.1016/0038-0717(87)90052-6
59
VillarinoS. H.PintoP.JacksonR. B.PiñeiroG. (2021). Plant rhizodeposition: a key factor for soil organic matter formation in stable fractions. Sci. Adv.7:eabd3176. doi: 10.1126/sciadv.abd3176,
60
WaggC.SchlaeppiK.BanerjeeS.KuramaeE. E.van der HeijdenM. G. A. (2019). Fungal-bacterial diversity and microbiome complexity predict ecosystem functioning. Nat. Commun.10:4841. doi: 10.1038/s41467-019-12798-y,
61
WangB.AnS.LiangC.LiuY.KuzyakovY. (2021). Microbial necromass as the source of soil organic carbon in global ecosystems. Soil Biol. Biochem.162:108422. doi: 10.1016/j.soilbio.2021.108422
62
WangW.FengY.WuR.WangX.HeX.ZhangM.et al. (2023). Long-term N addition reduced the diversity of arbuscular mycorrhizal fungi and understory herbs of a Korean pine plantation in northern China. Front. Ecol. Evol.11:1192267. doi: 10.3389/fevo.2023.1192267
63
WangW.GuanL.WenZ.MaX.FangJ.LiuX. (2020). Atmospheric nitrogen deposition to a southeast Tibetan forest ecosystem. Atmosphere11:1331. doi: 10.3390/atmos11121331
64
WangC.KuzyakovY. (2024). Mechanisms and implications of bacterial–fungal competition for soil resources. ISME J.18:wrae073. doi: 10.1093/ismejo/wrae073
65
WangD.LiM.WangJ.LiN.LiuJ.YinH. (2026). Microbial necromass carbon accumulation is associated with soil microbial life history strategies and multitrophic network complexity under nitrogen addition in alpine forests. J. Plant Ecol.19:rtag087. doi: 10.1093/jpe/rtag087
66
WangC.LiX.ZhangM.HanZ.XiaY.LianP.et al. (2024). Nitrogen deposition enhances accumulation of microbial and plant-derived carbon in forest soils: a global meta-analysis. Plant Soil505, 465–478. doi: 10.1007/s11104-024-06687-7
67
WhalenE. D.GrandyA. S.GeyerK. M.MorrisonE. W.FreyS. D. (2024). Microbial trait multifunctionality drives soil organic matter formation potential. Nat. Commun.15:10209. doi: 10.1038/s41467-024-53947-2,
68
WrightS.Franke-SnyderM.MortonJ.WrightS. F.MortonJ. B.UpadhyayaA. (1996). Time-course study and partial characterization of a protein on hyphae of arbuscular mycorrhizal fungi during active colonization of roots. Plant and Soil181, 193–203. doi: 10.1007/BF00012053
69
XuH.QuQ.LiG.LiuG.GeissenV.RitsemaC. J.et al. (2022). Impact of nitrogen addition on plant-soil-enzyme C–N–P stoichiometry and microbial nutrient limitation. Soil Biol. Biochem.170:108714. doi: 10.1016/j.soilbio.2022.108714
70
YangY.ChaiY.XieH.ZhangL.ZhangZ.YangX.et al. (2023). Responses of soil microbial diversity, network complexity and multifunctionality to three land-use changes. Sci. Total Environ.859:160255. doi: 10.1016/j.scitotenv.2022.160255,
71
ZhangM.BaiX.WangY.LiY.CuiY.HuS.et al. (2023). Soil microbial trait-based strategies drive the storage and stability of the soil carbon pool in Robinia pseudoacacia plantations. Catena222:106894. doi: 10.1016/j.catena.2022.106894
72
ZhangW.ShenW.ZhuS.WanS.LuoY.YanJ.et al. (2015). Can canopy addition of nitrogen better illustrate the effect of atmospheric nitrogen deposition on forest ecosystem?. Sci. Rep.5:11245. doi: 10.1038/srep11245,
73
ZhangJ.TangX.HeX.LiuJ. (2015). Glomalin-related soil protein responses to elevated CO2 and nitrogen addition in a subtropical forest: potential consequences for soil carbon accumulation. Soil Biol. Biochem.83, 142–149. doi: 10.1016/j.soilbio.2015.01.023
74
ZhangJ.TangX.ZhongS.YinG.GaoY.HeX. (2017). Recalcitrant carbon components in glomalin-related soil protein facilitate soil organic carbon preservation in tropical forests. Sci. Rep.7:2391. doi: 10.1038/s41598-017-02486-6,
75
ZhangY.WangY.ZhouG.RevilliniD.LiuH.WuS.et al. (2025). Soil microbial networks mediate long-term effects of nitrogen fertilization on ecosystem multiservices. J. Ecol.113, 2539–2558. doi: 10.1111/1365-2745.70112
76
ZhangJ.ZhouJ.SayerE. J.LambersH.LiuZ.LuX.et al. (2023). Nitrogen deposition enhances soil organic carbon and microbial residual carbon in a tropical forest. Plant Soil484, 217–235. doi: 10.1007/s11104-022-05787-6
77
ZhuX.MinK.FengK.XieH.HeH.ZhangX.et al. (2024). Microbial necromass contribution to soil carbon storage via community assembly processes. Sci. Total Environ.951:175749. doi: 10.1016/j.scitotenv.2024.175749,
78
ZhuL.ZhangX.TaoJ.WangX.HanY.YeY.et al. (2024). The effect of nitrogen addition on soil respiration in the Abies Georgei var. Smithii forest of Sygera Mountains. Ecol. Environ.33:1384. doi: 10.16258/j.cnki.1674-5906.2024.09.006
Summary
Keywords
alpine coniferous forest, bacterial–fungal co-occurrence network, C:N:P stoichiometry, GRSP, MNC, nitrogen deposition, SOC
Citation
Zhang S, Han Y, Cui Z, Wu Z, Hu M and Ye Y (2026) Six-year nitrogen addition alters microbial-derived carbon accumulation in association with soil stoichiometry and microbial community reorganization in an alpine coniferous forest. Front. For. Glob. Change 9:1905473. doi: 10.3389/ffgc.2026.1905473
Received
10 June 2026
Revised
05 July 2026
Accepted
27 July 2026
Published
07 August 2026
Volume
9 - 2026
Edited by
Ling Zhang, Jiangxi Agricultural University, China
Reviewed by
Zheng Hou, Southwest Forestry University, China
Zhanbo Yang, Northeast Normal University, China
Updates
Copyright
© 2026 Zhang, Han, Cui, Wu, Hu and Ye.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Yanhui Ye, yeyanhui@xza.edu.cn
† These authors share first authorship
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.