Ecostoichiometry Reveals the Separation of Microbial Adaptation Strategies in a Bamboo Forest in an Urban Wetland under Simulated Nitrogen Deposition

: The e ﬀ ect of nitrogen (N) deposition on N limitation, phosphorus (P) limitation and the related soil and microbial stoichiometries remains unclear. A simulated nitrogen deposition (SND) experiment (control, ambient, medium and high) and molecular techniques (high-throughput sequencing of 16S and ITS) were conducted to examine the variations in abiotic and biotic properties and to describe the responses of microbial (bacteria and fungi) adaptation strategies in a moso bamboo ( Phyllostachys edulis J. Houzeau) forest following SND. Soil water content (SWC) was positively correlated with the microbial community composition. Observed increases in total N and nitrate N contents and decreased ammonia N suggested that SND inﬂuenced nitriﬁcation. Chao1 and F:B showed that bacteria were more sensitive to SND than fungi. PCoA and linear discriminant analysis (LDA), coupled with e ﬀ ect size measurements (LefSe), conﬁrmed that microbial community composition, including the subgroups (below class level), responded to SND by employing di ﬀ erent adaptation strategies. Soil C:N indicated that the soil of the moso bamboo forest was under N limitation prior to SND. The increase in total P (TP), available P (AP) and microbial biomass P (MBP) suggested the acceleration of soil P cycling. Microbial biomass C (MBC) and microbial biomass N (MBN) were not a ﬀ ected by SND, which led to a signiﬁcant shift in MBC:MBP and MBN:MBP, suggesting that P utilization per unit of C or N was promoted. There was a negative gradient correlation between the fungal community composition and MBC:MBP, while bacteria were positively correlated with MBN:MBP. The results illustrated that the response of fungi to MBC was more sensitive than that of bacteria in the process of accelerated P cycling, while bacteria were sensitive to MBN. Prior to P limitation, SND eliminated the soil N limitation and stimulated soil microorganisms to absorb more P, resulting in an increase in MBP, but did not alter MBC or MBN. This study contributes to our understanding of the adaptation strategies of fungi and bacteria and their responses to soil and microbial stoichiometries. ammonium-N 4 + -N), carbon (WSOC), water dissolved organic nitrogen (WSON), microbial biomass carbon (MBC), biomass nitrogen (MBN), and biomass


Introduction
N deposition has become an important global change factor [1]. During the last century, the amount of reactive nitrogen (N) has doubled globally, and the biodiversity of soil microorganisms has responded sensitively [2]. Active N produced by human activities enters the atmosphere and then settles in terrestrial ecosystems, which interferes with the natural cycling of N, changes the availability of N, and thus not only has a wide impact on the fixing capacity of forest N, but also accelerates soil N leaching, causes changes in soil N use efficiency and soil C:N and C:P stoichiometric ratios, and results in soil acidification. N and P availability plays a crucial role in C cycling in terrestrial ecosystems [3][4][5][6][7]. However, the soil stoichiometric regulation of the microbial community assembly and its impact on microbial ecological adaptability in the soil of wetland forests remains unclear.
The change in soil pH and stoichiometry under the continuous increase in N deposition makes the study of the soil microbial feedback mechanism an important topic in global change research [1]. Microbes regulate core ecosystem processes, such as organic matter decomposition, soil C retention and nutrient (N, P) transformation, and therefore are a vital part of the global biogeochemical cycle [8]. Ecological stoichiometrics are helpful to understand nutrient cycling relationships within ecosystems by analyzing the imbalances in the supply and consumption of ternary elements (C:N:P) between organisms and their resources [9]. This study provides a method for studying the impact of N deposition on the structure and function of plant-microorganism-soil systems. Homeostasis is key to nutrient recycling and the substrate quality effects of stoichiometry; different organisms vary in how they regulate their elemental composition in response to variations in resource stoichiometry [10,11]. A stoichiometry meta-analysis found that the fungal C:N:P stoichiometry is, on average, 250:16:1 (by atoms) [7], which is close to the Redfield ratio (106:16:1) for marine phytoplankton. Cleveland and Liptzin [12] proposed a hypothesis of soil homeostasis and reported a value of 60:7:1 for soil bacteria. The variation in elemental composition is linked to the differences in organismal allocation to biochemical and structural components that differ in C, N, and P contents [9]. Therefore, comparing soil microbial C:N:P stoichiometry can be used as a tool to evaluate nutrient limitations [13,14]. González-Chávez et al. [15] used microbial N:P (MBN:MBP) to examine the nutrient limitations of soil cultivated with N fertilizer over a long-term period, indicating that the cultivated soil was limited by P. The effect of P limitation on C and N fixation in ecosystems may be underestimated [16]. Studies have found that N deposition is likely to increase soil microbial C:N (MBC:MBN), N:P (MBN:MBP) and C:P (MBC:MBP) [17][18][19]. Studies have suggested that elevated N deposition can lead to decreased soil microbial biomass [20], and MBC:MBN and MBN:MBP are negatively correlated with N deposition [21][22][23]. Studies have also documented nutrient stoichiometry decoupled from microbial behavior [24,25]. As the soil changes induced by N deposition are mainly the result of microbiological processes, different niche characteristics, such as substrates and soil properties, are selected by specifically adapted microbial communities [7,26]. The microbial acquisition of a selected substrate is regulated by relatively stable C, N and P levels [27]. Ecostoichiometry, which allows the separation of soil biotic and abiotic environments with varying degrees of microbial alteration, might help to elucidate microbial-mediated soil C, N and P cycling characteristics [7,28]. Studies that examine the degree to which soil microbial C:N:P stoichiometry in natural communities is driven by adjustments to variations in external N supplies or by shifts in the relative abundance are still lacking.
The Yangtze River delta, an area with conflicts between economic development and natural ecological protection, is one of the highest N deposition regions in the world (exceeding 30 kg N hm −2 a −1 ) [29]. This delta is also an important wetland protection area in China, with a wetland area of 618.55 × 10 4 hm 2 [30]. The delta provides an ideal place to study the soil microbial adaptation strategies of wetland forests affected by the increasing threat of elevated N deposition due to the combustion of fossil fuels and a growing demand for N in agriculture [31]. The dike-pond system is a typical ecosystem of urban wetlands formed by a dense network of slow-moving rivers and lakes in the delta. With the development of industrialization and urbanization, the original dike-pond wetland ecosystem has changed, e.g., wetland reduction, or even disappearance, and water eutrophication have occurred. However, little is known about the soil microbial ecostoichiometric response mechanism to human-induced nitrogen deposition in the urban wetlands of these developed regions.
The present study was undertaken to determine how microbial communities respond to soil ecostoichiometric variations. Ultimately, our goal was to understand the effects of N deposition on abiotic (soil C, N and P) and biotic (microbial biomass C, N and P) properties and their characteristics of niche separation. Hence, we examined the microbial variations in topsoil (0-20 cm) under simulated N deposition (SND) and evaluated the compositional changes in bacteria and fungi communities using high-throughput sequencing of the 16S ribosomal RNA (rRNA) gene and ribosomal ITS, respectively. We hypothesized that (i) SND would influence the community structure and microbial stoichiometry by changing the soil stoichiometry and (ii) SND would depress the abundance of fungi but stimulate the abundance of bacteria due to their different preferences for adaptable strategies.

Overview of the Study Area
The experimental site was Xixi National Wetland Park (120 • 03 E, 30 • 15 N) in the western suburb of Hangzhou, Zhejiang Province, China. The details of the study area were described by Li et al. [32]. This area has a monsoon climate within the northern margin of the subtropical zone. The average annual temperature is 16.4 • C, and the annual precipitation is 1100-1600 mm. The wetland is formed by ponds, rivers, lakes, swamps, dikes and islets with aquaculture in ponds and cultivation of economic tree species on dikes in delta regions, fertilizing manure and pond sediment. The main soil type is paddy soil.

Field Experiment and Soil Sampling
Moso bamboo (Phyllostachys edulis J. Houzeau), a giant and arbor-like bamboo species, is one of the main economic species in the compound production system of dike ponds, which have grown naturally in the reserve in recent decades to better protect wetlands. The dike width of the moso bamboo forest ranges from 16 to 28 m. There are few herbaceous plants because of the high canopy density, such as Ophiopogon japonicas Ker-Gawl., Cardamine hirsute L., and Pteris multifida Poir., beneath the forest. The dike surface is 80-120 cm away from the water surface. The density, height, and DBH of moso bamboo are 0.7-1.1 stalks·m 2 , 9.0-11.4 m, and 7.3-12.5 cm, respectively. The light conditions beneath the forest are 7.8-13.1% compared with full sunlight.
According to the actual N deposition in subtropical China (30-37 kg·hm −2 a −1 ) [29], four levels of SND were derived: control (CK, 0 kg·hm −2 ·a −1 , ambient deposition), low N (N30, 30 kg·hm −2 ·a −1 ), medium N (N60, 60 kg·hm −2 ·a −1 ), and high N (N90, 90 kg·hm −2 ·a −1 ). Three 3 × 3 m replicate plots were set up in each of the four SND levels, with a total of 12 plots established in this study. There was an interval of 10-20 m between each plot as a buffer to prevent interaction. SND spraying began in October 2016. Briefly, the annual N application rate was divided into 12 parts, once per month. NH 4 NO 3 was fully dissolved in 2.7 L pure water and uniformly sprayed on the surface of each plot of forestland. After surface litter and impurities were removed, soil samples were collected from a depth of 0-20 cm from three randomly selected points in each plot in October 2018. Subsequently, a total of 36 soil samples were collected. The samples were sieved (=2 mm) for homogenization, and three portions of the sieved soil samples were retrieved. One portion of the sieved sample was stored at −78 • C for DNA extraction and 16S and ITS sequencing. One portion was air-dried, and the remaining fresh, portion was stored at 4 • C and used for the determination of soil properties [32].

Measurement of Soil Properties
The following soil properties were determined for each sample and used in the subsequent statistical analyses: soil water content (SWC), pH, soil organic carbon (SOC), total nitrogen (TN), total phosphorus (TP), available phosphorus (AP), total sulfur (TS), nitrate-N (NO 3 − -N) and ammonium-N (NH 4 + -N), water dissolved organic carbon (WSOC), water dissolved organic nitrogen (WSON), microbial biomass carbon (MBC), microbial biomass nitrogen (MBN), and microbial biomass phosphorous (MBP). The soil water content (SWC) was determined by the oven-dry weight method. The pH was determined in a 1:2.5 (w/v) soil:water extract using a glass LE410 pH electrode (FiveEasy™ pH meter; Mettler-Toledo AG, Zurich, Switzerland). SOC, WSOC and WSON were measured using a total organic carbon (TOC) analyzer (multi N/C 3100 and HT 1300; Analytikjena, Jena, Germany). TC, TN and TS were determined using an elemental analyzer (vario MAX cube; Elementar Analysensysteme GmbH, Hanau, Germany). TP was measured by the Mo-Sb anti-spectrophotometric method. The soil C:N, C:P and N:P ratios were calculated using the SOC, TN and TP datasets. NO 3 − -N and NH 4 + -N were extracted with 2 M KCl and then determined using a continuous-flow ion autoanalyzer (Flowsys III; Systea, Anagni, Italy). MBC and MBN, determined using the chloroform-fumigation-extraction method [33], were calculated as the difference between fumigated and nonfumigated samples, adjusted by a coefficient (k EC , k EN = 0.45). MBP was determined according to Brookes et al. [34]. With the exception of a 5.0 g soil sample being used, the fumigation procedure was the same as that used for MBC and MBN with an adjusted coefficient (k EP = 0.4).

Data Processing and Statistical Analysis
Obtained DNA sequence reads were trimmed, and then chimeras were removed using USEARCH, and sample sequences were combined using QIIME_1.9.1 software. An operational taxonomic unit (OTU) was defined as a set containing DNA sequences with a similarity of >97%. OTUs representing less than 0.005% of all sequences were removed before analysis. Diversity indexes were calculated using QIIME software.
One-way ANOVA, multiple comparative analysis of OTU frequencies (α = 0.05) and nonparametric multivariate analysis of variance (PERMANOVA) were conducted using the Data Processing System (DPS) (17.10 for Windows, Zhejiang University, Hangzhou, China) [37]. The contributions of soil properties to the dissimilarities in microbial communities and differences in the relative abundances of microbial phyla were based on the DPS module of the Spearman correlation and best multiple regression. Alpha diversity type complexity, principal coordinates analysis (PCoA) and redundancy analysis (RDA) were conducted using the "vegan" and "ggplot" packages in R [38,39]. Absolute quantity (i.e., observed species) was used to calculate fungi:bacteria (F:B) [26]. Linear discriminant analysis (LDA) coupled with effect size measurements (LefSe) was applied to measure the main differential clades using the "stats" package in R [40,41].

The Effects of SND on Soil Physicochemical Properties and Ecostoichiometry
Increased N deposition via SND resulted in a decline in soil pH, most notably at N60 and N90, compared with that in CK (Table 1). SOC decreased at N90, and WSOC also decreased at N30, N60 and N90. WSON at N90 was different from that of CK and was found to decrease, but no significant difference was observed between N30 and N60, suggesting an increased risk of N leaching and water eutrophication.
There was no difference in MBC and MBN among the different treatments, but MBP showed a difference at N90 compared with the other treatments. NH 4 + -N and NO 3 − -N exhibited the opposite result, with NH 4 + -N decreasing with increasing SND, while NO 3 − -N increased at N90.
As SND increased, soil C:N decreased significantly at N90, which was different from that of N30 and CK, while N60 was different from CK ( Figure 1). There was no difference in C:P among treatments, while N:P reached the maximum value at N60. The trends of MBC:MBP and MBN:MBP were similar, with no difference among N30, N60 and N90, but both decreased significantly compared with the values observed for CK.  Table 1. Soil biochemical properties of sample plots in the moso bamboo forest under different SND levels. Values represent the mean ± SD (n = 9). Values with different letters within a column indicate that the difference was significant among the plots.

Effect of SND on the Alpha Diversity of Bacteria and Fungi
There was a stronger response in bacteria to SND than that for fungi (Table 2). Specifically, the Chao1 index reached its peak value at N30, with a significant difference compared with CK, but there were such no responses at N60 and N90. The OS also showed the same trend, indicating that the bacterial community of CK was at a disadvantage; the population composition of N30 was the most abundant, and SND stimulated the quantity and richness of the bacterial community at N30. The Shannon and Simpson indexes were not different among the treatments, indicating that SND had no role in improving the bacterial diversity and community evenness. Fungi had a certain inertia and showed no difference, and all the average values of the index (Chao1, OS, Shannon and Simpson) showed only a downward trend. F/B declined from ambient levels at higher N deposition rates.

Effects of SND on the Beta Diversity of Soil Bacteria and Fungi
The PCoA results of the bacterial community in the moso bamboo forest showed that the PC1 and PC2 dimensions calculated from the distance matrix explained 27.98% and 16.75%, respectively, of the OTU variation in the soil bacterial composition at the genus level (Figure 2A). The bacterial communities of N30, N60 and N90 were different from that of CK, which is consistent with the PERMANOVA results. There was no significant difference between N30 and N60, both of which were different from N90, indicating that N30 and N60 had similar effects on the composition of the bacterial community. The PC1 and PC2 dimensions calculated from the distance matrix explained 20.65% and 14.26%, respectively, of the OTU variation in the soil fungal composition at the genus level ( Figure 2B). N30, N60 and N90 influenced the fungal community structure for those fungal communities that differed from that in CK. N60 and N90 had similar influences (PERMANOVA, N60 vs. N90 p-value = 0.1014), which suggested that both N60 and N90 exerted a greater effect than N30 based on the distances from CK.
LEfSe detected differential clades at all SND levels, which clarified statistically significant differences om diversity among the bacterial and fungal communities (Figure 3). At an LDA score of ≥4, LEfSe showed that the most notably differential clades were Acidobacteria (5) and Chloroflexi (3) in the CK, Acidobacteria (2) and Planctomycetes (1) at N30, Acidobacteria (2) at N60, and Acidobacteria (2) at N90 ( Figure 3A). The cladogram also identified the Incertae_sedis (1), Ascomycota (3) and Glomeromycota (3) in the CK; Glomeromycota (3), the Cunninghamellaceae family and the Mucorales order at N30; the Dothideomycetes class at N60; and an unidentified phylum (3) at N90 for fungi ( Figure 3B). At an LDA score of ≥2, the numbers of bacterial discriminant clades from the CK to N90 were 58, 90, 47, and 108, respectively, while the discriminant clades of fungi from the CK to N90 numbered 67, 43, 53, and 56, respectively. At the CK level, the LEfSe indicated Actinobacteria (21), Chloroflexi (10), Proteobacteria (9), Acidobacteria (7) and Firmicutes (5) as the most differentially abundant bacterial taxa. It also revealed that, in contrast to CK, Proteobacteria (23, 11, and 40) was commonly found at the N30-N90 levels. Fungal LEfSe indicated that Ascomycota (26, 18, 37, and 22) and Basidiomycota (34, 12, 11, and 18) were the main discriminant clades, with Basidiomycota prominent in the CK. These results indicated that both bacterial and fungal groups differed under SND, thereby suggesting that both bacteria and fungi are sensitive to N deposition.  Each small circle at a different taxonomic level represents a taxon. The diameter of the small circle is proportional to the relative abundance of the OTU. The coloring of the non-significant species is yellow, and the different taxa are colored with groups, representing the microbial groups that play an important role in the group. The names of the taxa are represented in the legend on the right.Below each cladogram, discriminant clades among CK, N30, N60 and N90 treatment soil are reported at an LDA score ≥2 for bacteria (C) and fungi (D). The tables underneath each cladogram report phyla/classes that statistically significantly discriminate CK, N30, N60 and N90 treatment soils.

Association between Soil Properties and the Microbial Community Composition
Under SND conditions, the relative abundance of bacterial and fungal communities at the phylum level can be expressed by different physicochemical parameters based on the best model ( Figure 4). The correlation of the values with the differences in soil properties between bacterial and fungal phyla showed that microbial diversity exhibited strong, microhabitat-specific patterns. The major predictors were identified according to the correlation and the best multiple regression model, suggesting that this approach is helpful for us to understand the relationship between microbial community changes and microhabitat along the N deposition continuum.
Based on a VIF test, the inflation factors of all physicochemical parameters were smaller than 20.0, among which 11 parameters were selected for bacterial RDA ( Figure 5A). The four-dimensional axes revealed that the cumulative percentage variance (CPV) of bacterial genus data was 50.9%, and the CPV between the bacterial genus composition and the environment was 78.  Variance partitioning analysis (VPA, Figure 5C) with partial RDA showed that the C, N, P, soil CNP ratio and microbial biomass CNP ratio groups accounted for 5.8%, 13.2%, 8.3%, 5.1% and 6.2%, respectively, of the total bacterial genus variance after the fitting of covariables.  Figure 5D). The CPV of the fungal phylum data was 27.3%, indicating that there were other properties that were more influential than the factors we examined in this study.

Discussion
Soil acidification is an important cause of soil quality decline and further causes a series of ecological problems, such as ecosystem degradation, biodiversity reduction, and nutrient loss [42]. At present, it is generally accepted that long-term N addition will cause a significant decrease in soil microbial biomass because of soil acidification and a decline in soil microbial activity [18,22]. Our results showed that the N60 and N90 levels had an impact on soil pH, but pH had no gradient correlation with the microbial community composition, while SWC did have a gradient correlation, which may be related to the short implementation of SND, which had not yet affected the overall microorganism community [43]. The change in the microbial community's abundance of individual phyla was related to pH, which indicates that pH and other related parameters can be used to predict the dynamics of microbial abundance. TN increased with decreasing pH, which was similar to the results of nitrogen addition experiments in many ecosystems, indicating that N deposition could significantly improve soil N content [44]. In turn, N deposition can improve the rate of soil net nitrogen mineralization and greatly improve available N in the soil, thus increasing the amount of nitrogen available to plants and soil microorganisms [43,45]. In our experiment, ammonia N decreased, and nitrate N increased, which was similar to the research results of short-term N input by Hall and Matson [46], indicating that the net nitrification rate gradually increased with increasing N addition dose [47]. In general, the moso bamboo forest was short of N, especially with respect to the natural growth status, i.e., 20 years in our study, which is related to its preference for ammonium-N and strong clonal breeding habits [48]. Therefore, N deposition in moso bamboo forests can alleviate the degradation of ecosystems caused by N deficiency to a certain extent. Consequently, a slight N addition was beneficial to the stability of the moso bamboo forest and soil flora. It was suggested that 30 kg·hm −2 ·a −1 N can be used as a reference value for maintaining the development of bamboo ecosystems over a short period. However, the long-term application of chemical N fertilizer alone will accelerate the leaching loss of soil alkaline matter [6]. Furthermore, soil stoichiometry changed with the increase in nitrate N and, in the aerobic environment, the bacteria could not obtain enough available N (except for denitrifying bacteria) during competition with moso bamboo, which led to an imbalance in the bacterial community's composition and variation in life strategies and ecological adaptability [49]. Therefore, over the long term, N deposition is not conducive to the healthy development of the ecosystem.
The N and P pools are closely related, and their consumption processes control the N and P circulation within microorganisms [13,50]. The increase in the average value of TN in this study showed that exogenous N input alleviated nitrogen deficiency in the moso bamboo ecosystem, which is similar to the results of most studies [5,31,42,51]. Wang's results [52] suggest that a soil C:N greater than 25.0 is nitrogen-limiting for microorganisms. Our experiment shows that the soil C:N was between 12.6 and 26.8, and some of the control plots (CK) were in the critical state of N limitation. Therefore, in this non-N-limited environment of SND, N was not utilized by microorganisms but was nitrated or denitrified [53]. The soil C:N directly controls the nitrification rate and indirectly affects the nitrate used for denitrification [10]. This study also confirmed this point of view: with the decline in the C:N ratio and the increase in nitrate N, the development of the bacterial community had a negative gradient correlation with ammonium-N and C:N, which shows that SND affected soil C:N and was closely related to the decomposition ability of soil microorganisms [4]. However, in a non-N-limited environment, high growth rates are needed to increase the investment of P-rich ribosomal RNA, which means a decrease in the soil N:P when the soil status becomes P-limited [10]. In contrast, our moso bamboo forest may be in the transformation period from non-N limitation to P limitation but not but not yet experiencing P limitation, which was identified by the increased average value of TP and AP and the increasing soil N:P in N60 and N90. Moorhead and Sinsabaugh [54] showed a decrease in soil C:P, indicating that N had a positive effect on the mineralization of organic and inorganic P, which is consistent with our results. This finding suggests that N deposition may stimulate mineralization and increase biological demand for P [55], resulting in accelerated P cycling in forest ecosystems, especially litter decomposition and P acquisition [56], which may lead to P limitation in the long term.
The microbial community is constantly adapting to changes in specific microhabitat conditions provided by the soil [28]. The variation in microbial stoichiometry is the change in different adaptation strategies of microorganisms [4]. It was found that 12 years of N deposition in broad-leaved forests in the northern United States resulted in a decrease in soil microbial biomass and a 10% decrease in F:B [51]. In our study, the Chao1 index showed that bacteria were more sensitive to SND than fungi, which was related to their different life strategies (r-K strategy) [27]. Fierer et al. [57] found that F:B increased with increasing soil C:N in a meta-analysis of the global soil microbial community pattern, which is similar to our research results. After SND, F:B decreased due to the decline in C:N, mainly because K-strategy fungi had higher biomass C:N than bacteria [3]. Li et al. [48] suggested that N deposition can increase the soil MBC content, thus increasing MBC:MBN and reducing competition between plants and microorganisms [58]. Our results showed some differences from those conclusions, as they indicated that MBC decreased while MBC:MBN changed little. The reasons for this may be as follows: first, the active soil C pool of the moso bamboo forest was small and the degradation was very fast in the case of SND, which led to the loss of active organic C [5]. Therefore, there was not enough active organic matter to support high metabolic rates. Second, SND alleviated the rhizosphere microbial driving demand for nutrients, thus reducing the C input flux of plant root exudates; i.e., plants adopted the physiological strategy of a low N yield and low C input under the condition of N enrichment [59]. Consequently, sufficient active C could not be provided, which directly led to the decline in the microbial biomass. Third, due to the limitation of other nutrients, SND did not change the C ratio of microbial cells (no difference in MBC:MBN) and did not reduce the C ratio of microbial metabolic consumption [15]. In the context of wetland protection, bamboo forests require management that includes the application of an organic P fertilizer to help alleviate P limitation.
Our results showed that MBP increased and caused a significant decrease in MBC:MBP and MBN:MBP, which is similar to previous research results [21][22][23]. Deng's global meta-analysis [49] showed that N addition did not change microbial P but did enhance phosphatase activity. This different conclusion may, in part, be a matter of scale. The generalities derived from global meta-analyses tend not to detect stand-scale responses. Our results confirmed that N can release other nutrients (such as P) by producing N-rich enzymes, such as phosphatase, by shifting from a critical state of N limitation to a P-released state after sufficient N is provided. Therefore, short-term SND is distinct from the long-term ecosystem succession process of N deposition, in which MBN:MBP will increase, while short-term SND will result in a short period of P release followed by a P-limitation period. Furthermore, we found that MBC:MBP had a gradient correlation with the fungal community compositions, while MBN:MBP had a gradient correlation with bacteria, indicating that the P utilization efficiency per unit of C or N increased and SND raised the demand of soil microorganisms for P [60], implying that different microbial groups have their own unique adaptive strategies [11].

Conclusions
In summary, 2 years of SND reduced ammonia N and accumulated nitrate N, resulting in soil acidification, but the pH value showed no correlation with the microbial community compositional response. SND did not alter the alpha diversity of fungi but changed the bacterial Chao1 index, suggesting that bacteria were more sensitive than fungi. SND changed the F:B and beta diversity of both fungi and bacteria, suggesting that the microbial community composition was modified. The soil of the moso bamboo forest was in a state of N limitation and P cycling acceleration. During the short-term stimulation process of N and before P limitation, MBC:MBP and MBN:MBP experienced a short period of decline, indicating that the P utilization efficiency was enhanced. Under SND conditions, there may be heightened risk of bacterial breeding and N and P limitations over the long term in natural growth of the wetland moso bamboo forest, which will further affect the sustainable and healthy development of those wetland ecosystems. It is suggested that the necessary management interventions should be applied, such as the application of organic fertilizer.

Conflicts of Interest:
All the authors declare that they have no conflict of interest.