Long-term nitrogen addition increased soil microbial carbon use efficiency in subalpine forests on the eastern edge of the Qinghai–Tibet Plateau

Nitrogen (N) deposition increased forest carbon (C) sink significantly, hence exploring the microscopic mechanisms is critical to predicting future global ecosystem C cycle, especially the effects of enhanced N deposition on soil microbial carbon use efficiency (CUE), which still unclear. We evaluated the responses of soil microbial CUE to long-term (5 years) N addition in an evergreen broad-leaved forest and a mature coniferous forest by using a 13C isotope tracing method. The results showed that the soil microbial CUE ranged from 0.38 to 0.51, which was smaller than the results obtained from the previous studies based the same method and forest type. In evergreen broad-leaved forest, the microbial CUE had no significant changes in the low N-addition treatment, but it was increased by 9.23% and 12.69% in medium and high N-addition treatments compared to the control. In coniferous forest, soil microbial CUE was increased by 14.64%, 21.89% and 24.34% in low, medium and high N-addition treatments, respectively. Moreover, the soil C:P and N:P are negatively relate to soil microbial CUE. Our findings indicate that the enhancing N deposition can increase soil microbial CUE and ultimately promote C sequestration, especially in coniferous forest. The imbalance of soil stoichiometry is the main impact factor of CUE under N addition. However, we speculate that the key to increase forest soil microbial CUE is to promote the decomposition rate of litter and thus increase the available C content.


Introduction
N deposition plays an important role in terrestrial ecosystem C cycle. It was estimated that the global forest C sink caused by N deposition was about 0.72 Pg C yr −1 in the 2010s (Gurmesa et al. 2022). Previous studies have shown that N deposition regulates the microbial C cycle of forest soil by changing the ratio of soil nutrient elements, therefore, N deposition may have effects on forest soil C sequestration (Wamelink et al. 2009). In addition to affecting the growth of vegetation, soil N content also plays a decisive role in the growth and metabolism of soil microbial communities (Cleveland and Liptzin 2007). Researchers have carried out soil N addition experiments, and well understand the changes of soil microbial metabolism in grassland and cultivated land (Fang et al. 2018;Luo et al. 2020).
However, the responses of microbial C cycle process to N fertilization in forest soil remain controversial. Soil microbial CUE, defined as the ratio of microbial C allocated to growth, is an important synthetic representation of the microbial community metabolism. Despite it is vital for the forecast of global C cycle, mutability in CUE for terrestrial microorganisms is still poorly understood, and excluded from most biogeochemical models due to the varying methods and spatial heterogeneity (Manzoni et al. 2012). It has been speculated that the imbalance of soil stoichiometry, such as high C:N, can trigger overflow respiration, leading to relatively low microbial CUEs (Oliver et al. 2021;Pei et al. 2021;Subedi et al. 2021). Based on the ecological stoichiometry theory, the C:N ratio of soil microbial biomass is constrained within the range of 7-8.6 to keep internal balance of the elementals. It means that microorganisms must allocate larger proportions of C and energy to acquire C or nutrients in the case that the C:N ratio goes beyond this range, leading to a decrease in microbial CUE (Manzoni et al. 2012;Sinsabaugh et al. 2013). Manzoni et al. (2010) found that the maximum value of soil microbial CUE is about 0.3 under natural conditions, which is half of the thermodynamic calculation value. Hence, assessing the effects of N addition on forest soil microbial CUE can help us better understand the mechanisms of N deposition affecting soil C sink and enhance the predictive accuracy of relevant models.
It is generally accepted that N addition increases soil microbial CUE. For instance, Spohn et al. (2016) found that the long-term addition of N fertilizer to grassland soils gave resulted in a decrease of microbial C uptake and respiration rates, while the growth rates remained unchanged, resulting in an increase in microbial CUE. Soares and Rousk (2019) observed that the higher the N availability, the higher the CUE, but they believed that site-specific differences overrode the effect of N-availability. However, it is found from recent studies that N addition may have no effect in specific contexts. Li et al. (2021a, b) conducted a six-year N deposition experiment to evaluate the response of microbial CUE to N addition in a temperate forest, and the results showed that microbial CUE was increased by 133.18% with high N (7.5 g N m −2 yr −1 ) addition, while remained unchanged under low N addition (2.5 g N m −2 yr −1 ) in mineral soil. Taken together, there are two speculations for the enhanced soil microbial CUE induced by N addition. Firstly, the N-fertilizer in soils acts as an inhibitor for oxidative enzymes involved in the mineralization of complex compounds, leading to a decrease of microbial respiration and a reduction of the energy requirement of microbial N acquisition, and a further, increase of microbial CUE (Gallo et al. 2004;Manzoni et al. 2010;Spohn et al. 2016). Secondly, higher N availability generated a lower fungalto-bacterial ratio, resulting in a relative high CUE due to the fact that bacteria had a higher CUE than fungi (Maynard et al. 2017).
In this study, we planned to study the effects of N deposition on subalpine forests soil microbial CUE by artificially simulating N deposition. Therefore, our research objectives are as follows: 1) Estimate the effects of long-term N addition on subalpine soil microbial CUE, compare the sensitivity of soil microbial CUE to long-term N addition between evergreen broad-leaf forest and coniferous forest on Gongga Mountain; 2) Speculate the underlying mechanisms and influencing factors of the changes of forest soil microbial CUE under N addition.

Study area
The N-addition experiment was conducted in Mt. Gongga (29°20′ − 30°20′ N, 101°30′-102°15′ E), which is on the southeastern fringe of the Tibetan Plateau. Due to the large amount of fertilization on cultivated land in this area and the gradual increase in the flow of people and vehicles, the dry and wet N deposition rates of atmospheric N in the Mt. Gongga are relatively high. The rate of atmospheric N deposition (dry and wet deposition) in Mt. Gongga is about 8.46 kg ha −1 y −1 , which is greater than that in most areas of western China (Liu et al. 2008), so it is an ideal area for studying the influence of N deposition on forest soil microbial CUE. The N-addition experiment was established in an evergreen broad-leaf forest (the soil type is yellow brown soil, 2257 m) and a coniferous forest (the soil type is dark brown soil, 2839 m), these two vegetation types are typical and widely distributed in Mt. Gongga. In the evergreen broad-leaf forest, the predominant species is Lithocarpus cleistocarpus and is about 25 ~ 30 m high, and the stand density is about 525 trees per hectare. In the coniferous forest, the predominant species is Abies fabri and is about 20 ~ 30 m high, and the stand density is about 800 trees per hectare. East Asian monsoon is the dominant climate in Gongga Mountain, and the mean annul precipitation was 1403 mm at 2200 m and 1938 mm at 2900 m. The mean annual temperature was 13-14 °C at 2200 m and 3.5-5.0 °C at 2900 m. The soil is mostly acid soil with a pH value ranging from 4.5 to 6.

Nitrogen addition experiment
The N addition experiment was established in 2015, since the N deposition rate in the area is increasing and we designed three gradients of N-addition rate (applicated 10 kg N ha −1 yr −1 , 20 kg N ha −1 yr −1 and 40 kg N ha −1 yr −1 , respectively) and a control (0 kg N ha −1 yr −1 ), marked as N 10 , N 20 , N 40 and N 0 , respectively. The artificial N addition rates was about 1, 2 and 4 times of the natural atmospheric N deposition, respectively. There are 3 uniform plots (1 m × 2 m) per treatment for each forest type (24 plots in total). The plots were randomly selected and include plants (mainly Oxalis corniculata, Carex schneideri and Polygonum viviparum) growing in it, and the plant coverages ranges from 50 to 90%. Embed 5 cm deep PVC boards around the plots to isolate them from the outside soil. We applied the N fertilizer (urea) once a month from May to October on yearly basis (6 times a year in total), by mixing it in 500 ml deionized water, and then spreading it on the surface. In order for fertilizer to flow into the mineral soil, large litters with large surface areas were removed before each fertilization application (reset them after fertilizing). An equivalent volume of deionized water was sprayed in the area that extends outward 0.5 m around the quadrat.

Samples collection
Soil samples were collected in August 2020, and the litters and humus were removed before sampling at each site. The samples were taken with cutting rings randomly (diameter = 5 cm, volume = 100 cm 3 ) ranging from 0-10 cm in each of the replicate for measuring several soil physical indicators (e.g., water content, bulk density, and porosity, n = 3). Given the acreage of the plots, we collected four samples in S-shape from 0-10 cm in each of the replicate, mixed them together and sieved them (2 mm) immediately. Ice bags were used to keep the samples at a lower temperature during the transportation. Each sample was divided into three parts: i) For incubation, the subsample was stored in the frozen state (-20 °C) in Ziploc bags until further use; ii) For microbial properties testing, it was stored under 4 °C until further use; iii) For soil physical and chemical properties testing, the subsample was air-dried at room temperature.

Analysis of soil properties
Soil water holding capacity was determined by gravimetric method. In short, weighed the samples after soaking it for 12 h, take a part of it (about 20 g) and dried it (with a temperature of 105 °C) to calculate its moisture conversion factors, and calculated soil bulk densities and water holding capacities according to the conversion factors. A pH meter (Precision and Scientific Corp, China) was used to measuring soil pH (soil: water = 1:2.5). An elemental analyzer (Elementar Vario MACRO cube, Germany) was used to determine soil organic carbon (SOC) and total nitrogen (TN). Soil total phosphorus (TP) was extracted into sulfuric acid (98%) and then measured by an Auto Discrete Analyzer (Smartchem 200, AMS, Italy).
Soil microbial biomass carbon (MBC), phosphorus (MBP) and nitrogen (MBN) were determined by fumigation extraction method. K 2 SO 4 solution (0.5 M) was used to extract dissolved carbon (DOC) and dissolved nitrogen (DON) in fumigated (fumigated with chloroform for 24 h) and non-fumigated soil (Brookes et al. 1985;Vance et al. 1987). The conversion factors of calculating the MBC and MBN were 0.45 and 0.54, respectively (Wu et al. 1990). Dissolved phosphorus (DOP) was extracted into NaHCO 3 solution (0.5 M) from fumigated (fumigated with chloroform for 24 h) and non-fumigated soil, and the conversion factor was 0.40 for calculating MBP (Brookes et al. 1985

Determination of soil microbial CUE
For the understanding of the metabolic process of soil microorganisms, we applied the 13 C isotope tracing method to determine soil microbial 13 C-respiration and 13 C-growth rate following Jones et al. (2018). Before incubating, samples were kept for 3 days at 15 °C (average soil temperature at the sites during the growing season) to reactive the microorganisms and the water content of soils was kept at 60% of soil's water holding capacity to maintain an optimal conditions for microbial activity. At the end of the pre-incubation stage, 5 g of the pre-incubated soil was placed into a polypropylene tube (50 cm 3 ) and then received a quantified glucose solution (0.2 mg glucose-13 C g −1 soil) that included other reagents (0.1% MgCl 2 , 0.2% KH 2 PO 4 , and 0.1% K 2 SO 4 ) and sodium nitrate (0.1 M) to make it C:N ratio = 40 to mimic the circumstances of natural environment (Wadsö 2009). 2 ml sodium hydroxide solution (NaOH, 1 M) was injected into a polypropylene scintillation vial, and placed the scintillation vial upright on the soil surface to catch the respired 13 CO 2 , sealed the tubes and put them into a thermostat with a temperature of 15 °C. After 72 h, the polypropylene tubes and scintillation vials were harvested and the remaining 13 C-glucose was extracted in ice-cold 1 M NaCl (25 ml), the cultivation time was selected based on Glanville et al.
(2016) that most of the glucose have been consumed by the microorganisms. The content of 13 C in NaCl and NaOH was determined by Delta V Advantage (Thermo, America) and ISOPRIME100 (Elementar, German), respectively.
Microbial 13 C-uptake ( 13 C u ) can be estimated as follows: where 13 C t is the total amount of 13 C-glucose added to each of the sample, 13 C NaCl is the amount of 13 C extracted in the 1 M NaCl. Microbial CUE for 13 C-glucose was estimated by: where 13 C r is the amount of 13 C trapped by the 1 M NaOH, which is represents the 13 C-respiration. The threshold (TER) of C:N ratio was estimated by (Sterner and Elser 2002;Soong et al. 2020): where C:N MB is microbial biomass C:N ratio, NUE is the microbial N efficiency and it can be estimated by (Zhong et al. 2015): where C:N soil is C:N ratio of the soil.

Statistical analysis
The data were checked for normality and homogeneity of variance, and transformed in the case that they were not normally distributed. After that, a one-way ANOVA was performed to test the differences of the properties of soil (soil bulk density, water-holding capacity, pH, SOC, TN and TP) and the microorganisms (MBC, MBN and MBP) in the treatments, which was followed by a Tukey post-hoc test for multiple comparisons. The significant differences were calculated at the confidence interval of 95%. Similarly, the differences of the soil microbial metabolic parameters ( 13 C-uptake, 13 C-growth, 13 C-respiration and CUE) at varying N-addition rates and soil types was analyzed by using a two-way ANOVA. Furthermore, a liner regression model was established to describe the relationships between soil C:P (or N:P) ratio and microbial CUE. Besides, a redundancy analysis (RDA) was performed to reveal the influences of soil properties on microbial biomass and parameters in metabolic process. The one-way and two-way ANOVA analyses were carried out by using SPSS 25.0 software, and the RDA analysis was conducted using with the vegan package (version 2.5.7) of R software (version 4.4.1).

Results
Effects of N addition on soil properties in the two subalpine forest types Soil bulk density and water-holding capacity remained almost unchanged under long-term N addition. Soil pH slightly decreased with the increase of N addition rates, and evergreen broad-leaf forest had greater average soil pH than coniferous forest. Soil SOC and TN in evergreen broad-leaf forest first Vol.: (0123456789) increased and then decreased with the increase of N addition rates, while SOC and TN in the two forest types showed no consistent changes after N fertilizer was added. Soil C:P (SOC:TP) ratio in N addition treatments were generally less than that in controls in both coniferous forest and evergreen broad-leaf forest. Moreover, total phosphorous (TP), C:N (SOC:TN) and N:P (TON:TP) and N:P (TN:TP) did not show significant differences. In general, there was no significant differences in terms of the content of microbial biomass C, N, and P in the two types of forest soil. N addition decreased MBC and MBN content in evergreen broad-leaf forest, but had no significant effects on that in coniferous forest. Furthermore, N addition can increase the soil microbial phosphorus content in coniferous forests.
Effects of N addition on soil microbial CUE in the two subalpine forest types N addition treatments had no significant influences on microbial 13 C-uptake rate in evergreen broad-leaf forest in which 13 C-growth rates were enhanced in N 20 and N 40 treatments, but showed no significant changes in N 10 treatments. However, all N addition treatments are available to promote the microbial 13 C-growth in coniferous forest. Besides, N addition decreased microbial 13 C-respiration rate to different degrees, but the amplitude of the decrease was not significant in N 10 treatments in both soil types. We demonstrated that soil microbial 13 C-respiration was decreased by 1.97%, 11.00% and 6.99% in N 10 and N 20 and N 40 treatment compared with the controls respectively in evergreen broad-leaf forest, and in coniferous forest, the decreasing amplitude was 10.01%, 14.14% and 11.18%, respectively.
Consequently, microbial CUE shows different responses to N addition treatments in the two forest types. In evergreen broad-leaf forest, microbial CUE was not affected in N 10 treatment, which, however, was significantly increased in N 20 and N 40 treatments. In that case, in coniferous forest, all the N addition treatments had increased microbial CUE significantly compared to the control. The microbial CUE of N 10 , N 20 and N 40 treatments were increased 1.13%, 9.23% and 12.69% respectively in evergreen broad-leaf forest. While in coniferous forest, the microbial CUE increments were 14.64%, 21.89% and 24.34%, respectively. Thus, the microbial CUE in coniferous forest was more sensitive to long-term N addition than that in evergreen broad-leaf forest. There was no significant difference in soil microbial CUE between the two forest types (P = 0.716).

Relations of soil properties and microbial CUE under N addition
The redundancy analysis (RDA) plot shows the relationships between soil properties and microbial metabolic parameters (Fig. 1), revealing that N addition increases microbial CUE by decreasing soil C:N ratio. Constrained axes RDA1 and RDA2 accounted for 67.68% and 23.57%% of total variations, respectively. The pH value has the most significant influence on microbial community physiological characteristics under N addition, and it is positively correlated with MBC and MBN. SOC was negatively correlated with microbial CUE. As we can see from the results of RDA, the microbial 13 C-growth was negatively related to the C:N ratio, and microbial 13 C-respiration is positively related to the C:N ratio, resulting in an inverse relationship between the microbial CUEs and the C:N ratios. In addition, since SOC is proportional to C:N ratio, they were also inversely proportional to microbial CUEs. In this study, N addition resulted in a larger soil N:C ratio, thus promoting microbial growth, inhibiting microbial respiration, and ultimately leading to an increase in soil microbial CUE (Fig. 2). Furthermore, we observed that soil phosphorus content (P) and water-holding capacity had little effect on the growth of soil microorganisms.
Although soil P has little effect on the process of microbial heterooxygen respiration, the imbalance of C, N and P in the soil has a great impact on the CUE. It can be seen from the results of regression analysis among soil C:P and N:P ratio with microbial CUE that both the C:P (R 2 = 0.534, P < 0.01) and the N:P (R 2 = 0.499, P < 0.01) ratios were inversely related to the microbial CUE across the sites ( Fig. 3a and b). However, soil C:N ratio has no significant relationship with microbial CUE (R 2 = 0.085, P = 0.09).

Effects of N addition on soil microbial CUE
We found that the soil microbial CUE across all the sites ranges from 0.38 to 0.51 with an average of 0.44, the results were lower than those of other studies using the same method (Sauvadet et al. 2018;Takriti et al. 2018). Our results suggest that N additions can lead to an increase of soil microbial CUE, which is in line with the previous studies (Spohn et al. 2016;Poeplau et al. 2019). However, our finding contrasts with Widdig et al. (2020) and Riggs and Hobbie (2016), their researches showed that N addition had no impact on soil microbial CUE, and Silva-Sánchez et al. (2019) observed that N availability was negatively correlated with microbial CUE. We observed that the microbial CUE stopped growing when the N addition was increased to 20 kg·ha −1 ·yr −1 , therefore, we can speculate that the enhancement of N addition on soil microbial CUE peaked at nearly 20 kg· ha −1 ·yr −1 .
According to our observation, the increased soil microbial CUE was due to the declined microbial respiration and the advanced growth (Yuan et al. 2019;Li et al. 2021a). Our finding coincides with the previous work that microbial 13 C-respiraion was decreased and 13 C-growth was increased under N addition treatments (Söderström et al. 1983;Treseder 2008;Widdig et al. 2020) and there may be two mechanisms leading to this effect. On the one hand, microorganisms may consume less adenosine triphosphate (ATP) for the metabolism associated with N acquisition in N enriched soils, therefore, the excrescent C was allocated to growth, leading to the increase of microbial CUE (Manzoni et al. 2012). On the other hand, it has been confirmed that N addition can alter decomposer community composition to have more labile substrates, because soil N is a key element for microbial enzyme production, especially soil acid phosphatase (Allison and Vitousek 2005;Liu and Zhang 2019). Subedi et al. (2021) observed that greater pine litter mass was decomposed under N + P addition treatment compared with the control treatment in a long-term silvicultural study area. Moreover, previous studies demonstrated that microbial uptake was normally declined, which was induced by the decline of microbial biomass in N addition experiments (Wu et al. 2019;Soong et al. 2020;Widdig et al. 2020), but our result was different. We have also found that N addition has no significant effect on soil microbial 13 C-uptake rate, which is different from the findings of Spohn et al. (2016), who found that N additions reduced microbial 13 C-uptake rate. The substrate we added was glucose in the laboratory culture experiment, which is absorbed very quickly by microbial communities, therefore, there was no discernible difference in 13 C-uptake. Interestingly, however, the increasing ranges of soil microbial CUE in our study were relatively low compared with the previous researches (Spohn et al. 2016;Li et al. 2021a). The most likely explanation is that the C limitation restrained the increase amplitude of soil microbial CUE under N addition (as explained in detail below). In summary, the increase of N availability indeed improved the microbial CUE, but the limitation of available C was the key factor for it. Therefore, the relatively low increments of microbial CUE were induced by N addition.
We found that coniferous forest was more sensitive to N addition than evergreen broad-leaved forests. Although there was no significant difference in terms of the soil C:N ratios between the two forest types, the C:N ratios of the litter in the coniferous forest was greater than that in the broad-leaved forest. It can be inferred that the N limitation in the coniferous forest is more serious, thus, the microbial CUE improvement effect was more obvious after N addition.

The influencing factors of microbial CUE under N addition
We supposed that the relatively low pHs were responsible for the low average values of microbial CUE.
The pHs ranging from 4.4 to 6.5 with an average of 5.9, and some studies have shown that pH is positively correlated with microbial CUE in forest soil (Silva-Sánchez et al. 2019;Li et al. 2021b). Inorganic N fertilizer can result in soil acidification (Table 1) contributing to the reduction of soil base cations and the inhibition of microbial enzyme activities (Schleuss et al. 2019). Although soil acidification led to the relatively low microbial CUEs, our results of RDA showed that pH was not proportional to the microbial respiration and growth rate in this experiment. Ultimately led to inconsistent results with other studies, that is, pH value is not related to soil microbial CUE value. (Silva-Sánchez et al. 2019;Soares and Rousk 2019). The most likely reason is that the soil microbial CUE is more sensitive to N addition than pH in the study area. After all, the decrease of pH at the sites was caused by the addition of N . Moreover, the results of RDA analysis showed that soil pH had a great impact on soil MBC and MBN, and had a positive correlation with both of them. Chen and He (2004) believed that the soil had the highest MBC and MBN content at the original pH, and whether the pH increased or decreased, it will lead to a decrease in both of them, while N addition would lead to a decrease in soil pH, so we believe that N addition indirectly reduces soil MBC and MBN content. We found that SOC had a negative correlation with microbial growth (R 2 = -0.730, P < 0.01), resulting in an adverse impact on microbial CUE, which is consistent with the inference by Sinsabaugh et al. (2013). SOC is excessive before adding N fertilizer, which will lead to overflow respiration in microorganisms, then reduces microbial growth rate and microbial CUE (Manzoni et al. 2012). In addition, changes in soil C:N ratio also decreased microbial CUE (Sinsabaugh et al. 2013). Whether soil P has an effect on soil microbial CUE has been controversial. Widdig et al. (2020) came to a conclusion that soil P was not critical for microbial C cycling across a broad range of grassland sites, while Li et al. (2021a) believed that high availability of P will indirectly affect microbial process by reducing the energy and C investment for enzyme production. Some researchers think that the increase of availability of P can enhance soil microbial performance on the retrogradation of aromatic compounds (Manzoni et al. 2012;Chen et al. 2020). However, we think that soil P had no significant influence on microbial metabolism and CUE across the sites according to the results of our dataset analysis and the reasons will be discussed in detail later.
The soil C: N and C:P ratios within and across terrestrial ecosystems have a very large span and therefore microbial communities forced to adapt their foraging strategies to the available substrates by regulating the rates of respiration and growth. Despite high N fertilizer inputted, we found that soil C:N ratios and microbial biomass C:N ratios have no significant differences across all of the treatments, which is in accordance with the general observations that N addition have no influences on both of them (Cleveland and Liptzin 2007;Xu et al. 2013), indicating that microbial communities will adjust the rates of the processes of C and N cycling to make their internal biomass stoichiometry independent of the soil substrate stoichiometry. Surprisingly, long-term N additions had little change in soil total C, N, and Heuck et al. (2018) found that long-term N addition would significantly increase the activity of soil P (+ 180%), thereby increasing the rate of soil acid phosphatase production, which is one of the key microbial enzymes affecting microbial CUE. Therefore, we speculate that this is also one of the ways that N addition improves soil microbial CUE (Deng et al. 2017;Liu et al. 2021). Furthermore, soil C:P and N:P ratios were negatively correlated with microbial CUE, and C:N had no obvious relationship with microbial CUE. Overall, our results showed that the soil element imbalance of C and N relative to P has a significant impact on microbial CUE in the study area.
Therefore, combing the above two points of view, we think it is hard to find the relationships between individual elements and microbial C cycling, because the microorganisms regulated their community composition, metabolic processes and intracellular elements cycling to adapt the imbalanced stoichiometry, thereby resulting in a varying microbial CUE. The likely reason is that N addition causes changes in soil P availability and P cycling processes. Reasons for the changes in soil P availability under N addition include changes in pH, microbial extracellular enzymes, dissolved P uptake, and litter decomposition rates (Lu et al. 2012;Deng et al. 2017).
Moreover, C limitation is severe for microbial growth in our study area according to the theory of threshold (TER) elements ratio (especially C:N ratio), which was introduced to evaluate the C:N ratio of soil and limit the growth of microorganisms (Sterner and Elser 2002). We found that the average TER C: N (≈22.42) across the sites was slightly higher than the global average TER C:N (≈21), and the average soil C:N ratio in our study was 11.62, thus, the growth Table 1 Soil properties of the N addition and control treatments in two soil types "R": the effects of N-addition rate; "T": the effects of forest type; "R*T": the interaction of N-addition rate and forest type. Different lowercase letters indicate that significant differences in soil and soil microbial properties in the same forest type. "**" indicate that the soil microbial properties of different forest types have highly significant differences. "n.s" means there was no significant differences. Properties Evergreen broad-leaf forest of soil microorganisms were clearly C limited and it may constraint soil microbial CUE (Soong et al. 2020). We also found that the TER C:N was decreased in the three N addition treatments compared with the controls, so it can be speculated that the C limitation of microbial growth can be mitigated by N addition (Schleuss et al. 2019;Li et al. 2021a). Nevertheless, the mitigation effect stopped when N addition rate approached 20 kg· ha −1 ·yr −1 and once again, the C limitation became the major factor for the increase of CUE. In that case, we supposed that C limitation is the major factor for microbial C sequestration in the forests, which is similar to the condition in the current study area, therefore, the increase of C availability might be the most effective way to enhance microbial C sequestration.

Conclusions
In summary, we estimated soil microbial CUE in the two forest types based on a 13 C-glucose isotope tracing method and found that long-term N addition can inhibit microbial respiration and enhance microbial growth, resulting the increase of microbial CUE. Coniferous forest is more sensitive to N addition than evergreen broad-leaf forest due to the difference of the C:N ratios. Therefore, we speculate that N addition may promote soil C sequestration in forests, especially in coniferous forest. We further verified that it is not the individual element but the stoichiometry imbalance that regulates soil microbial C cycling, such as C:P and N:P ratios. Furthermore, we speculate that accelerating litter decomposition to increase the availability of SOC is a more effective way than N addition to increase soil microbial CUE, leading to an increase of soil C storage.