Effects of Stand Density on Soil Respiration and Soil Labile Organic Carbon and Their Inuence Mechanism in Larix Principis-Rupprechtii Plantations

Forest soil carbon pool plays a vital role in the global carbon sequestration and carbon emission. Forest management can regulate the sequestration and output of forest soil carbon pool to a certain extent, but mechanism of forest density effects on soil carbon pool still needs to be further researched. We established sample plots with density gradients in three-age stands of Larix principis-rupprechtii plantation and measured soil respiration (RS), soil organic carbon (SOC), soil dissolved organic carbon (DOC), and microbial biomass carbon (MBC), light fraction organic carbon (LFOC), and easily oxidizeable organic carbon (ROC). density is suggested to bridge the gaps in our comprehension of the Regulation of Forest Management on forest soil carbon pool. microbial respiration of the thinned plots compared to the control signicant


Introduction
Forest is the largest ecosystem which has tremendous amount of carbon sequestrated by plants (Franklin et al., 2009;Poorter et al., 2016;Khan et al., 2018). The soil carbon pool, which is the second largest carbon pool after aboveground forest carbon in the world, cannot be ignored (Sedjo, 1993;Lal, 2005). The carbon exchange between the soil and the atmosphere largely affects the global carbon cycle and climate change all the time (Dib et al., 2014;Tian et al., 2016;Gabriel et al., 2018). The soil respiration is the gas exchange between soil and atmosphere (Wei et al., 2010;Goldberg et al., 2017) and at the same time, soil organic carbon is also stored in forest soil in various forms and plays different roles (Tian et al., 2016;Jílková, 2020). Thus, understanding the characteristics of forest soil respiration and carbon sequestration is important for managing forest ecosystem. Through the proper management of forests, processes of greenhouse gas emission and soil organic carbon sequestration may be controlled to a certain extent.
The stand density, which is usually regulated by thinning, is considered as an indispensable in uence on the forest production. The principle of thinning maybe forming the local climax plant community as the basis of forest management research (Jens et al., 2000;Ming et al., 2018), which may be also based on the integrity principle of the forest ecosystem. Therefore, the effect of stand density is not only limited to the change of canopy structure and vertical forest structure (Jack and Long, 1991;, but also to the change of microclimate, interspeci c competition, and so on (Shao and Shugart, 1997 In the previous studies, it was considered that soil respiration (RS) is composed of autotrophic respiration (RA) and heterotrophic respiration (RH) whereby decomposition of microorganisms and the turnover of roots in the soil are the main forms of heterotrophic respiration and autotrophic respiration, respectively (Baggs, 2006;Hopkins et al., 2013;Xu and Shang, 2016). This process provides better understanding on the mechanism of soil respiration and the in uence of various factors on soil respiration. These studies, which were carried out on the forest soil within two years of thinning of the Carpinus betulus plantation, have reported signi cantly higher soil microbial respiration of the thinned plots compared to the control plots, and no signi cant difference was reported between soil respirations among the three thinning intensities (Akburak and Makineci, 2016). In the study of mature Masson pine forests, Leilei reported that thinning could effectively increase the rate of soil respiration in a short period (Lei et al., 2018). When studying the soil respiration of Shanxi Pinus tabulaeformis young forests that have been thinned, it is reported that moderate thinning positively affected soil respiration by changing soil temperature and humidity (Cheng et al., 2015). Similar results are also reported elsewhere (Zhang et al., 2018). Many researchers found that moderate thinning could make the stand microclimate changed, thereby optimizing the living environment of soil microorganisms, and reserve the su cient soil substrates and soil organic carbon for respiration. Few researchers have also reported different results, for example, thinning could cause the death of plant roots and reduce the autotrophic respiration of the soil and much sparse stands would reduce the activities of soil microorganisms (Park et al., 2009; Mosca et al., 2017).
Soil organic carbon (SOC) is the C contained in soil organic matter (SOM), which is an important indicator of measuring soil carbon sequestration (Lull et al., 2020), and this substantially affected by land use changes, forest management, natural and human interference, and so on. Forest soil organic Carbon is attributed to strong dynamic changes (Cambardella and Elliott, 1992;Liang et al., 1997). However, due to the complexity of the composition, structure, and existence of soil organic matter, the performance of a certain functional characteristic of the soil is often the result of a chemical mixture's simultaneous action with similar chemical components, structural characteristics, and functional groups. It is not the total amount of soil organic carbon that characterizes the soil carbon pool activity, but the soil labile organic carbon.The degree of activity is not only the total amount of soil organic carbon, but also the soil labile organic carbon. The active component of soil organic carbon is the most active and unstable C in the soil, which has the characteristics of availability, easy oxidation, and solubility. Generally, soil carbon may be grouped based on stability of SOC and measure soil dissolved organic carbon (DOC), such as microbial biomass carbon (MBC), light fraction organic carbon (LFOC), and easily oxidizeable organic carbon (ROC). This is considered as the most quantitative expression of soil labile organic carbon (LOC) (Hu et al., 2010).
It has been reported that the decrease of stand density increases soil temperature, humidity, soil respiration, soil SOC, N, P, K, etc. (Wic Baena et al., 2013;Zhang et al., 2018). However, many studies have found that the changes of soil temperature, humidity, and soil respiration by thinning could become stable after certain period of thinning, and even return to the original level (Fernandez et al., 2012;Olajuyigbe et al., 2012;Bai et al., 2016). These studies also reported that the root system or the plant remaining in the stand after thinning acted as the exogenous carbon, which led to the change of soil rather than the effect of stand density. Most of the existing studies focus on the short-term "stress response" of forest soil carbon pool after thinning (Ryu et al., 2009;Bolat, 2013;Zhao et al., 2019). Although thinning directly changes the stand density, the impact of stand density on the soil should be the steady state achieved by biochemical action under the in uence of different stand densities. The effect of stand density on the soil carbon pool is caused by the interaction of litter return, root growth and respiration, microbial activity, and mineral turnover after thinning. Our study intends to explore more on the effect of different stand density on the soil respiration and soil labile organic carbon and their mechanism. This is not clear whether the soil carbon pool of different age plantations would have different responses to different stand densities.
This study considers paying enough attention to the factors of forest development and growth in carrying out more effective human intervention and forest management on the forests of all ages. As the substrate of soil respiration, soil organic carbon cannot not be discussed in isolation, the hypothesis presented in this study was that both stand density and stand age are essential factors in uencing soil respiration, soil organic C and soil labile organic C, and a close correlation between soil labile organic C and soil respiration. Therefore, objectives of this study were to: 1) Examine whether RS, RH, RA and SOC, MBC, DOC, LFOC, ROC in mineral soil would be signi cantly affected by stand density and stand age during the studied period (Five years after thinning); 2)Find out the trend of the variables mentioned above with the stand density, and compare the trends for different stand ages; 3 Identify the crucial factors shaping the RS, RH and RA in soil organic carbon pools.

Study area
The study area is located in Yeshagou (37 ° 44 ′ N, 111 ° 30 ′ E) of Xiaowenshan forest farm in Pangquangou Nature Reserve, Lvliang City, Shanxi Province, China, with an altitude of 1760-2210 M (Fig. 1). The climate of this area is temperate continental monsoon climate, cold and dry in winter and hot and humid in summer. The annual average temperature is 4.2 ℃, the average yearly precipitation is 822.6 mm, and the annual average relative humidity is 70.9%. The soil is leached cinnamon soil with a humus layer thickness of 3-7 cm. In the study area, the forest types were mainly pure Larix principis-rupprechtii forest and Pinus tabulaeformis plantation, occasionally accompanied by a few Betula platyphylla and Quercus liaotungensis; the main shrub species are Spiraea saliifolia, Rosa xanthina, and Lespedeza bicolor.
Toward the end of April 2020, we selected sub-compartments with similar site conditions, slopes, and slope positions of different ages, with an area of more than 2 ha, according to the average height of dominant trees (Larix principis-rupprechtii), and the age of sub-compartments were 27-(27a), 36-(36a), and 48-year-old (48a), respectively. In order to prevent the human interference and shortterm effect after thinning, the latest thinning operation was conducted sub-compartment on 2015 and another thinning on the same area on 2020. Based on the stand density difference formed by thinning, nine standard sample plots (20 m x 20 m) with different stand densities were laid out in the stands with different ages (stand density calculated according to the actual density on the sample plot), and a 5 m wide buffer zone was set around the sample plots. A total of 27 sample plots were laid out with inter-plot distance limited to 40 m. We set three sample plots with similar stand densities as replications, dividing the nine sample plots of each age into three density levels, namely high density (HD), medium density (MD), and low density (LD). We measured the trees with DBH greater than 5 cm in each plot, and record their DBH, height, and other parameters.

Field survey and soil sampling
When laying out the sample plot, we randomly placed six polyvinyl chloride (PVC) material rings with inner diameters of 20 cm and heights of 10 cm wihin the sample plot, and driven them into the soil by about 6 cm with a hammer. Three of them were treated with root removal (trench method) to distinguish the autotrophic breathing from the heterotrophic breathing. The root removal method involved selecting an area of size 0.5 m × 0.5 m and digging trenches vertically to a depth of 0.6-0.8 m around this area with a shovel (up to the depth of the root system accessed), cutting off the roots (but not removing) and inserting the nylon net with 100 mesh for preventing root growth (Kuzyakov, 2006). The vegetation of the sample plot was clipped and all the trench-digging sample plots were regularly cleaned for surface vegetation and litter to distinguish autotrophic respiration and heterotrophic respiration.
In order to avoid the in uence of soil disturbance on the soil respiration, soil respiration was measured twice in mid-July 2020. Soil carbon ux automatic measurement system (Li-cor 8100a, Li-COR, Inc., Lincoln, Nebraska) was applied to measure soil respiration three times on each respiratory ring. The measurement was made between 9:00 am to 15:00 whereby the soil respiration of the same age and different density was also measured in the same period. A total 972 measurements of 162 respiratory rings were made. The soil was sampled at the depth of 0-10 cm, 10-20 cm, and 20-30 cm by using a 5 cm circular soil auger. Three samples from each soil layer were collected, which resulted in 9 soil samples per sample plot, and 243 soil samples in total. The collected soil samples were put into the numbered aseptic soil bags and stored at low temperature, and brought back to the laboratory. The ne roots and stones were removed from the soil samples, sieved through 2 mm holes, and divided into two parts. One sample was fresh and stored in a 4 ℃ refrigerator to determine MBC and DOC; other samples collected were dried in the laboratory to determine soil physical and chemical properties, and other labile organic carbon components (SOC, ROC, LFOC). The summary characteristics of the sample plots are presented in Table 1. Table 1 The summary characteristics of the sample plots. SWC, soil moisture content; BD, soil bulk density; TN, soil total nitrogen content; TP, soil total phosphorus content. Values are as the mean ± standard deviation (SD

Data processing
We calculated RA as below: where RA, autotrophic respiration; RS, soil respiration; RH, heterotrophic respiration (obtained by trench method).
We calculated MBC as below: where CU and NU are the carbon and nitrogen contents of fumigated soil extract, K is the conversion coe cient of fumigated extraction method (0.45) (Brookes et al., 1985) Based on the testing normality and consistency of the data, the statistical analysis was carried out the multi-factor analysis of variance, linear and quadratic regressions, and structural equation modeling, correlation analysis, and redundancy analysis using , and all the illustrations in the article were made using this software version.

Factors affecting SOC, soil labile organic carbon, and soil respiration
We used the multi-factor analysis of variance to test the factors that may affect RS, SOC, and labile organic carbon. Stand density signi cantly affected RS, RH, and RA (  Analysis results of factors affecting RS, RH, RA and soil labile organic carbon. In the table * represents statistical signi cance, in which, P < 0.1;* P < 0.05; ** P < 0.01; *** P < 0.001. RS, soil respiration; RH, heterotrophic respiration; RA, autotrophic respiration.

Effects of stand density on soil respiration
We analyzed the differences of RS, RH, RA between different density levels. We use the advantages of density sequence diagrams to discover the trend of these indicators with stand density and try to nd the mechanism of stand density on soil respiration.

Difference of RS, RH, RA of different stand density levels
For the 27a stand, the RH of MD was signi cantly higher than that of HD and slightly higher than LD; the RS of MD was slightly higher than that of LD and HD; the RA of HD was the highest, followed by MD and LD was the lowest (Fig. 2A). For the 36a stand, the RH of MD was signi cantly higher than that of HD and slightly higher than LD; the RS of MD was signi cantly higher than the other two grades; there was no signi cant difference in RA (Fig. 2B). For the 48a stand, the RS of MD was signi cantly higher than HD and LD, RH was slightly higher than LD and higher than HD, RA of MD was signi cantly higher than LD and HD (Fig. 2C).

Variation of RS, RH, RA for stands with different ages with stand density levels
We attempted to characterize the variations of RS, RH, and RA for the stands with stand density and ages using regression analyses (Fig. 3). The quadratic function was used to t the changing trend of RS, RH, and RA at different ages, and this function tted the data well, describing stand density variations for RS and RH by more than 50% (R 2 > 0.5). The older the stand age, the higher was the RS and RH rate. However, when the stand density was too large, the RS and RH of the three age stands were the same. The model curve of the quadratic function seems to be the highest at the same time, maintaining higher RS and RH levels for the 48a stand, followed by the 36a stand. This indicated that the older stand was more sensitive to stand density than what?. In terms of autotrophic respiration, except for stand with 48 years, the autotrophic respiration of 36a and 27a stands showed a clear upward trend with an increase of stand density.
We tried to explore more about the mechanism of stand density effects on RS after the effect of stand density on different ages of Larix principis-rupprechtii plantation was clari ed. Therefore, we used R 4.0.3 Lavan package to establish the structural equation model (SEM) for the four observed variables of three different aged-stands. It was found that Lavaan normally ended after 60 iterations, and the model's p-value (p = 0.653, χ 2 = 1.628,CFI = 1.000) was greater than 0.05, so the structural equation model was acceptable (Table 3). We found that the in uence of forest density of Larix principis-rupprechtii plantation on soil respiration at different ages mainly came from the regulation of heterotrophic respiration, and a small part came from the regulation of autotrophic respiration. The direct effect of heterotrophic respiration on soil respiration was 0.84, autotrophic respiration was 0.64. stand, MBC of MD was slightly higher than that of LD and signi cantly higher than HD; DOC had no signi cant difference; LFOC of MD and LD was signi cantly higher than HD; ROC had no signi cant difference, but ROC of MD was lower than LD and HD. In the 36a stand, the MBC of MD was signi cantly higher than that of LD and HD; there was no signi cant difference in DOC; the LFOC of LD and MD was signi cantly higher than that of HD; there was no signi cant difference in ROC, but the ROC of MD was lower than that of LD and HD. For the 48a stand, the MBC of MD was signi cantly higher than that of HD and slightly higher than that of LD; the DOC of LD was slightly higher than that of MD and signi cantly higher than HD. 3.3.2 The variation of SOC and labile organic carbon with stand density at different stand ages The SOC of different ages showed a downward trend with stand density changes and could be adequately described by linear regression. Although the SOC of each age stand with the increase of stand density, SOC remained high for 48 years, followed by 36 years and the lowest at 27 years (Fig. 5A). The DOC did not show an obvious trend with stand density (Fig. 5B). The LFOC of three age stands rst increased and then decreased with the increase of stand density, and the trend of the 48a stand was the most obvious, and the changes for s36a and 27a stands were not signi cant (Fig. 5C). The trend of soil ROC changes was opposite to that of LFOC, and the trend of soil ROC of three aged-stands showed the rst decreasing trend and then increasing trend (Fig. 5D). Soil MBC rst increased and then decreased with an increase of the stand density. The MBC content of 48a stands reduced to 400 mg/kg when the density was higher than 1300 trees/ha, which was lower than 36a stand with similar density (Fig. 5E).
Based on these trends, we could conclude that although the general trend of SOC and soil labile organic carbon affected by density was the same in different age stands, there were some differences in their sensitivity to the density changes. For the 48a stand, density change response is the strongest, as this was especially re ected in the LFOC, ROC, and MBC. For the 27a stand, response was weaker than that for 36a and 48a, and the range of variation with stand density was rather small.

The correlations between soil organic carbon and labile organic carbon and soil respiration
We further explored whether there was a certain correlation between soil respiration and soil labile organic carbon under the background of stand density regulation (Fig. 6). The correlation analysis showed that RS was signi cantly negatively correlated with

Explanation of soil organic carbon pool to the soil respiration variability
After understanding the correlations between SOC, soil labile organic carbon and soil respiration, we carried out the redundant analyses of soil respiration and soil organic carbon using SOC and soil labile organic carbon to explain the variation of soil respiration rate. The axes 1 and 2 described 99.99% variations of the variable of interest, indicating well representation of the original explanatory variable. The RDA1 and RDA2 contained almost all the information of the axes and jointly explained 56.05% of the variance of RS, Rh, and RA (Fig. 7). See Table 5 for detailed parameter estimates. The RS and RH in the different density levels show signi cant differences (Fig. 2). The RS and RH of MD of the three aged-stands were higher than LD and HD, and the RA of the 48a stand was signi cantly higher than LD and HD, but there was no signi cant difference in RA of the other two age stands. We consider that for the three stand ages, moderate stand density promotes RS and RH, combine the trends of RS, RH, and RA to give a reasonable explanation.
The RS and RH rst increase and then decrease with an increase of the stand density. This trend can be well tted by the quadratic function ( Figure ?), which was also reported in the previous studies, and coincided with the idea of Intermediate-Disturbance Hypothesis (Connell, 1979 what we found in our study. The only difference is that RS and RH rst increase and then decrease with the stand density, and they could not keep a high level when the density was too low. We assume that there may be two reasons, such as under the effect of long-term low density, the soil microenvironment created by low stand density and upper shading could not be suitable for soil microbial activities (Yang et al., 2017). A litter with low stand density may not be suitable for soil microbial activities, and the amount returned was less than that of high density, which resulted in the limitation of organic carbon components as respiratory substrates in some soil carbon pools. RA's trend for 48a stand with stand density is similar to RS and RH, and RA shows an upward trend with stand density for 36a and 27a stands. (Fig. 3C). we consider that there is a threshold for the capacity of the soil to hold plant roots.
An increase in the root density within the threshold will cause an increase in the soil autotrophic respiration. This is the reason for the increase in soil autotrophic respiration for 27a and 36a stands, while in 48a plantations, too high forest stand density causes root density to exceed this threshold. Plant roots and root microbes compete ercely (Grubb, 2000), even leading to reduced root microbial activity, root death, and even plant death (Kuzyakov and Larionova, 2005), so for 48a stand, soil RA will rst increases and then decrease with an increasing stand density.
The trends appeared in RS, RH, and RA were signi cantly age-related (Fig. 3), and was fastest in the 48-year-old stand, followed by the You et al., 2014), Therefore, the difference of soil respiration and heterotrophic respiration rate in stand age and the increase with age were caused.
The regulation mechanism of stand density on RS is that the stand density mainly regulates soil respiration by changing RH and then by regulating RA, which is empirically proved by the structural equation model (Figure ?). The change of RA and its effect on soil respiration are weaker than heterotrophic respiration (Bond-Lamberty et al., 2004). The reason may be that soil organic carbon status and Soil Microenvironment of different stand densities are different. In other words, from the perspective of soil respiration mechanism, it is the substrate state, process enzyme activity, temperature, and humidity may have differences, which are also the major factors affecting RH.

The effect of stand density on SOC and labile organic carbon of different age stands
Many studies have con rmed that soil depth substantially affects organic carbon and soil nutrients, so it does not appear as the discussion content of this study (Baldrian et al., 2012). We can see that in the three age stands, the differences in soil labile organic carbon at different stand density levels are mainly re ected in three indicators: MBC, LFOC, and ROC. Among them, the MBC and LFOC of MD are signi cantly higher than LD and HD. ROC is lower than LD and HD. In other words, moderate stand density promotes the accumulation of MBC and LFOC and inhibits the accumulation of ROC. We know that SOC could increase with the decrease of stand density to a certain extent. Although the reduction of stand density increases the loss of litter (Lim et al., 2012; Lull et al., 2020), the appropriate soil environment increases the biochemical process rate, thus promoting the increase of SOC, which is consistent with the conclusion of this study. It is not di cult to nd the reason from the composition of LFOC and MBC. The composition of LFOC is usually the soil organic matter with soil particle density less than 20 g/cm 3 (Wu et al., 2003), and the main components are animal and plant residues, mycelium, spore, monosaccharide, polysaccharide, and hemilignin (Christensen, 1992;B. et al., 2001). The MBC refers to the total C content of living bacteria, fungi, algae and soil microorganisms in the soil (Wander et al., 1994). When the stand density was moderate, the soil microenvironment was more suitable for microbial activities, and the activity and quantity of soil microorganisms and the intermediate products of microbial activities increased. Therefore, the contents of LFOC and MBC were the highest when the density was moderate, showing a trend of increasing rst and then decreasing and could be well tted by the quadratic function.Doc has no obvious trend with stand density. The DOC is the part of organic carbon in the soil that can be dissolved in water by extraction, including simple organic molecules, such as carbohydrates, simple amino acids, and small molecular proteins (Bolan et al., 2008). The DOC has strong mobility, which will be stronger in the rainy summer, so it would not be affected by strong stand density. The reason for the change of ROC has been explained in the following discussion combined with soil respiration.
According microbial activity is promoted. Therefore, the content of SOC and soil labile organic carbon will increase with age.

Sensitivity of soil carbon pool indicators of forest stands of different ages to density changes
Soil respiration, SOC, and soil labile organic carbon are all affected by forest density to varying degrees (Table or Figure ?). However, the sensitivity of forests of different ages to density is different, especially in the three indicators of ROC, LFOC, and MBC. From the perspective of the change range and the tting equation, the change range of the 48a forest stands affected by density is obviously larger than that of 36a and 27a, and 36a ranks second. We consider that the older the plantation, the greater the importance of individual trees in the sample plot. This is re ected in the larger crown width, root area, etc. The death or felling of each tree in the sample plot will have a more signi cant impact on the soil and upper shading, and the density change has a more substantial effect on the forest soil microenvironment. Therefore, soil respiration and labile organic carbon will show strong sensitivity.

The relationship between soil labile organic carbon and soil respiration
Through the correlation analysis, we can clearly see that RS, Rh and MBC, LFOC, and ROC showed a strong correlation, which with LFOC, MBC showed a positive correlation, and ROC showed a negative correlation (Fig. 6). However, the correlation between RA and soil labile organic carbon for the 48a stands did not have much practical meaning. The trend of RA was largely determined by stand density and root density.
Soil readily oxidized organic carbon (ROC) is rapidly oxidized and decomposed by microorganisms and soil enzymes. We can understand that LFOC is similar to the "warehouse" of soil respiration. MBC is more like a "tool" of soil respiration. It is a feature formed by the long-term accumulation of different stand densities, and ROC is more like "raw material," which is an immediate feature. We consider that ROC will decompose in time at a higher soil respiration rate and stay in dynamic stability with lower content.
In comparison, it is at a higher content dynamic stability when the soil respiration rate is lower. Therefore, ROC is negatively correlated with RS and RH, which also explains the trend that ROC rst decreases and then increases with stand density. Because more powerful "tools should support the high rate of soil respiration," LFOC, MBC and RS, RH show a signi cant positive correlation. This allows us to link the relationship between labile organic carbon and soil respiration. The results of RDA also proved this point well. We found that SOC and soil labile organic carbon have a high degree of explanation (56.05%) for the variance variation of RS, RH, and RA.
On the one hand, stand density affects the soil microenvironment, such as temperature and humidity (Table or Figure?). On the other hand, it regulates the state and content of labile organic carbon in soil carbon pool and affects soil respiration from another aspect, that is, stand density soil labile organic carbon soil respiration, which is also an important path of stand density regulating soil respiration LFOC, MBC and ROC in carbon components are responsible for "warehouse" and "tools" and "raw materials." We speculate that the enzymes related to the soil C cycle play the role of "controlling the speed," and then speculate whether the existing RS model can be improved by using the characteristics of labile organic carbon (respiration substrate characteristics) and enzyme activity (rate characterization), so as to make the model more ecological and scienti c, which will be the focus of our subsequent study.

Proposal of the optimal density for each age of Larix principisrupprechtii plantation
Although this is a statement with substantial limitations, it is an essential reference for regional forest management.
Taking afforestation density of 3300 plants/ha as an example, the forest stand for 27 years should maintain about 1650 plants/ha (50%), the forest stand for 36 years should be about 1250 plants/ha (38%), and the forest stand for 48 years should be maintained at 900 plants/ha (27%), so that both the above-ground vegetation carbon pool and the soil carbon storage carbon sequestration can be maintained at a high level, while the soil labile organic carbon content is at a high level and stable, the soil microbial activity is high, and the soil quality is good. This is what a forest manager wants to see.

Conclusion
Among the forest stands of three ages, RS and RH of different stand density levels was signi cantly different, MD was higher than LD and HD. Moderate density promotes RS rate and RH rate; RS and RH of three ages stand to increase with stand density and then decrease. The quadratic function can better t this trend. Except for 48 years, RA increases gently with the increase of stand density; stand density affects both RH and RA, but it mainly affects RS by regulating RH.
Among the forest stands of three ages, the MBC, LFOC, and ROC (0-30cm) of different density levels are signi cantly different. The MBC and LFOC of MD are higher than LD and HD, and the ROC of MD is lower than LD and HD; moderate Density promotes the sequestration of MBC and LFOC, and inhibits the sequestration of ROC. With the increase of forest density, LFOC and MBC rst increased and then decreased, and ROC rst decreased and then increased, the quadratic function can t these changing trends.
The RS, RH, and RA rates of older forest stands are relatively fast, and the contents of SOC, MBC, LFOC, DOC, and ROC are higher, and they are more sensitive to changes in stand density.

Declaration of Competing Interest
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper. concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.  Structural equation model of the in uence mechanism of stand density on RH, RA, and RS. Stand density 2 is the square of stand density; all lines are direct effects, the red line is a negative effect, and the green line is a positive effect.