Contribution of Fine Roots to Soil Organic Carbon Accumulation in Different Desert Communities in the Sangong River Basin

This study explored the relationship between soil organic carbon (SOC) and root distribution, with the aim of evaluating the carbon stocks and sequestration potential under five plant communities (Alhagi sparsifolia, Tamarix ramosissima, Reaumuria soongorica, Haloxylon ammodendron, and Phragmites communis) in an arid region, the Sangong River watershed desert ecosystem. Root biomass, ecological factors, and SOC in different layers of a 0–100 cm soil profile were investigated. The results demonstrated that almost all living fine root biomass (11.78–34.41 g/m2) and dead fine root biomass (5.64–15.45 g/m2) levels were highest in the 10–20 cm layer, except for the P. communis community, which showed the highest living and dead fine root biomass at a depth of 60–70 cm. Fine root biomass showed strong seasonal dynamics in the five communities from June to October. The biomass levels of the A. sparsifolia (138.31 g/m2) and H. ammodendron (229.73 g/m2) communities were highest in August, whereas those of the T. ramosissima (87.76 g/m2), R. soongorica (66.29 g/m2), and P. communis (148.31 g/m2) communities were highest in September. The SOC of the five communities displayed strong changes with increasing soil depth. The mean SOC value across all five communities was 77.36% at 0–30 cm. The highest SOC values of the A. sparsifolia (3.08 g/kg), T. ramosissima (2.35 g/kg), and R. soongorica (2.34 g/kg) communities were found in June, and the highest value of the H. ammodendron (2.25 and 2.31 g/kg, p > 0.05) community was found in June and September. The highest SOC values of the P. communis (1.88 g/kg) community were found in July. Fine root production and turnover rate were 50.67–486.92 g/m2/year and 1.25–1.98 times per year. The relationships among SOC, fine root biomass, and ecological factors (soil water content and soil bulk density) were significant for all five communities. Based on the results, higher soil water content and soil conductivity favored the decomposition of root litter and increased fine root turnover, thereby facilitating SOC formation. Higher pH and bulk density levels are not conducive to soil biological activity and SOC mineralization, leading to increased SOC levels in desert regions.


Introduction
The soil carbon pool is one of the key carbon pools in terrestrial ecosystems and plays an essential role in the global carbon cycle [1,2]. Arid and semi-arid areas account for more than 40% of the terrestrial surface of Earth [3,4] and contain nearly 10% of the global soil organic carbon (SOC) stock [5]. Arid and semi-arid areas are extremely fragile ecosystems and are particularly sensitive to global change in the terrestrial ecosystems, and are characterized by their vulnerability to external environmental influences such as extreme weather, erratic precipitation, frequent droughts, and anthropogenic destruction [4,6]. Consequently, slight changes can have doubly significant impacts on the regional or global carbon cycle [7]. Although arid and semi-arid regions are important sources and sinks of carbon, long-term studies of SOC have focused on forests, grasslands, and croplands some of the mechanisms remain unclear. Therefore, we investigated the biomass and turnover of fine roots in desert communities, explored the contribution of fine roots to the vertical distribution of SOC in deeper soil depths (0-100 cm), and analyzed the effect of soil factors on SOC in five plant communities in depth. This study aimed to evaluate the contribution of fine root biomass to carbon stocks and sequestration potential in desert plant communities and to provide a theoretical basis and data support for research on soil carbon in arid and semi-arid ecosystems.

Study Area
The present research was conducted at the Fukang Station of Desert Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences (43 • 45 -45 • 29 N 87 • 43 -88 • 44 E), at the southern edge of the Junggar Basin ( Figure 1). The region has a typical temperate continental arid climate, with a mean annual rainfall of 160 mm and a mean annual temperature of 6.6 • C. The natural vegetation is dominated by Reaumuria soongorica (Pall.) Maxim, Haloxylon ammodendron (C. A. Meyer) Bunge, Tamarix ramosissima Ledeb, Alhagi sparsifolia Shap, and Phragmites communis (Cav.) Trin. ex Steud. The soil type is gray desert salt-alkali soil and dry sandy loam soil. Five plant communities were selected for analysis, based on the dominant natural vegetation. Three randomly selected 25 × 25 m experimental plots were established in each of the five communities, resulting in a total of 15 plots (Table 1). turnover [27,31,33] In this study, the plant communities (Alhagi sparsifolia, Tamarix ramosissima, Reaumuria soongorica, Haloxylon ammodendron, and Phragmites communis) were the dominant species in the desert area. Although the contribution of fine roots to SOC is greater in desert plant communities, some of the mechanisms remain unclear. Therefore, we investigated the biomass and turnover of fine roots in desert communities, explored the contribution of fine roots to the vertical distribution of SOC in deeper soil depths (0-100 cm), and analyzed the effect of soil factors on SOC in five plant communities in depth. This study aimed to evaluate the contribution of fine root biomass to carbon stocks and sequestration potential in desert plant communities and to provide a theoretical basis and data support for research on soil carbon in arid and semi-arid ecosystems.

Study Area
The present research was conducted at the Fukang Station of Desert Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences (43°45′-45°29′ N 87°43′-88°44′ E), at the southern edge of the Junggar Basin ( Figure 1). The region has a typical temperate continental arid climate, with a mean annual rainfall of 160 mm and a mean annual temperature of 6.6 °C. The natural vegetation is dominated by Reaumuria soongorica (Pall.) Maxim, Haloxylon ammodendron (C. A. Meyer) Bunge, Tamarix ramosissima Ledeb, Alhagi sparsifolia Shap, and Phragmites communis (Cav.) Trin. ex Steud. The soil type is gray desert salt-alkali soil and dry sandy loam soil. Five plant communities were selected for analysis, based on the dominant natural vegetation. Three randomly selected 25 × 25 m experimental plots were established in each of the five communities, resulting in a total of 15 plots (Table 1).   Location of the study area and experimental sites (Haloxlon ammodendron, Phragmites communis, Tamarix ramosissima, Alhagi sparsifolia, and Reamuria soongorica communities). Each of the five communities was used to investigate the seasonality of live fine root biomass and turnover rate in each month from May to October 2010. Fine root production from May to October was used to represent the annual fine root production. We considered the average distance (D) between two dominant plants as the standard, and samples were collected in three quadrats, i.e., near the plant, at 1/4 D, and at 1/2 D. Each fine root sample was taken from three plots of the freshly cut profile walls by collecting a 50 × 50 cm soil core. The soil core samples were divided into 10 soil depths of 0-10, 10-20, 20-30, 30-40, 40-50 and 50-60, 60-70, 70-80, 80-90, and 90-100 cm until few fine roots were found. The collected fine root samples were sealed in plastic bags and transported to the laboratory. In the laboratory, roots with a diameter more than 2 mm were discarded, and live and dead fine roots less than 2 mm were manually washed and separated based on their color and luster, resilience, consistency, smell, degree of cohesion between the cortex and the periderm, and the appearance of the phloem. The separated fine roots were stored on the basis depth interval and oven-dried to a constant mass at 65 • C for determination of the mass and for chemical analysis.
Fine root production and turnover were estimated and calculated on the basis of the maximum-minimum method, specifically, based on the following equations: [20,34]: where M, D, P, and T indicate the annual fine root mortality, decomposition, production, and turnover, respectively. Mmax and Mmin are the maximum and minimum values of dead fine root biomass, whereas Pmax, Pmin, and Pa are the maximum, minimum, and average values of live fine root biomass during the entire growing season.

Measurement of Fine Root Decomposition
A decomposition experiment with fine roots was established in each of the five communities in 2010, which were collected from the top 100 cm of the soil in each of the five communities by sieving roots during the 2009 growth season. Nylon mesh bags (15 × 20 cm and 0.12 mm mesh size) were used for determining root decomposition [35]. Approximately 5 g of fine-root samples and 500 g of the soil in which the fine roots had been sampled were placed in the five ecosystems at about a depth of 20-30 cm in the soil horizon in early May 2010. Bags were deployed randomly in the five ecosystems using the coordinates of permanently marked grids. Three or more bags from each ecosystem were collected once a month. Upon collection, bags were immediately returned to the laboratory, the fine soil particles were removed, and fine roots were removed from bags, dried at 65 • C to a constant mass, and weighed. The fine root samples were ground and passed through a 100-mesh sieve for subsequent analyses.
The fine root exponential decay function (4) was used to fit the curve of the percentage of mass remaining, and the time of decomposition incubation and annual fine root decomposition were calculated according to the fitted functions [36], using the following equation: where W 0 is the initial dry weight, W t is the dry weight remaining after the decomposition interval, k is the decomposition coefficient, and t is the decomposition interval (days).

Soil Physical and Chemical Properties
Soil samples were collected in similar places to the fine root samples at a depth of 100 cm at seven intervals of 0-5, 5-10, 10-20, 20-30, 30-40, 40-50 and 50-100 cm with a cutting ring. The soil samples were naturally dried and sieved through a 2 mm mesh size in order to remove plant materials, small insects, and stones. Soil pH (1:5 solid-to-water (w/v)) and soil electrical conductivity (SEC) (1:5 solid-to-water (w/v)) were determined in deionized water with a Eutech PC700 pH/EC meter (Thermo Fisher Scientific Inc., Waltham, MA, USA). Soil bulk density (SBD) was determined for each depth interval using a cutting ring and calculated as the ratio of oven-dry soil weight to cutting ring volume. Soil water content was determined by drying for 3 days at 105 • C. The rest of the soil samples were air-dried and ground to pass through a 100-mesh sieve for subsequent analyses.
Soil bulk density (SBD, g/cm 3 ) and soil water content (SWC, %) were calculated using Equations (5) and (6), respectively: where M S is the dry soil weight (g), V t is the total volume of the soil core (cm 3 ), and M W is the fresh soil weight (g).

Measurement of SOC and Fine Root Organic Carbon
Soil carbon and fine root carbon were determined by a wet oxidation technique [37]. The fine root carbon input was determined using Equation (7): where C input is the fine root organic carbon input to soil by decomposition (g/m 2 /a) and FOC is the fine root organic carbon content (%). An asymptotic nonlinear model was used to describe the SOC stock's depth distribution, using the following Equation (8) [18,38,39]: where y is the cumulative SOC content from the soil surface to a certain soil depth d in centimeters, and the β value describes the shape of the vertical distribution of SOC within the whole sampled soil profile. High values of β correspond to a larger proportion of SOC at greater soil depths.

Data and Statistical Analysis
All calculations were performed using SPSS 16.0 software (SPSS Inc., Chicago, IL, USA), including ANOVA and homogeneity of variance tests, multiple comparisons, and regression analyses. The differences in the fine root productivity, decomposition, decomposition of supplementary soil organic carbon (SOC), and turnover rate of the five plant communities were compared by the LSD test. The dependence of SOC on all variables and soil depths were analyzed further by regression analysis. Mean differences were considered significant at p < 0.05. SigmaPlot for Windows Version 10.0 was used for graphics.

Fine Root Vertical Distribution and Seasonal Dynamics
The vertical distribution of living fine root biomass first increased and then decreased with soil depth in the five communities. However, in the A. sparsifolia community, it only decreased with soil depth. Living fine root biomass was highest in the 10-20 cm soil profile, except for the A. sparsifolia and P. communis communities ( Figure 2). The highest living fine root biomass at 10-20 cm was 11.78-34.41 g/m 2 for the T. ramosissima, R. soongorica, and H. ammodendron communities, whereas the highest living fine root biomass at 0-10 cm and 60-70 cm was 19.25 and 54.79 g/m 2 for A. sparsifolia and P. communis, respectively. The fine root biomass was mainly concentrated at a depth of 0-30 cm. The living fine root biomass at a depth of 0-30 cm accounted for 63.86-86.0% of the total living fine root biomass for the A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron communities, but living fine root biomass at 40-90 cm accounted for 84.65% of the total living fine root biomass for the P. communis community.
with soil depth. Dead fine root biomass was highest at 10-20 cm, except for the P. communis community ( Figure 2). The highest dead fine root biomass at 10-20 cm was 5.64-15.45 g/m 2 for the A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron communities. The highest dead fine root biomass at 60-70 cm was 15.89 g/m 2 for the P. communis community. The dead fine root biomass at 0-30 cm accounted for 73.04-85.03% of the total dead fine root biomass for the A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron communities, and the dead fine root biomass at 40-90 cm accounted for 84.65% of the total dead fine root biomass for the P. communis community.  The vertical distribution of dead fine root biomass first increased and then decreased with soil depth. Dead fine root biomass was highest at 10-20 cm, except for the P. communis community ( Figure 2). The highest dead fine root biomass at 10-20 cm was 5.64-15.45 g/m 2 for the A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron communities. The highest dead fine root biomass at 60-70 cm was 15.89 g/m 2 for the P. communis community. The dead fine root biomass at 0-30 cm accounted for 73.04-85.03% of the total dead fine root biomass for the A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron communities, and the dead fine root biomass at 40-90 cm accounted for 84.65% of the total dead fine root biomass for the P. communis community.
From June to October, the living and dead fine root biomass of the five communities showed strong seasonal dynamics, increasing and then decreasing ( Figure 3). Living fine roots in five communities peaked in August or September (48.45-405.75 g/m 2 ), with the lowest values found in June (31.40-255.73 g/m 2 ). The maximum value of living fine root biomass (405.75 g/m 2 ) occurred in September for P. communis, and the minimum value of living fine root biomass (31.40 g/m 2 ) occurred in June for R. soongorica. Dead fine root biomass in the five communities peaked in June, July, or August, with the maximum value of dead fine root biomass (23.48 g/m 2 ) found in July for H. ammodendron; the minimum value was found in June, September, and October; and the minimum value for T. ramosissima (1.72 g/m 2 ) found in June. lowest values found in June (31.40-255.73 g/m 2 ). The maximum value of living fine root biomass (405.75 g/m 2 ) occurred in September for P. communis, and the minimum value of living fine root biomass (31.40 g/m 2 ) occurred in June for R. soongorica. Dead fine root biomass in the five communities peaked in June, July, or August, with the maximum value of dead fine root biomass (23.48 g/m 2 ) found in July for H. ammodendron; the minimum value was found in June, September, and October; and the minimum value for T. ramosissima (1.72 g/m 2 ) found in June.

Fine Root Biomass and Production
Fine root biomass showed strong seasonal dynamics in the five communities from June to October (Figure 4). The highest fine root biomass was found for the P. communis community and the lowest was for the R. soongorica community. The monthly mean fine root biomass was 54.19-438.89 g/m 2 for the five communities. The seasonal variation in biomass was similar between the A. sparsifolia and H. ammodendron communities, with an initial increase followed by a decrease; the highest values of 138.31 and 229.73 g/m 2 , respectively, were measured in August. The seasonal variation in biomass was similar among the T. ramosissima, R. soongorica, and P. communis communities; biomass first increased and then decreased, with the highest values of 66.29-523.41 g/m 2 being measured in September.

Fine Root Biomass and Production
Fine root biomass showed strong seasonal dynamics in the five communities from June to October (Figure 4). The highest fine root biomass was found for the P. communis community and the lowest was for the R. soongorica community. The monthly mean fine root biomass was 54.19-438.89 g/m 2 for the five communities. The seasonal variation in biomass was similar between the A. sparsifolia and H. ammodendron communities, with an initial increase followed by a decrease; the highest values of 138.31 and 229.73 g/m 2 , respectively, were measured in August. The seasonal variation in biomass was similar among the T. ramosissima, R. soongorica, and P. communis communities; biomass first increased and then decreased, with the highest values of 66.29-523.41 g/m 2 being measured in September.

SOC Vertical Distribution and Seasonal Variations
The SOC of the five communities changed greatly with increasing soil depth ( Figure  5). The A. sparsifolia community showed the highest SOC content in all soil layers compared with the other four communities. In addition to the 30-40 cm soil layer, the P. communis community in all soil layers had the lowest SOC content compared with the other four communities.

SOC Vertical Distribution and Seasonal Variations
The SOC of the five communities changed greatly with increasing soil depth ( Figure 5). The A. sparsifolia community showed the highest SOC content in all soil layers compared with the other four communities. In addition to the 30-40 cm soil layer, the P. communis community in all soil layers had the lowest SOC content compared with the other four communities.

SOC Vertical Distribution and Seasonal Variations
The SOC of the five communities changed greatly with increasing soil depth ( Figure  5). The A. sparsifolia community showed the highest SOC content in all soil layers compared with the other four communities. In addition to the 30-40 cm soil layer, the P. communis community in all soil layers had the lowest SOC content compared with the other four communities. The β values of the SOC ranged from 0.9472 to 0.9570 for the five communities ( Figure  6). Among all communities, the accumulation of SOC with soil depth differed significantly ( Figure 5). The mean SOC value was only 74.36% for the H. ammodendron, T. ramosissima, R. soongorica, and P. communis communities at 0-30 cm, whereas for the A. sparsifolia community, 80.36% of the SOC was distributed in the 0-30 cm soil layer.
The SOC contents in the five communities varied strongly with seasons ( Table 2). The highest monthly mean values of SOC occurred in the A. sparsifolia community and the lowest ones in the P. communis community. The monthly average values were 1.65-2.89 g/kg for the five communities. The highest SOC values of the A. sparsifolia (3.08 g/kg), T. ramosissima (2.35 g/kg), and R. soongorica (2.34 g/kg) communities were found in June, and The β values of the SOC ranged from 0.9472 to 0.9570 for the five communities ( Figure 6). Among all communities, the accumulation of SOC with soil depth differed significantly ( Figure 5). The mean SOC value was only 74.36% for the H. ammodendron, T. ramosissima, R. soongorica, and P. communis communities at 0-30 cm, whereas for the A. sparsifolia community, 80.36% of the SOC was distributed in the 0-30 cm soil layer.    The SOC contents in the five communities varied strongly with seasons ( Table 2). The highest monthly mean values of SOC occurred in the A. sparsifolia community and the lowest ones in the P. communis community. The monthly average values were 1.65-2.89 g/kg for the five communities. The highest SOC values of the A. sparsifolia (3.08 g/kg), T. ramosissima (2.35 g/kg), and R. soongorica (2.34 g/kg) communities were found in June, and the highest values for the H. ammodendron (2.25 and 2.31 g/kg, p > 0.05) community were found in June and September, and the lowest ones were 2.79, 2.07, 1.68, and 1.62 g/kg, respectively. The lowest and highest SOC values of the P. communis communities were 1.19 g/kg in June and 1.88 g/kg in July.

Effects of Fine Roots on SOC Dynamics
Fine root production, annual decomposition, and SOC supplementation followed the order P. communis > H. ammodendron > A. sparsifolia > T. ramosissima > R. soongorica. The fine root productivity of the P. communis community was 9.61 times that of the R. soongorica community. In this study, fine root production was 50.67-486.92 g/m 2 /yr, decomposition was 16.08-130.24 g/m 2 /yr, and supplementary SOC was 7.14-50.42 g/m 2 /yr. The fine root turnover rate followed the order T. ramosissima > A. sparsifolia > P. communis > R. soongorica > H. ammodendron and ranged from 1.25 to 1.98 times/yr. The maximum fine root net productivity and fine root turnover rate were found for the P. communis and T. ramosissima communities (Table 3).

Environmental Factors Affecting SOC Dynamics
The relationships between SOC and ecological factors (soil water content, soil bulk density, soil conductivity, and pH) were significant (Table 4). According to the data, the SOC first decreased and then increased with increasing soil water content, soil bulk density, soil conductivity, and pH. When soil water content, bulk density, electrical conductivity, and pH were 21%, 1.36 g/cm 3 , 4.75 ms/cm, and 9.63, respectively, the SOC was lowest. Fine root biomass significantly affected the SOC values of all plant communities ( Table 5). With increasing levels of fine root biomass, the SOC values first increased and then decreased; peak levels were reached at a fine root biomass of 12.5-45.0 g/m 2 . However, the change trend of SOC with fine root biomass in the P. communis community showed the opposite pattern, first decreasing and then increasing; at a fine root biomass of 277.77 g/m 2 , the lowest SOC value was reached. Note: y, soil organic carbon (SOC) (g/kg); x, fine root biomass (g/m 2 ). Significance levels: **, p < 0.01; *, p < 0.05.

Relationship between the Vertical Distribution of and Seasonal Variations in SOC and Fine Root Biomass in Each Plant Community
The ability to predict and improve the consequences of global change partly depends on an improved understanding of the distribution and control of SOC, as well as how vegetation changes affect the distribution of SOC throughout the soil column [40]. Different vegetation types can significantly affect the vertical distribution of SOC [27,41,42] According to these scientists, SOC accumulation at deep soil layers in shrub-dominated environments is caused by vertical root dispersion. Based on global estimates, the SOC of desert ecosystems accounts for approximately 44% at a depth of 0-30 cm [41], and in past examinations in northwestern China, over half of the SOC content in shrub plots was aggregated at the 0-40 cm layer [14,43]. In the present study, the SOC levels ranged from 73-80% for the five plant communities and were therefore higher than the global estimate ( Figure 5). As an explanation, different shrub species differ in their fine root biomass levels, leading to different SOC values [14,41,43]. For example, the fine root biomass in desert ecosystems accounts for approximately 61% of the 0-100 cm layer and is concentrated in the first 0-30 cm [41]. In the present study, the living fine root biomass was 64-86%, whereas the dead fine root biomass was 73-85%, above the global estimates ( Figure 2). These results are similar to those reported by other authors, namely that over 60% of the fine root biomass is gathered in the initial 30 cm of soil [41,[44][45][46]. In addition, in the present study, the live and dead fine root biomass at 40-90 cm accounted for 84.65% of the total live and dead fine root biomass for the P. communis community. This is in contrast to the results reported in a past study, where giant reed development occurred in the shallow soil layers; however, according to the authors, these herbaceous crops tend to extend an enormous extent of their fine root biomass to more profound soil layers [47]. In another study, giant reed development occurred in the deeper soil layers, related to the occurrence of groundwater [47]. In the present study, when fine roots were excavated, a large amount of groundwater was also found at greater soil depths.
In the current study, the highest SOC values of the A. sparsifolia (3.08 g/kg), T. ramosissima (2.35 g/kg), and R. soongorica (2.34 g/kg) communities were found in June, and the highest values of the H. ammodendron (2.25 and 2.31 g/kg, p > 0.05) community were found in June and September (Table 2). However, the lowest fine root biomass (43.17-329.89 g/m 2 ) of the five communities occurred in June. Fine root biomass (66.29-523.41 g/m 2 ) and living fine root biomass (48.45-405.75 g/m 2 ) peaked in August and September in the five communities (Figures 3 and 4). Due to the high temperature and low rainfall in summer, the growth of fine roots and the activity of soil animals and microorganisms were limited, with a consequent reduction in SOC mineralization. The SOC reached the maximum in June, when the fine root biomass was lowest. Generally, at the end of the growing season, microbial activity increases, which will increase SOC mineralization. Subsequently, the plants reduce nutrient input into the leaves and increase energy input into the roots, leading to a reduction in SOC. For another, fine root growth occurs mainly in July and August, consistent with the precipitation, which amounted to 30% of the rainfall for the whole year, resulting in a maximum fine root biomass in August and September. Thus, the peak of living fine root biomass reflects the responses of fine root production to rainfall. Some studies have shown that the majority of fine roots developed in spring and died off in summer because of the high temperatures [48], which may be a potential reason for the peak of fine root mortality in July and August. In addition, in the present study, although the fine roots of the P. communis community were concentrated in the 40-90 cm layer, the SOC was concentrated in the upper 0-30 cm, and the SOC content of the P. communis community was the lowest among the five communities. The species P. communis is a tall perennial grass species [49] and occurs in the form of standing litter in the late growing season, which reduces the input of litter, whereas litter and fine root decomposition mainly occur in the upper soil layer [42]. Generally, the deep soil layer has a large bulk density and a low porosity, reducing the activity of microorganisms and thereby impeding fine root decomposition, resulting in a reduction in the input of SOC.

Relationship between SOC and Ecological Factors
SOC stocks are influenced by a combination of environmental factors and organisms. In most cases, environmental factors, including soil temperature, available supplements, soil moisture content, soil texture, and soil pH, are closely related to fine root production and turnover [22,50,51]. The biological factors include fine root characteristics and microbial composition [41,52]. In their literature review, Gill and Jackson [22] reported that soil temperature is, to a great extent, connected with fine root turnover, but in point-scale studies, the effects of slight temperature differences in fine root production and turnover could be ignored, and temperature appears to affect the initiation of fine root production [53]. In this study, regression analysis showed that SOC was lowest when the soil moisture content, bulk density, conductivity, and pH were 21%, 1.36 g/cm 3 , 4.75 ms/cm, and 9.63 respectively (Table 4). Higher water content and soil bulk density are not conducive to aerobic respiration by aerobic microorganisms, resulting in the slower decomposition of organic matter [54][55][56]. Soil pH affects not only the dissolution of organic matter (fine roots), soil nutrients, and redox status (conductivity), but also the structure and activity of the microbial community. In general, at pH > 8.0, the activity of soil microorganisms decreases and root decomposition is slow [57][58][59]. In addition, the composition, structure, and physical condition of fine roots are closely related to SOC. In this study, fine root production, annual decomposition, and SOC replenishment followed the order P. communis > H. ammodendron > A. sparsifolia > T. ramosissima > R. soongorica (Table 3). In principle, simple organic substances such as monosaccharides, polysaccharides, and amino acids are easily converted to SOC by microorganisms, while complex organic substances such as lignin and polyphenols are difficult for microorganisms to use and do not contribute to the formation of SOC [60]. In summary, suitable soil moisture and high soil conductivity facilitate the decomposition of root litter and thus the formation of SOC [54,55]. Higher soil pH and soil bulk density levels are not conducive to soil biological activity and organic matter decomposition, and when they are more unfavorable to SOC mineralization, they lead to increased SOC levels [54,56].

SOC of Decomposed Fine Roots
In the present study, the fine root distribution patterns of the A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron communities were similar to those in various arid areas (Figures 2-4); generally, topsoil layers contain more fine roots [14,27,61,62]. In arid areas, a high fine root biomass in the top layer might increase the capacity of plants to retain soil assets (especially water) and adjust to dry season periods [62], whereas it might diminish their C investment [42]. In this study, there was a distinction in the fine root biomass among the different plant communities (Figures 2-4), probably resulting from the hereditary differences [42]. Previous studies have demonstrated that communities with intense and shallow root systems and those with deep roots can coexist and interact with neighboring shrubs when resources are scarce [14]. In our study, the differences in the vertical distribution of fine roots among communities suggest that they might have different adaption mechanisms when resources are limited [6]. Fine root production varied among the different community plots (Table 3). Besides, the fine root production of P. communis was significantly higher than that of A. sparsifolia, T. ramosissima, R. soongorica, and H. ammodendron, most likely because in semi-arid and arid regions, the topsoil is at least generally or periodically dry [63]. Moreover, because of the heterogeneity of soil nutrients and water, fine root production shows different vertical patterns [64]. In sandy soils, the soil water content increases with depth, indicating that plants with a greater extent of fine roots at more profound depths can utilize water stored at in deeper soil layers, which explains the variation in parched regions [62,63].
An increase in SOC stocks fundamentally relies upon the rate of fine root decomposition [34]. The turnover rate in the present work ranged from 1.25 to 1.98 times/yr (Table 3), falling in the range of shrubs in temperate zones [8]. Huang obtained a fine root turnover rate (1.28-2.12 times/yr) similar to that observed here in the arid regions of northwestern China [27]. In a previous study, for temperate tree species across the eastern US, the average turnover rate was 1.27 times/yr, which was below the average turnover rate (1.6 times/yr) in the present study. Here, the turnover rate was closer to the fine root turnover rate (0.68-2.20 times/yr) of tropical forests. Generally, fine roots in arid regions have high turnover rates, which is an "adaptive strategy". Fine root sloughing, with subsequent microbial immobilization, could help retain root nutrients and mineral elements in the upper soil horizons when plants experience drought, competition, and nutrient deficiency [65,66].
Our results show that a high content of fine root biomass and a high fine root production support the return of organic mass and nutrients through fine root turnover [67]. Nonetheless, higher fine root biomass and production were not related to a greater SOC content in communities (Figures 2-6). Moreover, the SOC stocks did not respond to high fine root biomass [41], and soil C accumulation was not affected by the level of standing fine root biomass and its production [68,69]. A possible explanation is that in addition to fine root, leaf, and branch litter, the insoluble organic matter from roots, refractory microbial biomass, and living root exudates can, to a great extent, contribute to SOC accumulation at more profound soil depths in some regions [70,71] For example, some studies have indicated the importance of SOC originating from living roots as root exudates [71], and there is substantial evidence that living root inputs, not root or shoot litter, play a significant part in SOC formation [72]. Overall, these results suggest that the fine root litter of different communities contributes to SOC accumulation at different rates and that the mechanism of SOC accumulation varies among plant communities.

Conclusions
In this study, an asymptotic single-parameter global model was used to describe the depth distribution of the SOC reservoir (Figure 6), and the results implied that soil organic carbon levels were up to 70% below 0-30 cm depths in five plant communities. At the same time, the survey showed that live and dead fine root biomass at 0-30 cm accounted for 63.86-86.0% and 73.04-85.03% of the total fine root biomass (except for the P. communis community), respectively (Figures 2 and 3). Regression analysis showed a significant positive correlation (p < 0.05) between the vertical distribution of SOC and fine root biomass (Table 5). Furthermore, SOC and fine root biomass showed a reciprocal relationship with seasonal changes, with SOC reaching a maximum (1.88-3.08 g/kg) in summer (June and July) ( Table 2); however, the lowest values of fine root biomass were found in June (43.17-329.89 g/m 2 ) ( Figure 4). Subsequently, SOC declined to various degrees in August, September, and October (Table 2), while fine root biomass returned to its peak in August and September (66.29-523.41 g/m 2 ) (Figure 4). The differences in fine root biomass and SOC among the different plant communities in this study were attributed to two factors. Regression analysis showed that soil water content and soil bulk density had a significant effect on SOC (p < 0.05) ( Table 4). In addition, the results implied that the turnover rate of fine roots ranged from 1.25 to 1.98 times/yr (Table 3). Genetic differences in plant communities mainly resulted in fast and slow rates of fine root turnover and decomposition, as well as heterogeneity in soil texture and moisture, resulting in fine root production showing different vertical patterns and different mechanisms of soil carbon stock accumulation.
Author Contributions: Conceptualization, X.Z. and B.J.; methodology, X.L. and S.T.; data curation, S.T. and X.L.; writing-original draft preparation, S.T. and X.Z.; writing-review and editing, S.T. and X.Z.; supervision, X.Z. and B.J. All authors have read and agreed to the published version of the manuscript.