Changes in the Soil Labile Organic Carbon Fractions following Bedrock Exposure Rate in a Karst Context

Abstract

Soil labile organic carbon fractions (SLOCFs) mainly include microbial biomass carbon (MBC), dissolved organic carbon (DOC), easily oxidized organic carbon (EOC) and light fraction organic carbon (LFOC). The link between bedrock exposure rates with SLOCFs and the carbon pool management index under karst rocky desertification has not been well understood. We selected the bedrock exposure rate and vegetation coverage of 30–50% (light bedrock exposure, LBE), 50–70% (moderate bedrock exposure, MBE) and >70% (intense bedrock exposure, IBE) as the experimental sample plots according to the classification standard of karst rocky desertification, and then selected a sample plot of 0–30% (secondary forest, SF) as the control. This study compared the concentrations and stocks of soil organic carbon (SOC) and SLOCFs and analyzed the relevant carbon pool management index on karst landforms at Anshun, S.W. China. The aims were to determine the relationship between bedrock exposure rates and SLOCFs and to identify the most limiting factors for SLOCFs in karst rocky desertification areas. We found that (1) the concentrations and stocks of SLOCFs declined with increasing soil depth. SOC, DOC and MBC showed IBE > LBE > MBE > SF; LFOC decreased with increasing bedrock exposure rate, and EOC did not show obvious regularity. (2) The carbon pool management index and sensitivity index had significant differences under different bedrock exposure rates. Redundancy analysis and linear regression showed that the increase in bedrock exposure rate had a great impact on MBC, DOC, EOC and SOC. In conclusion, the increase of bedrock exposure rate has no side impact on the DOC, EOC and MBC of the soil, but side effects are exhibited by LFOC. Secondary forest improves the integrity of karst landscapes, and does not change the soil properties as well as the concentrations and stocks of SLOCFs in karst rocky desertification areas.

1. Introduction

Karst ecosystems, as a natural landscape in arid and semi-arid areas, are widely distributed all over the world and pose a serious threat to the ecological environment and socioeconomic advancement [1], with an area of 22 million km² and accounting for approximately 12% of the global land area [2,3]. Therefore, it is one of the main ecologically fragile zones in our terrestrial ecosystem [3]. Karst rocky desertification (KRD) is a process and result of the interaction between ecosystem vulnerability and the unreasonable economic activities of humanity under karst landform conditions, and it is also the main ecological environmental problem in typical karst areas [4]. The main characteristics of KRD areas are a thin soil layer, low vegetation coverage, and large areas of exposed bedrock caused by serious soil and water loss [5]. Additionally, a series of social and economic problems, such as population poverty, economy, science and technology and culture, have further aggravated the ecological and environmental problems of KRD region [6], which has been the main reason for the aggravation problem of KRD in Southwest China for many years [5,7].

The soil organic carbon (SOC) pool is the largest carbon pool in terrestrial ecosystems and is affected by many external factors, such as climate type, plant diversity, microbial community structure, and topography, and it is also closely related to soil type, vegetation coverage, and land use type [8,9], and therefore it plays an important role in atmospheric CO₂ level balance [10]. In addition, SOC stocks are affected by land use, soil fertility, soil temperature, and water [11,12]. Meanwhile, small changes of SOC in terrestrial ecosystems may lead to variations in the carbon budget and the stability of the climate system [13]. Moreover, SOC is the key factor controlling soil fertility and agricultural production [14], which are usually affected by the land-use conversion, climatic factors, vegetation, management practices, clay content, etc. [15], and the loss of SOC can lead to a decline in soil fertility and quality, and affect the stability of terrestrial ecosystems [16,17]. Karst ecosystems, especially in KRD areas, comprise a significant part of the Earth’s surface system, and its SOC changes will have an impact on the matter cycle of other regions and the whole Earth system [18,19]. Meanwhile, SOC sequestration significantly increases with restoration age, and the large scale of land-use change under the ‘Grain-for-Green’ Program will significantly increase SOC stocks [20,21]. These inconsistent results from scholars show that SOC dynamics are a complex process that is affected by multiple factors among different bedrock exposure rate areas [10,22].

Soil labile organic carbon fractions (SLOCFs) mainly include microbial biomass carbon (MBC), dissolved organic carbon (DOC), easily oxidized organic carbon (EOC), and light fraction organic carbon (LFOC). MBC and DOC are mainly controlled by soil microorganisms, while microorganisms are affected by soil environment and biological factors, mainly including temperature, humidity, and soil permeability [23]. However, they account for only a small part of SOC, and are closely related to the migration, fixation and carbon dioxide release of SOC in soil ecosystem [24,25]. EOC accounts for 13–28% of soil total organic carbon, and it moves relatively fast, and is unstable and easy to oxidize, decompose and mineralize in the soil [26,27]. EOC can be used as an indicator of changes in soil quality and fertility, easily moving with the solvent, and participating in a wide range of the carbon cycle [13,28]. LFOC is highly labile and consists of soil organic matter for microbes and plants as a major source of soil nutrients [24,29]. Soil carbon sequestration in the fragile ecological environment is significantly affected by the relative distribution of SLOCFs [24]. Additionally, the pool of SLOCFs is the most active fraction of SOC and has a rapid turnover rate, which governs the production and flux of CO₂ from the soil to the atmosphere [26]. Moreover, these small fractions of SOC are more sensitive to land-use change or management practices than SOC [27]. The relationship between soil properties and plant diversity [30,31], microbial community structure and enzyme activity [17,32], soil organic carbon components and their sensitivity to KRD control measures, and the dynamic process of SOC [10,19] in karst areas has been widely studied, but studies on SLOCFs under different bedrock exposure rates in karst areas are rare. Thus, it is essential to explore the distribution characteristics of SLOCFs at different bedrock exposure rates of KRD areas, and we hypothesized that increasing bedrock exposure rate would have a serious restrictive effect on SLOCFs in the karst ecosystems of Southwest China, and expected this to provide a scientific basis for regional remediation, the SOC cycle, and soil quality management in KRD areas.

Here, SLOCFs for three different bedrock exposure rates in KRD areas, namely, the light bedrock exposure rate (LBE), the moderate bedrock exposure rate (MBE), and the intense bedrock exposure rate (IBE), were studied. These samples were compared with the reference secondary forest (SF). The aims were (i) to explore the distribution characteristics of SLOCFs in different bedrock exposure rates of karst rocky desertification; (ii) to quantify the stocks of SLOCFs under different bedrock exposure rates; and (iii) to understand the response of SLOCFs to bedrock exposure rates in karst rocky desertification areas of Southwest China.

2. Materials and Methods

2.1. Study Site

We conducted the experiment at the sides of the Beipan River, Guanling–Zhenfeng junction, Guizhou Province, in Southwest China (Figure 1) (25°39′12″~25°41′09″ N, 105°39′59″~105°40′27″ E, with an average altitude of 756 m), covering a total area of 51.62 km², of which the KRD area accounts for 53.82%. The study area experiences a subtropical mountainous monsoon climate with a mean annual precipitation of 1100 mm (83% occurs during May to October) and an average annual temperature of 18.4 °C [32]. The soil is mainly sandy loamy calcareous soil that lacks an aggregate structure and has a poor water holding capacity. This study site is seriously disturbed by human beings; most of the plants are secondary forests, and the main crops are Zanthoxylum bungeanum and Pitaya.

2.2. Soil Sample Collection

We selected the bedrock exposure rate and vegetation coverage of 30–50%, 50–70% and >70% as the experimental sample plots according to the classification standard of KRD, and then selected a sample plot of 0–30% as the control, named LBE, MBE, IBE and SF, respectively [5]. All KRD sample plots were divided into three large 20 m² plots and then divided into five 2 m² small plots in 20 m² plots for soil sample collection, and the soil layers were divided into 0–10 cm, 10–20 cm and 20–30 cm. Finally, the soil samples were collected and taken back to the laboratory to screen out roots, stones and other debris with a 2 mm particle size and were naturally dried to measure the SLOCFs of the soil.

2.3. Soil Sample Determination

Soil organic carbon (SOC), total nitrogen (TN), soil bulk density (BD) and SLOCFs (MBC, DOC, EOC and LFOC) were determined in our study. SOC was determined using the dichromate oxidation method [33]. TN was analyzed by the Kjeldahl method [34]. The LFOC, EOC, MBC and DOC were determined using the density fractionation method [35], chemical oxidation method using KMnO₄ [36], chloroform fumigation–K₂SO₄ extraction method, and K₂SO₄ extraction [24], respectively. MBC was fumigated with chloroform (CHCl₃), then extracted with 0.5 mol L⁻¹ K₂SO_4(aq), and finally put the sample into TOC analyzer for detection and analysis. DOC was extracted with 0.5 mol L⁻¹ K₂SO_4(aq) and then determined by TOC analyzer. EOC was oxidized with 333 mmol L⁻¹ KMnO4 and determined by spectrophotometer. LFOC used NaI solution to separate heavy organic carbon in soil, and finally used the same determination method as SOC.

2.4. Data Analysis

The data and chart analyses of the SLOCFs were conducted by GraphPad Prism 8.0. The relationships between the bedrock exposure rate and MBC, EOC, DOC, and LFOC were analyzed using Canoco 5.0 software and all data were processed by Excel 2010 software.

The soil organic carbon stocks (SOCs) (kg m⁻²) in different soil layers were calculated as follows [37]:

$$\mathrm{SOCs}=\frac{\mathrm{OC}\times \mathrm{D}\times \mathrm{BD}}{100}$$

(1)

where OC represents the SOC, MBC, DOC, EOC, and LFOC concentrations (g kg⁻¹ in whole soil) of total soil when calculating SOC, MBC, DOC, EOC, and LFOC stocks in total soil, respectively; D is the thickness (cm) of the 0–10 cm soil depth; and BD is the soil bulk density (g cm⁻³).

The sensitivity index (SI) was calculated as follows [38]:

$$\mathrm{SI}=\frac{{\left({\mathrm{C}}_{\mathrm{pool}}\right)}_{\mathrm{experiment}}-{\left({\mathrm{C}}_{\mathrm{pool}}\right)}_{\mathrm{control}}}{{\left({\mathrm{C}}_{\mathrm{pool}}\right)}_{\mathrm{experiment}}}$$

(2)

where (C pool) measures refer to the different C pools under LBE, MBE and IBE and (C pool) reference means the different C pools under SF.

The carbon pool management index (CPMI) values were determined using EOC concentrations, which were calculated using the following equation according to [36]. Additionally, the carbon pool index (CPI), liability index (LI), and soil liability of carbon (L) were calculated as follows [19]:

$$\mathrm{CPMI}=\mathrm{CPI}\times \mathrm{L}\mathrm{I}\times 100$$

(3)

$$\mathrm{CPI}=\frac{{\mathrm{SOC}}_{\mathrm{experiment}}}{{\mathrm{SOC}}_{\mathrm{control}}}$$

(4)

$$\mathrm{LI}=\frac{{\mathrm{L}}_{\mathrm{experiment}}}{{\mathrm{L}}_{\mathrm{control}}}$$

(5)

$$\mathrm{L}=\frac{\mathrm{EOC}}{\mathrm{SOC}-\mathrm{EOC}}$$

(6)

where SOC experiment is the total SOC concentration (g kg⁻¹) in LBE, MBE and IBE and SOC control is the total SOC concentration (g kg⁻¹) in SF.

3. Results

3.1. Soil Labile Organic Carbon Fraction Stocks

In different bedrock exposure areas, the concentrations and stocks of SOC and LFOC decreased with increasing soil layer depth (Figure 2). We found that the concentrations of SOC from SF to IBE were 13.65 ± 0.45, 26.80 ± 1.30, 19.20 ± 1.95 and 30.82 ± 1.67 (g kg⁻¹) at a depth of 0–10 cm, respectively. Overall, the changing trend of MBC concentration under different bedrock exposure rates was SF (0.10 ± 0.06) < LBE (0.19 ± 0.04) < MBE (0.20 ± 0.06) < IBE (0.30 ± 0.06 g kg⁻¹), and showed a positive correlation with increasing bedrock exposure rate. However, DOC concentration was SF (0.10 ± 0.05) < MBE (0.14 ± 0.04) < LBE (0.18 ± 0.04) < IBE (0.24 ± 0.04 g kg⁻¹). EOC concentration changes without regularity at different bedrock exposure rates and in different soil layers. LFOC decreased with increasing bedrock exposure rate, imposing serious side effects on LFOC. Additionally, there was a significant difference in SLOCF concentrations among the different bedrock exposure rates (p < 0.0.5).

As shown in Figure 3, the SLOCF stocks were also greatly affected by the bedrock exposure rate. In the topsoil layer (0–10 cm), there was no difference in SOC, MBC, DOC and EOC stocks between different bedrock exposure rates, and the overall stocks were the highest in IBE, followed by LBE. However, the stocks of LFOC decreased with increasing bedrock exposure rate. With increasing soil layer depth, the stocks of SOC showed little change in topsoil, but the stocks of SLOCFs in the 10–20 cm soil layer were quite different. In addition, the stocks of MBC were IBE > MBE > LBE > SF under different bedrock exposure rates; DOC and EOC were IBE > LBE > MBE > SF and IBE > SF > MBE > LBE, respectively. However, the LFOC stocks gradually decreased with increasing bedrock exposure rate. As the soil layer reached 20–30 cm, the stocks of SOC, MBC and DOC showed IBE > LBE > MBE > SF, EOC was MBE > IBE > LBE > SF, and LFOC was consistent with that of other soil layers.

3.2. Proportions of Soil Labile Organic Carbon Fractions

Different bedrock exposure rates and soil layers had a great impact on the proportion of SLOCFs at our study site. As shown in Figure 4, there were no significant differences among different bedrock exposure rates in MBC/SOC in the 0–10 cm soil layer (p > 0.0.5), and the highest proportions appeared in SF and MBE. The overall trend of DOC/SOC was consistent with MBC/SOC, but this proportion was greatly affected by the bedrock exposure rate. However, the difference between EOC/SOC and LFOC/SOC was obvious under different bedrock exposure rates, which gradually decreased with increasing bedrock exposure rate in rocky desertification areas.

Interestingly, we found that there was no significant difference between the 10–20 cm and 20–30 cm DOC/SOC in different bedrock exposure areas, which indicated that DOC/SOC was less affected by rocky desertification with increasing soil layer depth. In contrast, MBC/SOC changed significantly at depths of 10–20 cm and 20–30 cm compared with that a depth of 0–10 cm. The proportion of LFOC/SOC in the second soil layer was consistent with that in the first soil layer, and it was SF > MBE > LBE > IBE at 20–30 cm.

3.3. Sensitivity of Soil Labile Organic Carbon Fractions

The CPI values changed significantly for different bedrock exposure rates in karst rocky desertification areas (Figure 5 and

Table 1. Soil liability of carbon (L), liability index (LI), carbon pool management index (CPMI), and carbon pool index (CPI) in karst rocky desertification areas with different bedrock exposure rates.

Measures	Soil Layers (cm)	CPI	L	LI	CPMI
SF	0–10	−	0.12 ± 0.01 Aa	−	−
	10–20	−	0.18 ± 0.03 Aa	−	−
	20–30	−	0.13 ± 0.03 Aa	−	−
LBE	0–10	1.48 ± 0.41 Ab	0.09 ± 0.01 ABa	0.79 ± 0.14 Aa	115.5 ±32.1 Aa
	10–20	2.39 ± 0.44 Ab	0.03 ± 0.01 Cb	0.20 ± 0.05 Bb	47.46 ± 13.33 Ab
	20–30	3.74 ± 0.72 Aa	0.03 ± 0.01 Bb	0.26 ± 0.02Bb	97.3 ± 22.8 Aab
MBE	0–10	0.77 ± 0.21 Aa	0.07 ± 0.01 Ba	0.61 ± 0.15Aa	45.09 ± 8.93 Aa
	10–20	0.76 ±0.06 Ca	0.08 ± 0.02 Ba	0.43 ± 0.05Aa	32.72 ± 4.70 Aa
	20–30	0.72 ± 0.19 Ba	0.08 ± 0.01 Aa	0.70 ± 0.23Aa	49.30 ± 18.80 Aa
IBE	0–10	1.65 ± 0.60 Aa	0.07 ± 0.02 Ba	0.60 ± 0.28Aa	96.70 ± 61.80 Aa
	10–20	1.64 ± 0.19 Ba	0.06 ± 0.01 BCa	0.31 ± 0.03 Ba	50.59 ± 4.17 Aa
	20–30	1.61 ± 0.52 Ba	0.03 ± 0.01 Ba	0.28 ± 0.07 Ba	45.90 ± 24.6 Aa

Note: SF (secondary forest); LBE (light bedrock exposure); MBE (moderate bedrock exposure); IBE (intense bedrock exposure). Different lowercase letters represent significant differences between different soil layers, and different uppercase letters represent significant differences between different treatments (p < 0.05).

). In the 0–10 cm soil layer, with increasing bedrock exposure rate, CPI values first decreased and then increased, and the value was the highest in IBE (1.65 ± 0.60). With increasing soil layer depth, the trend of change was consistent with that at 0–10 cm. In the LBE sample plot, the CPI value increased gradually with increasing soil layer depth, and there was a significant difference between the different soil layers. However, the MBE and IBE gradually decreased with increasing soil layer depth. The LI value gradually decreased with increasing bedrock exposure rate in 0–10 cm. In the 10–20 cm and 20–30 cm soil layers, the LI value first increased and then decreased with increasing bedrock exposure rate, showing a “U”-shaped change. In general, it first increased and then decreased with increasing soil layer depth, with the highest LBE in the 0–10 cm soil layer (0.79 ± 0.14) and the lowest LBE in the 10–20 cm layer (0.20 ± 0.05).

We found that CMPI values in LBE and IBE decreased gradually with increasing soil layer depth, and there were significant differences among different soil layers (p < 0.05). In the 10–20 cm soil layer, CMPI first decreased and then increased with increasing bedrock exposure rates. However, in the 20–30 cm soil layer, it first increased and then decreased. There was no significant difference in the L values with increasing bedrock exposure rate between LBE, MBE and IBE (p > 0.05), while SF showed a significant difference compared to the other samples (0.12 ± 0.01). LBE and IBE gradually decreased with increasing soil layer depth, while SF first increased and then decreased; however, there was little change in MBE.

Other than for MBC, the SI values of the SLOCFs were in the order IBE > LBE > MBE at 0–10 cm. However, the SI value of MBC increased with increasing bedrock exposure rate. In the 10–20 cm soil layer, the response of SI values of EOC and LFOC to bedrock exposure rate was MBE > IBE > LBE; SOC, MBC and DOC were LBE > IBE > MBE, LBE > MBE > IBE and IBE > LBE > MBE, respectively. In the 20–30 cm soil layer, the SI values of SOC, MBC and DOC were LBE > IBE > MBE, and those of EOC and LFOC were MBE > LBE > IBE.

3.4. Relationship between Soil Labile Organic Carbon Fractions and Bedrock Exposure Rate

Through linear regression fitting of SLOCFs and SOC in karst rocky desertification areas with different bedrock exposure rates (Figure 6), we found that linear fitting of LFOC in the SF sample was p < 0.001, F = 266.3; with increasing bedrock exposure rate, the p-value and F value of the linear regression equation gradually decreased, which indicated that LFOC experienced serious side effects with increasing bedrock exposure rate in rocky desertification areas. The F and p-values of MBC and SOC in the IBE plot were 1.07 and 0.34, respectively, and in SF they were 49.05 and 0.0002, respectively, but there was no significant difference between LBE and IBE. Interestingly, we found that the different bedrock exposure rates in rocky desertification areas had no significant impact on DOC, only in MBE, but almost no impact in SF, LBE or IBE. The response of EOC to the bedrock exposure rate was generally consistent with that of LFOC. They were all in areas with higher bedrock exposure rates, which were greatly affected.

As shown in the redundancy analysis in Figure 7 and

Table 2. Statistical table of redundancy analysis of soil labile organic carbon fractions and organic carbon related indexes.

Axis	I	II	III	IV	Explains (%)	Pseudo-F	P
Eigenvalues	0.7029	0.0162	0.0004	0.00	−	−	−
Explained variation	70.29	71.91	71.95	71.95	−	−	−
Pseudo-canonical correlation	0.8470	0.9121	0.8963	0.3604	−	−	−
Explained fitted variation	97.69	99.94	100.00	100.00	−	−	−
SOC	−	−	−	−	62.3	16.5	0.002
MBC	−	−	−	−	56.8	13.1	0.002
DOC	−	−	−	−	36.9	5.9	0.05

, in rocky desertification areas with different bedrock exposure rates, the response of SLOCFs and carbon pool related indexes was different. Changes in bedrock exposure rate had a significant effect on the DOC, SOC and MBC. The explained variances were 36.9%, 62.3% and 56.8%, respectively; p-values were ≤ 0.05, and Pseudo-F were 5.9, 16.5 and 13.1, respectively. The explained variance values of Axis 1, Axis 2, Axis 3 and Axis 4 were 70.29%, 1.62%, 0.4% and 0.0%, respectively. Overall, in rocky desertification areas under different bedrock exposure rates, CPMI, CPI and LI were several indicators with strong correlations with SLOCFs and bedrock exposure rate. These carbon-related indexes could be considered further in future explorations of soil quality and remediation in karst rocky desertification areas.

4. Discussion

SLOCFs (MBC, DOC, EOC and LFOC) are the most active factions in the SOC cycle, especially in karst rocky desertification areas, which reflects the important response characteristics of soil quality to different bedrock exposure rates [37,39]. As shown in Figure 2 and Figure 7, we found that the changes of MBC, DOC and EOC were highly correlated with the content of SOC. MBC and DOC increased with increasing SOC concentration, which is consistent with the findings of Qin [40]. Additionally, MBC represents the size of active microbial biomass, which largely depends on the supply of SOC [41,42] Therefore, the MBC content in our study was consistent with SOC and showed a trend of IBE > LBE > MBE > SF. DOC is produced by the decomposition of organic matter driven by soil microorganisms, which is mainly dependent on the degree of degradation of organic matter by microorganisms and the amount of organic matter in the ecological environment [43,44,45]. In this study, the distribution trend of DOC concentrations was consistent with MBC at different bedrock exposure rates. The main factors affecting MBC and DOC in rocky desertification areas could be the fact that that SF is richer in plant species composition, more shrubs and more organic matter [46], but rainwater scouring is more serious, microorganisms do not decompose organic matter sufficiently, and organic matter is lost by rainwater [47]. In contrast, IBE showed that there was less organic matter and less lost, and most of it was concentrated in soil [43]. In addition, those with a high bedrock exposure rate may have higher activity with respect to the types of microorganisms degrading organic matter.

The concentrations of EOC and LFOC are different from MBC and DOC, and not all of them increase with increasing SOC. EOC is produced by plant residue leaching, root secretion, and the soil microorganisms driving the decomposition of organic matter [48,49]. There was no regularity in EOC distribution. LBE, IBE and MBE had the highest EOC concentrations at depths of 0–10 cm, 10–20 cm and 20–30 cm, respectively. The reason for this may be that the LBE had a high vegetation coverage and a large amount of organic matter in the topsoil. Although IBE had a limited soil area, the rain erosion was serious, resulting in the easy decomposition and loss of EOC at 10–20 cm. In MBE, due to the combined action of rain erosion and plant roots, EOC was largely stored in deeper soil layers [19,30,48]. In addition, the concentrations of LFOC are quite different from those of SOC. LFOC is a highly unstable component of active organic carbon, which is mainly decomposed by plant organic matter in soil [17,36]. We found that in the study area, the content of LFOC always maintained the trend of SF > LBE > MBE > IBE and 0–10 cm > 10–20 cm > 20–30 cm. The main factor of this phenomenon may be the gradual decrease in the composition of the plant community, while the content of organic matter decreases from SF to IBE [19,50]. Therefore, the trend of LFOC indicates that it may be more affected by plant organic matter than by other factors in rocky desertification areas.

The stocks of SLOCFs are determined by carbon input and output [35,51], and are closely related to soil properties (soil bulk density, soil layer thickness, etc.) [14]. In karst rocky desertification areas, vegetation restoration reduces the bedrock exposure rate, improves soil quality, and enhances vegetation coverage, but does not change the gravel content and soil depth [52]. Moreover, the interception of rainfall by vegetation further leads to the reduction of soil capacity, and the input of vegetation litter has a significant impact on the increase of exogenous carbon in the soil, especially for LFOC [53,54]. With increasing bedrock exposure rate in rocky desertification areas, the vegetation coverage rate gradually decreases and the soil area decreases, but the shallow fissure soil increases and the gravel content in the soil decreases greatly [30,55]. Additionally, the concave ground formed by special bedrock characteristics causes rainwater scouring to converge into the soil, resulting in higher soil bulk density in areas with higher bedrock exposure rates [56]. Therefore, different bedrock exposure rates lead to great differences in soil carbon reserves. In karst rocky desertification areas, the labile organic carbon stocks of different exposure rates of bedrock are affected by many factors.

The carbon pool management index (CPMI) is an index including the soil organic carbon pool (carbon pool index, CPI) and soil organic carbon stability (stability index, LI), which is widely regarded to be a sensitive index for evaluating and reflecting soil quality and soil organic carbon activity [14,57]. Therefore, the change of CPMI in habitats with different bedrock exposure rates plays an essential role in the embodiment of carbon input and soil quality [58], especially for the variety of soil surface vegetation coverage caused by different bedrock exposure rates in rocky desertification areas, which affects the difference of soil labile organic carbon components [48]. In the current study, with different grades of rocky desertification, some soil CPMI decreased with increasing degree of rocky desertification, and some increased first and then decreased [19]. Different scholars have reached different conclusions on the research of CPMI in karst rocky desertification areas, which is mainly driven by external factors such as the complex division of rocky desertification, the serious impact of human disturbance, soil layer, slope, and altitude [58]. Therefore, research into the management index of karst regional carbon pools should be further systematically and scientifically incorporated into references of similar research sample plots in the future.

5. Conclusions

The stocks and concentrations of SOC decreased with increasing soil depth and showed IBE > LBE > MBE > SF in karst rocky desertification areas with different bedrock exposure rates. MBC and DOC increased with increasing bedrock exposure rate; EOC and LFOC had higher concentrations in SF and LBE. However, the ratio of SLOCFs to SOC was the highest in SF, and the SI decreased with increasing bedrock exposure rate. Redundancy analysis showed that the increase of bedrock exposure rate had a more obvious response to EOC, DOC and MBC, while the response of LFOC to vegetation restoration was more obvious. These findings further deepen our understanding of rocky desertification areas with different bedrock exposure rates and can be used to formulate strategies and practices for desertification restoration and the return of farmland to forest in similar karst areas.