The Carbon Storage of Reforestation Plantings on Degraded Lands of the Red Soil Region, Jiangxi Province, China

Abstract

To assess the effects of reforestation on ecosystem carbon storage, a long-term Forest Restoration Experimental Project (FREP) was established in 1991 on southern degraded red soil in Taihe County, Jiangxi Province, China. In this study, we selected five types of plantations: Schima superba plantation (SS), Liquidambar formosana plantation (LF), Pinus massoniana plantation (PM), Pinus elliottii plantation (PE), and P. elliottii and broadleaf mixed plantation (MEB). The unforested land was used as an experimental control check (CK). We aimed to assess the changes in carbon storage in plantations and the factors affecting them. Thirty years after reforestation, the ecosystem carbon storage of the five types of plantations was significantly higher than that of the control site, and there were also significant differences in the ecosystem carbon storage between the different plantation types (p < 0.05). The ecosystem carbon storage of SS, MEB, LF, PM, and PE were 211.71 Mg ha⁻¹, 199.02 Mg ha⁻¹, 160.96 Mg ha⁻¹, 155.01 Mg ha⁻¹, and 142.88 Mg ha⁻¹, respectively. Compared to the CK, these values were increased by 436.8%, 404.6%, 308.1%, 293.1%, and 262.3%, respectively. The ecosystem carbon storage was significantly positively correlated with soil porosity, total nitrogen (TN), and stand density, and was significantly negatively correlated with pH, Pielou’s evenness index (PEI), and the Shannon–Weiner diversity index (SWDI). The soil water content (SWC), bulk density (BD), SWDI, and stand density can be used as indicators of the impact of reforestation plantings on ecosystem carbon storage. The research results has shown that reforestation plantings significantly increase ecosystem carbon storage, and that afforestation should be encouraged on degraded land.

1. Introduction

The current global climate crisis is becoming increasingly serious in terms of frequency and intensity, and its impact is become more widespread. According to the IPCC (2019), the global average temperature has increased by 1.53 °C since preindustrial times [1]. Increased CO₂ emissions from land use, land cover change, and human activities account for approximately 66% of the global warming effect and are the main cause of global warming [2]. Limiting the global average temperature requires both reducing carbon dioxide emissions and removing carbon dioxide from the atmosphere through terrestrial carbon sinks [3]. Considerable efforts have been made to develop affordable carbon (C) capture systems during the past few centuries. Carbon sequestration (CS) in terrestrial ecosystems, especially forests, has large environmental and economic impacts and provides a negative feedback to global warming [4]. From the Kyoto Protocol (1997) and the Copenhagen Accord (2009) to the Glasgow Climate Convention (2021), carbon sequestration by forests has been recognized as the greenest, most economical, and most scale effective technology pathway [5]. It is therefore important to achieve the sustainable management of forests and increase the carbon storage of forest plantations.

Forest ecosystems are the world’s largest terrestrial carbon sinks; they store carbon in different pools that are essential for regulating the carbon cycle [6]. Biomass production (BP) consists of photosynthetically derived carbon used for biomass growth, that is, for building up leaves, wood, and roots [7]. Forests cover only 28% of terrestrial ecosystems but store 80% of aboveground biomass carbon and 40% of belowground biomass carbon, with an incremental carbon sink effect of approximately 65%–90% of that of terrestrial ecosystems [8,9]. Soil carbon storage (SCS) refers to carbon that is stored in soil as soil organic carbon (SOC) [10]. Globally, forest ecosystems total approximately 861.66 Pg C, with their soil carbon storage accounting for the highest proportion (44%), followed by aboveground and belowground biomass (42%), dead wood (8%), and surface litter (5%) [4]; however, the forest ecosystems have been damaged by extensive human activities, leading to a significant loss of global carbon storage [11]. Reforestation is considered the ecological measure with the highest potential for sink enhancement [12] and plays a crucial role in slowing the increase in CO₂ concentration. Reforestation has dramatically altered the dynamics of organic carbon at the ecosystem, regional, and global scales [13]. After reforestation, carbon storage in ecosystems and vegetation layers generally tends to increase, with vegetation having the greatest impact on ecosystem carbon storage [14]; however, due to a combination of vegetation, site, and soil factors, soil carbon storage changes exhibit four patterns: increasing continuously [15,16], varying [17], decreasing [18], and decreasing followed by increasing [19]. In terms of the vertical distribution of an ecosystem’s carbon storage, the contribution of carbon storage by the vegetation layer increases with vegetation growth and rehabilitation, while the contribution of carbon storage in the soil layer decreases [20]. In addition, plantation carbon storage depends on many factors, such as plantation type and structure, environmental variables, disturbances, and management practices [21,22,23]. There is still a great deal of controversy regarding the effect of species richness, i.e., the number of species in a stand, on carbon storage. Dayamba et al. [24] showed a positive correlation between species richness and biomass productivity, and Suo et al. [25] noted a negative correlation in temperate forests, while Vilà et al. [26] found no relationship between productivity and species richness in coniferous forests.

The red loamy hills of southern China are rich in water and heat resources and were historically covered by evergreen forests [27]; however, heavy human interference—particularly harvesting, burning, and trash—combined with overgrazing by animals has decreased the area’s evergreen vegetation, transforming it into a “red desert” and rendering it among the most degraded regions on the planet [28]. To prevent soil erosion, China has been implementing large-scale afforestation and reforestation initiatives in southern degraded red soil areas since the 1970s [29]. The plantation type also significantly influences vegetation and soil carbon accumulation [30]. A previous study conducted in the red soil region of southern China found that the total amount of ecosystem carbon pools in broadleaf and mixed plantations was found to be 25% and 16% greater, respectively, than that in coniferous plantations 19 years after reforestation [27]. Furthermore, based on a comparison of the various rebuilt plantation types with the experimental control (unforested land), the 0–20 cm soil organic carbon content was ranked as follows: Schima superba > Liquidambar formosana > mixed plantation > Pinus elliottii > Pinus massoniana, which increased by 59%, 51%, 50%, 37%, and 27%, respectively [31]. Furthermore, in comparison to severely degraded sites due to erosion, the soil organic carbon storage at a depth of 0–20 cm in the thistles of the eighth and thirty-sixth years increased by 177% and 558%, respectively [15]; however, quantitative knowledge regarding the contributions of various patterns to the recovery of plantation ecosystems via the carbon melting effect is still lacking, even though researchers have already carried out preliminary studies on the carbon reservoir characteristics of plantation ecosystems in the red soil erosion zone in southern China.

In our research area, a long-term Forest Restoration Experimental Project (FREP) was designed and implemented in 1991. This project provides an ideal platform for evaluating the carbon sink effect of reforestation. The objectives of this study were as follows: (i) to characterize the changes in biomass carbon sequestration (trees and the rest of the vegetation) in different plantations, (ii) to determine the characteristics of soil carbon storage in different plantations, and (iii) to elucidate the characteristics of ecosystem carbon storage in different plantations and their influencing factors. This research aimed to evaluate the carbon storage of various plantation ecosystems in degraded hilly red soil areas in southern China. Hopefully, these findings will aid in the selection of appropriate tree species for afforestation on degraded red soil lands and evaluate how well different reconstruction types can rebuild plantation ecosystems. Moreover, these findings can provide scientists and field practitioners with accurate and trustworthy information for managing degraded lands effectively for the purpose of producing biomass, reducing CO₂ emissions, and mitigating global warming.

2. Materials and Methods

2.1. Research Site

The research site is located in Taihe County, Jiangxi Province, China, at 26°54′~26°55′ N, 114°48′~114°49′ E (Figure 1), which has a typical subtropical humid monsoon climate that is hot and humid with an average annual temperature of 18.6 °C, annual rainfall of 1363 mm, and an average annual sunshine duration of 1756.4 h. The dominant soils are classified as Ferralsols (FAO/Unesco) which primarily developed from Quaternary red soil. The topography is typical of the Jitai Basin, with elevations ranging from 70 to 113 m, and the mean slope gradient of this area is approximately 15 °. Before the 1990s, the native evergreen broad-leaved forest had long since disappeared, with only a few arid and semiarid grasses (i.e., Setaria viridis, Imperata koenigii, Cynodon dactylon, Heteropogon contortus, Cymbopogon goeringii, and Arundinella anomala) scattered around [28], and the soil organic matter content was extremely low (6.5 g/kg) due to the long-term intensity of woodcutting, thatch cutting, digging, and overgrazing (Figure 2). In 1991, a long-term Forest Restoration Experimental Project (FREP) with single and mixed tree species, including Schima superba, Liquidambar formosana, Pinus massoniana, P. elliottii, and mixed tree species of P. eliottii and broad-leaved trees, was established by our research group and the local government to rehabilitate the functions of the harshly degraded ecological ecosystem with the support of the local government. The FREP covered an area >133.33 ha. The initial planting space of the plantation is shown in

Table 1. Summary of plantation types and their mean characteristics.

Plantation Types	Area (ha)	Initial Spacing	Plantation Density (n/ha)	Age (a)	Mean DBH (cm)	Mean Tree Height (m)	Mean Sectional Area (m2 ha−1)	Soil Bulk Density (g cm−3)	Canopy Density
SS	11.61	2 m × 2 m	1870 ± 378	30	11.65 ± 5.86	15.82 ± 4.77	19.97 ±1.17	1.46	0.87 ± 0.01
LF	4.77	2 m × 1 m	1655 ± 322	30	11.82 ± 5.81	15.36 ± 4.73	17.70 ± 3.24	1.40	0.63 ± 0.08
PM	17.16	2 m × 1.5 m	1850 ± 362	30	12.81 ± 5.73	15.21 ± 4.98	17.16 ± 2.56	1.24	0.50 ± 0.16
PE	17.74	2 m × 3 m	815 ± 368	30	16.14 ± 7.55	15.23 ± 5.23	18.45 ± 2.83	1.45	0.56 ± 0.04
MEB	20.27	2 m × 2 m	1310 ± 381	30	15.69 ± 7.81	14.97 ± 4.89	26.02 ± 1.67	1.41	0.72 ± 0.13
CK	-	-	-	30	-	-	-	1.44	-

SS, Schima superba plantation; LF, Liquidambar formosana plantation; PM, Pinus massoniana plantation; PE, Pinus elliottii plantation; MEB, P. elliottii and broadleaf mixed plantation; CK, experimental control check. The same applies below.

. There was no additional management of these plantations except for the watering and weeding of trees at the beginning of planting to improve the survival rate. Afterwards, it was allowed to grow naturally without any human management activities.

2.2. Data Collection

In July 2022, we selected five stands each from the Schima superba (SS), Liquidambar formosana (LF), Pinus massoniana (PM), Pinus elliottii (PE), P. eliottii and broadleaf mixed plantations (MEB) and unforested land in different hills in the FREP. SS was designated as an evergreen broad-leaved plantation, LF was designated as a deciduous broad-leaved plantation, PM was designated as a coniferous plantation, PE was designated as a coniferous plantation, and MEB was designated as a coniferous broad-leaved plantation [27]. The unforested land with no planting by humans and no tree establishment was used as an experimental control check (CK) [27].

In each stand, we set up a 20 m × 20 m plot for vegetation and soil sampling, with a total of 30 plots sampled [32]. The spatial distribution of the sampling plots is shown in Figure 1, and the plots were away from adjacent plots (at least 100 m). The basic information of each plantation type is shown in Table 1. In each plot, five subplots of 2 m × 2 m were placed at the four corners and at the center of each major plot to enumerate the shrubs and small regeneration trees. To measure the litter and herbaceous biomass, a 1 m × 1 m subplot was placed inside each shrub plot [27]. A total of 150 shrub and herb plots, and 150 litter plots were used to record the shrub, herb, and litter biomass data. The fresh weights of the herbs were obtained by harvesting and weighing them directly in the field. The plot location, tree species, diameter at breast height, and tree height were recorded. Trees with a diameter at breast height ≤ 5 cm were considered small regeneration trees [33]. The plant species, basal diameter, height, and coverage were recorded for each subplot. Subsamples were taken back to the laboratory and oven-dried to a constant weight at 65 °C. The water contents were calculated and used to convert the fresh weight to the dry weight. The Simpson dominance index (SDI), Shannon–Wiener diversity index (SWDI), Pielou’s evenness index (PEI), dominance index (DI), and Margalef’s richness index (MRI) were used to measure the plant diversity at each sample location [34].

Five “S”-shaped undisturbed soil samples were randomly and uniformly taken from each sample plot (20 m × 20 m) [32]. The soil samples were then gently and uniformly combined to generate a composite sample for each soil layer (0–10 cm, 10–20 cm, 20–30 cm, and 30–40 cm). Each sample was delivered to the laboratory in a refrigerated container after being carefully packaged in aluminium sample boxes. To analyse the chemical parameters of the soil, all 120 samples were cleared of roots and rocks, allowed to air dry, and then partially sieved using a 2 mm sieve. Moreover, bulk soil from each soil layer was also collected with a 100 cm³ ring cutter to measure the bulk density of each sample [32]. The ring knife method was used to determine the bulk density and porosity of the soil, while the drying method was used to determine the moisture content of the soil [28]. The soil pH was determined in a 1:5 (v/v) soil/water suspension using a digital pH electrode. The SOC was determined using the K₂Cr₂O₇-H₂SO₄ (Jiangxi Ganyi Instrument Co., Ltd., Ganzhou, China) Walkley–Black oxidation–titration method. The HCLO₄-H₂SO₄ (Jiangxi Ganyi Instrument Co., Ltd., Ganzhou, China)decoction method was used on a fully automated intermittent chemical analyser (Smart Chem 200 Alliance Corp., Melbourne, VIC, Australia) to determine the total nitrogen and phosphorus [35].

2.3. Estimation of Biomass and Soil Carbon Storage

The standing tree biomass model and carbon measuring parameter standards in China [36,37,38,39] were used to estimate the amount of biomass production in the stem, bark, branches, leaves, aboveground, and belowground parts of each sample plot. The formula and corresponding parameters are as follows (

Table 2. Biomass equations of the main tree species.

Species	MA = a D b H c			MB = a D b H c			${\mathbf{g}}_1$ = a D b H c			${\mathbf{g}}_2$ = a D b H c			${\mathbf{g}}_3$ = a D b H c
	a	b	c	a	b	c	a	b	c	a	b	c	a	b	c
S. superba	0.12045	2.06446	0.38265	0.08177	2.32395	−0.24289	0.75547	0	−0.49904	0.94206	0.26840	−0.74914	1.00699	−0.24787	−0.64570
L. formosana	0.08909	2.25544	0.30414	0.12052	2.42178	−0.40370	0.23463	0.30953	−0.47515	0.85343	0.69115	−1.00246	0.77866	0	−0.78326
P. massoniana	0.66615	2.09317	0.49763	0.00883	2.73828	−0.08026	0.48729	0	−0.60207	1.59113	0.96127	−1.80294	2.98814	0.61586	−2.02349
P.elliottii	0.04744	2.10359	0.63108	0.03526	2.10359	0.23419	0.84286	−0.16538	−0.39610	0.76157	0.58421	−1.03111	3.48107	0	−1.18444

M_A, Aboveground biomass; M_B, belowground biomass; D, diameter at breast height; H, tree height; a, b, c are measurement parameters; M_total = M_A + M_B.

and

Table 3. Carbon content conversion coefficient of the main tree species.

Species	Carbon Content Conversion Coefficient
	Stem	Bark	Branch	Leaves	Aboveground	Belowground
Schima superba	0.4683	0.4738	0.4712	0.4879	0.4712	0.4677
Liquidambar formosana	0.4737	0.4487	0.4694	0.4583	0.4690	0.4604
Pinus massoniana	0.5186	0.4994	0.5174	0.5785	0.5254	0.5082
Pinus elliottii	0.4744	0.4612	0.4868	0.4810	0.4756	0.4664

).

$${\mathrm{M}}_1=\frac{1}{{1+\mathrm{g}}_1+{\mathrm{g}}_2+{\mathrm{g}}_3}\times {\mathrm{M}}_{\mathrm{A}}$$

$${\mathrm{M}}_2=\frac{{\mathrm{g}}_1}{{1+\mathrm{g}}_1+{\mathrm{g}}_2+{\mathrm{g}}_3}\times {\mathrm{M}}_{\mathrm{A}}$$

$${\mathrm{M}}_3=\frac{{\mathrm{g}}_2}{{1+\mathrm{g}}_1+{\mathrm{g}}_2+{\mathrm{g}}_3}\times {\mathrm{M}}_{\mathrm{A}}$$

$${\mathrm{M}}_4=\frac{{\mathrm{g}}_3}{{1+\mathrm{g}}_1+{\mathrm{g}}_2+{\mathrm{g}}_3}\times {\mathrm{M}}_{\mathrm{A}}$$

where M₁, M₂, M₃, and M₄ are the biomass of stem, bark, branches, and leaves, respectively, in kilograms (kg); g₁, g₂, and g₃ are the proportional functions of the bark, branches, and leaves relative to a stem biomass, respectively, of 1; and M_A is the estimated value of the aboveground biomass.

The conversion coefficients come from the standards used in China.

To estimate the shrub biomass production, every new shrub in a 2 m × 2 m quadrat had to be uprooted first. After each harvest, the aboveground and belowground material was separated into various categories and weighed in the field. The samples were transported to the laboratory as a mixed sample weighing no less than 300 g from each belowground section. To determine the amount of biomass production in the herbs and litter, all the litter in each 1 m × 1 m quadrat was collected, weighed in the field, and mixed thoroughly. At least 300 g of each sample was removed and subsequently transported to the laboratory. Next, fresh herbs were removed from every 1 m × 1 m quadrat, separated into aboveground and belowground components, and weighed in the field. At least 100 g of the aboveground and belowground sections of each plant in each plot were collected. Subsamples (shrubs, herbs, and litter) were delivered to the laboratory and subjected to two processes: first, they were weighed, sampled twice, and dried at 65 °C to calculate their biomass; second, they were ground and crushed, and the carbon content was measured. The biomass production was calculated by multiplying the biomass by the measured carbon content. The shrub and herb biomass production is the sum of the aboveground and underground biomass production. Every component of the vegetation biomass production was calculated and scaled per hectare.

The soil carbon storage was determined using the following formula recommended by Chen et al. [40]:

SOC storage = SOC × BD × T × (1 – F) × 10⁻¹

where SOC is the soil organic carbon (g kg⁻¹), BD is the bulk density (g m⁻³), T is the soil thickness (cm), and F is the mass percentage (%) of gravel > 2 mm. the unit conversion coefficient is 10⁻¹.

The total soil carbon storage was calculated by adding the values of soil carbon storage from the four soil layers [1,27]. Ecosystem carbon storage was calculated by adding the vegetation biomass production (trees, shrubs, herbs, and litter) and total soil carbon storage [41].

2.4. Statistical Analysis

The Shapiro–Wilk test (data normality) and Bartlett’s test for the homogeneity of variance were used to test the normality of all variables. Then, one-way ANOVA was used to analyse the changes in vegetation biomass production, soil carbon storage, and ecosystem carbon storage under the different plantation types, and multiple comparisons (least significant difference (LSD)) were used to determine the differences in vegetation biomass production, soil carbon storage, and ecosystem carbon storage (p < 0.05). The Pearson correlation coefficient and redundancy analysis (RDA) were used to explore the main factors affecting vegetation biomass production, soil carbon storage, and ecosystem carbon storage. The data analysis was performed using SPSS 27.0 (IBM, Armonk, NY, USA), correlation heatmaps were created using Origin 2023b (OriginLab, Northampton, MA, USA), the RDA was performed using Canoco 5 (Microcomputer Power, Ithaca, NY, USA), and the study region overview map was created using ArcGIS 10.8 (ESRI, Redlands, CA, USA).

3. Results

3.1. Tree Biomass Production

Our study revealed significant differences (p < 0.05) in the tree biomass production of the different plantation ecosystems (

Table 4. Tree biomass production (Mg ha⁻¹).

Plantation Types	Tree Biomass Production
	Stem	Bark	Branch	Leaves	Aboveground	Belowground	Total
SS	54.80 ± 10.23 a	9.27 ± 1.53 a	12.22 ± 1.67 a	4.31 ± 0.66 a	80.73 ± 13.74 a	17.99 ± 2.42 ab	98.72 ± 16.07 a
LF	38.91 ± 2.99 b	5.32 ± 0.52 b	12.05 ± 0.59 a	3.03 ± 0.25 b	59.25 ± 4.14 b	15.58 ± 0.82 bc	74.83 ± 4.83 bc
PM	55.93 ± 12.37 a	6.06 ± 1.78 b	9.18 ± 2.88 b	3.80 ± 1.41 ab	76.89 ± 15.34 a	12.86 ± 3.29 cd	89.75 ± 17.63 ab
PE	32.21 ± 15.20 b	5.02 ± 1.94 b	8.35 ± 2.99 b	3.88 ± 1.15 ab	49.59 ± 20.93 b	11.65 ± 4.27 d	61.24 ± 25.17 c
MEB	49.85 ± 9.46a	8.60 ± 2.23 a	12.65 ± 2.73 a	4.67 ± 1.57 a	76.35 ± 15.05 a	19.88 ± 4.30 a	96.22 ± 18.11 a
CK	0 c	0 c	0 c	0 c	0 d	0 e	0 d

The values are presented as the means ± standard deviations. Different letters in the same column indicate a significant difference between different plantation types at p < 0.05. The same applies below.

). The total amount of aboveground tree biomass production of the various plantation ecosystems ranged from 49.59 Mg ha⁻¹ to 80.73 Mg ha⁻¹, among which that of SS was the highest (80.73 Mg ha⁻¹), followed by PM (76.89 Mg ha⁻¹), MEB (76.35 Mg ha⁻¹), and PE (49.59 Mg ha⁻¹). The minimum belowground tree biomass production was found in MEB (19.88 Mg ha⁻¹), while the lowest amount was found in PE (11.65 Mg ha⁻¹). The total tree biomass production of the various plantation ecosystems followed a similar pattern to that of the aboveground tree biomass production; SS had a relatively high tree biomass production (98.72 Mg ha⁻¹).

3.2. Shrub and Rest of Vegetation Biomass Production

Our study found significant differences (p < 0.05) in biomass production between the shrubs and the rest of the vegetation in terms of different components of the plantation ecosystem under the different plantation types (

Table 5. Shrub biomass production and that of the rest of the vegetation (Mg ha⁻¹).

Plantation Types	Shrubs and Herbs Biomass Production			Litter Biomass Production	Vegetation Biomass Production (Tree + Shrubs/Herbs + Litter)
	Aboveground	Belowground	Total
SS	0.48 ± 0.18 ab	0.28 ± 0.09 a	0.75 ± 0.27 ab	1.96 ± 0.16 c	101.73 ± 16.26 a
LF	0.51 ± 0.10 ab	0.25 ± 0.04 a	0.76 ± 0.12 ab	2.05 ± 0.27 c	77.42 ± 4.77 b
PM	0.36 ± 0.07 b	0.24 ± 0.04 a	0.60 ± 0.11 ab	2.66 ± 0.33 bc	93.71 ± 18.68 a
PE	0.75 ± 0.21 a	0.29 ± 0.06 a	1.05 ± 0.27 a	3.21 ± 0.27 ab	65.73 ± 24.65 c
MEB	0.26 ± 0.04 b	0.24 ± 0.05 a	0.50 ± 0.08 b	3.66 ± 0.26 a	100.38 ± 19.07 a
CK	0.20 ± 0.03 c	0.16 ± 0.02 b	0.36 ± 0.04 c	0.98 ± 0.03 d	1.36 ± 0.04 d

The values are presented as the means ± standard deviations. Different letters in the same column indicate a significant difference between different plantation types at p < 0.05. The same applies below.

). The Shrub and herb biomass production varied significantly between plantation ecosystems, ranging from 0.50 Mg ha⁻¹ for MEB to 1.05 Mg ha⁻¹ for PE. The study’s findings also demonstrated that the amount of biomass production in the litter varied significantly among the recovered plantations (p < 0.05). The litter biomass production of the five plantations ranged from 1.96 to 3.66 Mg ha⁻¹, with a sequential decrease in the following order: MEB, PE, PM, LF, and SS.

Overall, the vegetation biomass production in the plantations was significantly greater than that in the CK treatment (p < 0.05), specifically SS (101.73 Mg ha⁻¹) > MEB (100.38 Mg ha⁻¹) > PM (93.71 Mg ha⁻¹) > LF (77.42 Mg ha⁻¹) > PE (65.73 Mg ha⁻¹), which increased by 738.1%, 728.1%, 690.1%, 559.3%, and 473.3%, respectively, compared with that in the CK treatment (1.36 Mg ha⁻¹). Furthermore, more than 90% of the vegetation biomass production enhanced by plantations was found in the tree layer. The grass and shrub layers had the lowest increase in biomass production, ranging from 0.50 to 1.60%.

3.3. Soil Carbon Storage

Table 6. Soil carbon storage (Mg ha⁻¹).

Plantation Types	Soil Depth (cm)				Total
	0–10	10–20	20–30	30–40
SS	50.03 ± 10.85 a	26.08 ± 10.33 a	19.06 ± 8.34 a	14.81 ± 8.63 a	109.97 ± 18.90 a
LF	39.35 ± 6.44 ab	18.10 ± 2.68 b	15.11 ± 2.49 ab	10.97 ± 2.25 abc	83.53 ± 8.71 bc
PM	28.26 ± 5.15 b	16.42 ± 3.01 b	12.54 ± 4.99 bc	10.26 ± 4.93 bc	67.47 ± 21.48 c
PE	31.45 ± 13.63 b	17.68 ± 5.47 b	12.73 ± 3.35 bc	9.11 ± 2.87 bc	70.97 ± 21.39 c
MEB	48.25 ± 8.81 a	21.11 ± 7.95 ab	16.16 ± 6.92 ab	13.13 ± 6.54 ab	98.64 ± 12.95 ab
CK	11.30 ± 2.38 c	9.98 ± 2.43 c	9.47 ± 2.32 c	7.34 ± 1.76 c	38.08 ± 8.88 d

Different letters in the same column indicate a significant difference between different plantation types at p < 0.05.

shows significant (p < 0.05) variations in the soil carbon storage of the different plantation types. Across all the soil depths (0–10 cm, 10–20 cm, 20–30 cm, and 30–40 cm), SS had the largest amount of soil carbon storage (50.03 Mg ha⁻¹, 26.08 Mg ha⁻¹, 19.06 Mg ha⁻¹, and 14.81 Mg ha⁻¹), followed by MEB (48.25 Mg ha⁻¹, 21.11 Mg ha⁻¹, 16.16 Mg ha⁻¹, and 13.13 Mg ha⁻¹). In the three soil depth ranges of 0–10 cm, 10–20 cm, and 20–30 cm, PM had the lowest soil carbon storage (28.23 Mg ha⁻¹, 16.42 Mg ha⁻¹, and 12.54 Mg ha⁻¹), and in the soil depth range of 30–40 cm, PE had the lowest soil carbon storage (9.11 Mg ha⁻¹). Furthermore, the soil carbon storage in SS and MEB at all soil depths and in LF at the 0–10, 10–20, and 20–30 cm soil depths was significantly greater than that in the CK (p < 0.05). Only at the 0–10 cm and 10–20 cm soil depths did there appear to be significant increases (p < 0.05) in PM and PE. In general, the soil carbon storage was ranked as follows: SS (109.97 Mg ha⁻¹) > MEB (98.64 Mg ha⁻¹) > LF (83.53 Mg ha⁻¹) > PE (70.97 Mg ha⁻¹) > PM (67.47 Mg ha⁻¹). These values are increased by 188.8%, 159.1%, 119.4%, 86.4% and 77.2%, respectively, compared with the CK (38.08 Mg ha⁻¹).

3.4. Ecosystem Carbon Storage

Table 7. Ecosystem carbon storage (Mg ha⁻¹).

Plantation Types	Vegetation Biomass Production	Soil Carbon Storage	Ecosystem Carbon Storage
SS	101.73 ± 16.26 a	109.97 ± 18.90 a	211.71 ± 55.24 a
LF	77.42 ± 4.77 b	83.53 ± 8.71 bc	160.96 ± 5.81 b
PM	93.71 ±18.68 a	67.47 ± 21.48 c	155.01 ± 30.99 b
PE	65.73 ± 24.65 c	70.97 ± 21.39 c	142.88 ± 34.6 b
MEB	100.38 ± 19.07 a	98.64 ± 12.95 ab	199.02 ± 52.68 a
CK	1.36 ± 0.04 d	38.08 ± 8.88 d	39.44 ± 7.39 c

Different letters in the same column indicate a significant difference between different vegetation plantation types at p < 0.05.

shows the ecosystem carbon storage was significantly higher than those in the CK (p < 0.05). The ecosystem carbon storage followed the trend SS (211.71 Mg ha⁻¹) > MEB (199.02 Mg ha⁻¹) > LF (160.96 Mg ha⁻¹) > PM (155.01 Mg ha⁻¹) > PE (142.88 Mg ha⁻¹), and compared with CK (39.44 Mg ha⁻¹), the storage amounts were higher by 436.8%, 404.6%, 308.1%, 293.1%, and 262.3%, respectively. This study also revealed that the vertical distribution characteristics of the ecosystem carbon storage in the four plantations (SS, LF, PE, and MEB) were 0–40 cm soil layer > tree layer > litter layer > shrub and herb layer; however, the vertical distribution of the ecosystem carbon storage of PM was as follows: tree layer > 0–40 cm soil layer > litter layer > shrub and herb layer.

3.5. Correlations with Environmental Indices

The linear relationships between vegetation biomass production, soil carbon storage, ecosystem carbon storage and other ecosystem indicators were determined by the Pearson correlation analysis (Figure 3). The results revealed that soil porosity and stand density were positively correlated with the ECS, VBP, and SCS; moreover, the pH was negatively correlated with the ECS, VBP, and SCS. BD was significantly negatively correlated with the VBP and ECD. TN and MRI were significantly positively correlated with the SCS and ECD. SWDI and PEI were negatively correlated with the SCS and ECD.

To learn more about the variables influencing the ecosystem carbon storage, RDA was employed. Figure 4 displays the RDA results. Together, axis 1 and axis 2 accounted for 71.07% of the variation in the ecosystem carbon storage, with axis 1 accounting for 65.4% of the total variation. Stand density (32.3%), SWDI (11.4%), BD (6.8%), and SWC (5.7%) were the main factors influencing the ecosystem carbon storage. The results of the RDA provide a reference for reducing the dimensionality of the data, and variables without significant correlations can be indexed to calculate their contribution to the ecosystem carbon storage.

4. Discussion

4.1. Ecosystem Carbon Storage

There is enormous potential for sequestering carbon in the steep, deteriorated red soil regions of southern China. In the red soil erosion hilly area of Jiangxi, China, we found that the carbon storage of reforestation plantings ecosystems was 142.88 Mg ha⁻¹–211.71 Mg ha⁻¹. Liu et al. [42] estimated that the carbon storage of the Pinus massoniana plantation ecosystem in the degraded red soil hilly area of Changting, Fujian Province, China, after 30 years of reconstruction was 107.34 Mg ha⁻¹, while Preston et al. [43] estimated that the carbon storage of the Pinus koraiensis plantation ecosystem after a 16-year reforestation project of degraded soil in an industrial park in northeastern Canada was 90.2 Mg ha⁻¹. Sahoo et al. [2] estimated the carbon storage of the subtropical plantation ecosystem in northeastern India to be 158.03 Mg ha⁻¹ through a biomass model. This is because the typical reforestation paradigm involves the use of high-survival-rate pioneer species to colonize the landscape, followed by the gradual introduction of local apex species when environmental conditions allow for it. As for the current study area, however, there are few examples of creative direct references to local apex species [27] for successfully restoring severely eroded and degraded areas of soils. The successful reforestation in the study area is significant because it serves as a model for large-scale afforestation and reforestation projects in the red soil erosion areas of southern China. Additionally, it was discovered that the four types of plantations—SS, LF, PE, and MEB—had a vertical ecosystem carbon storage distribution that was similar to the findings of Wang et al. [44] and were characterized by soil layers ranging from 0 to 40 cm, followed by tree layers, litter layers, and shrub–herb layers. This could be because less biomass carbon is stored during reforestation because the plants have not yet reached maturity [45]; however, the tree layer > 0–40 cm soil layer > litter layer > shrub and herb layer had a vertical ecosystem carbon storage distribution in PM, which is in line with the findings of Liu et al. [42], who showed that only the top layer of the soil increased in carbon storage and that reforestation in Pinus massoniana plantations led to a slower increase in soil carbon pools and a faster recovery of tree growth.

Table 5 shows that the observed vegetation biomass production varied from 1.36 Mg ha⁻¹ under unforested land to 101.73 Mg ha⁻¹ under SS. Because more tree species have been added to reforestation plantings, more biomass has been allocated to the ground by plants as a result [44]. This is in contrast to unforested land, where biomass was still low three to four years after anthropogenic disturbances stopped, despite the rapid growth and development of scrubby grasses. These findings suggest that reforestation plantings has the potential to significantly increase vegetation biomass production. According to the present study, the tree layer accounted for more than 90% of the overall increase in vegetative biomass production caused by reforestation plantings, with the litter layer occurring secondarily. The shrub layer experienced the least increase in biomass production, contributing only 0.50%–1.60% of the total. This is due to the comparatively low levels of soil nutrients in the regenerated forests found in the red loamy hill erosion regions of southern China, as well as the structural homogeneity of the understorey plant communities, which are dominated by herbaceous plants such as Gleichenia linearis. The understorey vegetation of reforestation plantings experienced a minimal increase in biomass production during the growth process because of the sparse understorey vegetation, such as understorey shrubs and grasses. This finding is generally consistent with that of other studies [41,42] that found that the stem and the belowground root system were the primary plant carbon stores. Increased biomass production was found primarily in the stem, followed by the branches, roots, and bark and leaves. Notably, SS, MEB, and PM played dominant roles in biomass production. This study showed that the biomass production of SS was 74.8 times greater than that of the CK and 1.5 times greater than that of PE. Geographical location, plantation age, species composition [30,46], species richness [47], stand density, leaf area index, canopy design, and the genetic makeup of different trees [1,48] were all implicated in these variations in biomass production. Although the geographical location and plantation age of this study were identical, the structural makeup of the plantations and their genetic makeup accounted for the majority of the variation in biomass production.

Soil stores large amounts of organic carbon, which helps maintain soil quality and the carbon cycle. The organic carbon stored in the top 40 cm of soil in the reforestation plantings amounted to 67.47 Mg ha⁻¹–109.97 Mg ha⁻¹, which was equivalent to 44%–52% of the ecosystem carbon storage. Notably, prior research conducted in the study area ten years ago [49] revealed lower levels of organic carbon in the soil layer of reforestation plantings soils at a depth of 0–40 cm than in unforested land; however, subsequent research conducted in the study area (17 years [27] and 30 years) revealed greater SOC storage in reforestation plantings than in unforested land. This is due to the slow decomposition of soil carbon in the early stage of reforestation, which is an important reason for the increase in organic carbon in the degraded red soil area in southern China [50]; however, with the increase in the reforestation period, the total biomass of the reforestation plantings increased, the amount of existing surface litter gradually increased, the quality of the litter layer improved, the litter layer decomposed faster, and the input of organic matter increased [51]. Additionally, the area of bare land decreased, and the erosion caused by surface runoff decreased; therefore, the quality of the soil environment improved and stabilized gradually, and the enhancement of the physical protection of soil aggregates slowed the loss of soil SOC [52]. It was also found that soil carbon storage was significantly greater (p < 0.05) in SS than in PM and PE. The decomposition rates of the PE and PM litter were low due to the large quantities of refractory chemicals they contained, which also slowed the conversion of particulate organic matter to mineral soil. In contrast, SS litter broke down more quickly [53]. In addition, SS can allocate more biomass to its roots, especially fine roots, which can fix more carbon and transfer more root debris to the soil [54]. In addition, the well-developed and widely dispersed belowground root system of SS enhances the porosity and storing-water function of the soil [55], boosts plant output, and slows the rate at which organic carbon is mineralized [10]. As the soil depth increases, the degree of the effect of the plant litter and root systems decreases since deep soil is formed from the same parent material [51]. As a result, the variation in SOC content between the various plantation types decreased as the soil depth increased. A reduction in microbial diversity and abundance caused by litter monoculture results in significantly greater rates of residue decomposition [56]. Therefore, while broadleaf plantations have greater litter volumes, soil microbial activity is greater in mixed plantations, with faster rates of litter and belowground root system decomposition. This means that there were no significant differences (p < 0.05) in SOC storage among MEB, SS, or LF within the 0 to 40 cm range.

4.2. Factors Influencing Ecosystem Carbon Storage

Plantations absorb significant amounts of atmospheric CO₂ and play a vital role in reducing climate change. Therefore, to manage and improve ecosystem carbon pools in constantly changing environments, it is essential to understand the possible synergies between plantation carbon storage and environmental parameters [33]. Similar to the study by Gogoi et al. [33], the ecosystem carbon storage was significantly positively correlated with soil porosity, TN, and plantation density. The rate at which biomass carbon storage expands is accelerated by a higher plantation density. A higher porosity of the soil facilitates the movement and exchange of gases and water, which help to better retain the particulate matter in organic carbon, alter particulate organic carbon, and encourage the build-up of organic carbon [32]. An increased soil N content may increase plant productivity and carbon sequestration in soil [31]. In forestry, as soil acidity increases, the soil pH gradually decreases, the soil becomes compacted, the bulk density increases, the soil structure is destroyed, and the nutrient storage capacity of the soil decreases. Therefore, an increase in pH is associated with a decrease in ecosystem carbon storage [57]. Notably, the ecosystem carbon storage was found to be significantly negatively correlated with both the PEI and SWDI in this study. These findings align with those from a temperate forest study conducted by Suo et al. [25]; however, the majority of prior research has demonstrated that a greater diversity of plant life can have a positive impact on soil carbon storage, which in turn influences ecosystem carbon storage [33,58]. This may be due to the effects of competition for limited resources among species or the inclusion of low-productivity species in species-rich communities [59]. After several years of rehabilitation, the CK in this study contained lush grasses and shrubs with high species richness; nevertheless, due to the absence of an arboreal layer, the ecosystem carbon storage was low. The primary variables governing the capacity of the restored plant ecosystem to store carbon in the degraded red soil region of southern China were further determined via redundancy analysis. The RDA results indicated that the ecosystem carbon storage was mainly affected by the combined effects of stand density, SWDI, BD, and SWC. Increased stand density directly affects both aboveground and belowground carbon pools and is one of the main factors determining ecosystem carbon storage, SWC supporting plant growth, and microbial activity. A study conducted in the Eastern Himalayas in India on six major forest types revealed that stand density and SWC had significant impacts on the ecosystem carbon storage [33].

4.3. Implication for Carbon Sink Management

Approximately two-thirds of terrestrial carbon is accounted for by the yearly increase in carbon stored in forests, which accounts for approximately 80% of the total carbon in terrestrial vegetation. With their ability to regulate the climate among other ecosystem services, they rank among the most significant terrestrial ecosystems. The southern red soil hilly region has a total water and soil loss area of 131,200 km², or 15% of the overall land area according to monitoring results from the Chinese Ministry of Water Resources. After the Loess Plateau, the southern red soil hilly region now the second-largest area in China where soil and water are lost [60]. There has been significant damage to natural vegetation in the red soil hilly areas of the southern region due to specific natural conditions, such as high population density, prominent man–land conflicts, staggered mountains and hills, undulating terrains, heavy and concentrated rainfall, heavy rainstorms, and strong weathering. Furthermore, there are numerous development and building projects of various kinds, a rapid increase in the area of newly created human-caused soil and water erosion, and a rapid pace of socioeconomic development in the region. Soil erosion has emerged as a prevalent hindrance to regional production growth, farmer prosperity and poverty alleviation, ecological environment enhancement, land remediation and protection, and improvements in people’s quality of life. Therefore, reforestation in degraded red soil areas in southern China is crucial for preventing soil erosion and sequestering carbon. According to our research, replanting in damaged red soil regions can greatly increase the ecosystem carbon storage. This is because replacing plantation vegetation on degraded land can progressively enhance vegetation coverage, which can reduce runoff and soil loss by causing the accumulation of litter mass, the development of root networks, and an improvement in the physical and chemical qualities of the soil [41]. Furthermore, studies indicate that planting broad-leaved tree species or planting them in conjunction with native coniferous species is the optimal approach for future reforestation in degraded red soil areas. Concurrently, efforts must be directed towards establishing a sensible density for afforestation, enhancing the soil’s physical composition, and augmenting its porosity to enhance and reinstate the plantation ecosystem carbon storage. Enhancing global carbon accounting, creating more effective climate policies, and forecasting future climate change all depend on the quantitative reconstruction of biomass and carbon storage in reforestation plantings ecosystems. Moreover, it can also make positive contributions to the United Nations Framework Convention on Climate Change’s plantation protection agenda, sustainable plantation management, and increasing plantation carbon storage in developing countries such as China (REDD⁺).

This study has several limitations, and further extensive sampling in additional regions will be necessary to enhance the dependability of the findings and thus broaden their applicability [61].

5. Conclusions

Thirty years after reforestation, the FREP experiment significantly increased all major carbon pools in barren degraded landscapes, but there is still a large potential for increasing carbon sinks. The ecosystem carbon storage of the different reforestation plantings in the study area were significantly greater than those of the CK. Additionally, there were differences in ecosystem carbon storage between different plantations. SS and MEB had greater ecosystem carbon storage. This could be because SS and MEB had a higher potential growth rates and were better suited to the environmental conditions in the red soil of southern China. Therefore, it is best to plant broadleaf tree species or combine them with native coniferous species to improve organic carbon sequestration and ecological environment rehabilitation. Rebuilding degraded red soil can alter the way plantation ecosystems store carbon. This complex process is influenced primarily by SWC, BD, SWDI, and stand density. Overall, better land management choices, lower carbon dioxide emissions, and climate change mitigation in other vulnerable ecosystems, such as the red soil erosion zones in southern China, can be achieved through the knowledge of the carbon storage of regenerated plantation ecosystems.