Vegetation Restoration Significantly Improved Soil Aggregate Stability in the East Qinling Mountains

Abstract

Although plant restoration is essential for improving soil structure and stability, there are still few systematic assessments of its impacts across various restored vegetation species, especially in environmentally sensitive areas like the East Qinling Mountains. In order to provide a scientific foundation for optimizing restoration tactics and enhancing soil erosion control and ecosystem services in the area, this study attempts to assess the impacts of different recovered plant types on soil aggregate stability and to clarify the underlying mechanisms. The Pinus tabuliformis Carrière, Quercus variabilis Blume, Robinia pseudoacacia L., Pinus tabulaeformis-Quercus variabilis mixed forest, Platycladus orientalis (L.) Franco and abandoned grassland were the six vegetation types represented by the sixteen plots. Farmland was used as a control. Soil samples were taken from three depths (0–5 cm, 5–20 cm, and 20–40 cm) and evaluated for root biomass, soil organic matter (SOM), and water-stable aggregate dispersion. Mean weight diameter (MWD), fractal dimension (D), macroaggregate content of diameter > 0.25 mm (R_0.25), and percentage of aggregate disruption (PAD) were used to evaluate aggregate stability. One-way ANOVA, LSD multiple comparisons, and Spearman correlation analysis were among the statistical analyses. In comparison to grassland and farming, forested regions, particularly mixed forests, showed considerably higher proportions of macroaggregates (>0.25 mm) and superior aggregate stability (higher MWD and R_0.25, lower D and PAD). Increased litter and coarse root inputs, which encouraged big water-stable aggregates (WSAs) and reinforced their positive connection with SOM, were the driving forces behind this development. Robinia pseudoacacia L. and Platycladus orientalis (L.) Franco displayed the highest SOM concentration and root biomass (1201.45 and 679.66 g/m², respectively). At all depths, mixed forests showed the most stable soil structure. In contrast to agriculture, vegetation restoration dramatically changed the mechanical composition of the soil, increasing the differentiation of particle-size fractions across soil layers and decreasing the amount of surface clay. Soil aggregate stability is greatly enhanced by vegetation restoration, with mixed forests offering the greatest advantages because of their varied root systems and increased input of organic matter. These results emphasize how crucial it is to choose the right vegetation types for restoration efforts in order to improve soil structure, reduce erosion, and promote ecological sustainability in the East Qinling Mountains.

1. Introduction

The formation and stabilization of soil aggregates are closely related to vegetation restoration. Vegetation restoration enhances soil organic matter content via root exudates, litter decomposition, and other pathways, thereby promoting the formation and stabilization of soil aggregates [1]. The stability of soil aggregates directly affects soil structure stability, which is crucial for controlling soil erosion, enhancing carbon storage, and maintaining ecosystem health [2]. Consequently, studying soil aggregates is indispensable for comprehending the structure and function of soil ecosystems, improving soil quality, and sustaining ecosystem services. In recent years, with the intensification of global climate change and human activities, soil degradation has become increasingly severe, prompting greater attention to research on soil aggregates.

As a critical ecological barrier and water conservation area in China, the stability and erosion resistance of soil structure in the Qinling Mountains significantly influence regional ecological security and water resource safety [3]. However, due to the complex topography, concentrated rainfall, and human activities in the Qinling Mountains, soil degradation and erosion are severe challenges in some areas [4]. Soil aggregates, as the fundamental units of soil structure, directly affect soil erosion resistance and water retention capacity. Therefore, research on soil aggregates in the Qinling Mountains helps elucidate the effects of different vegetation types on soil structure and function, which is essential for evaluating soil erosion mechanisms, assessing vegetation restoration effects, and formulating soil and water conservation strategies.

Previous studies have extensively investigated the distribution, ecological stoichiometry, carbon sequestration function, water retention effects, and stability mechanisms of soil aggregates under various vegetation restoration types in the Qinling Mountains and adjacent areas [5,6,7,8,9,10,11,12,13,14]. These studies have deepened our understanding of how vegetation restoration influences soil aggregates in the region, revealing the composition, stability mechanisms, and ecological processes of soil aggregates under typical restored vegetation. This provides a scientific foundation for optimizing regional vegetation restoration models, enhancing ecosystem services, and addressing climate change. However, despite these advancements, critical knowledge gaps remain. First, most existing studies have focused on the effects of single vegetation types or land-use patterns on surface soil aggregates (0–10 cm), with limited systematic comparisons across multiple restored vegetation types within the same ecological context. This lack of comparative analysis hinders the ability to identify which restoration strategies most effectively enhance soil structure, particularly in ecologically sensitive regions like the East Qinling Mountains. Second, the majority of research has concentrated on surface soil layers, leaving the dynamics of deeper soil aggregates (below 20 cm) largely unexplored. Given that deep root systems of certain vegetation types (e.g., Robinia pseudoacacia L. and mixed forests) can substantially influence subsoil structure through carbon input and bioturbation, understanding depth-dependent aggregate stability is essential for evaluating the full-profile effects of vegetation restoration. Third, while the roles of litter and root biomass in aggregate formation are well recognized, the differential contributions of fine roots versus coarse roots, as well as the interactive effects of soil texture and organic matter on aggregate stability, remain poorly quantified. Specifically, it is unclear whether vegetation-induced changes in soil mechanical composition actively contribute to aggregate stability or merely reflect pre-existing textural differences.

To address these gaps, this study systematically evaluates soil aggregate characteristics across six representative restored vegetation types (Pinus tabuliformis Carrière, Quercus variabilis Blume, Robinia pseudoacacia L., Pinus tabulaeformis-Quercus variabilis mixed forest, Platycladus orientalis (L.) Franco, and abandoned grassland) in the East Qinling Mountains, using adjacent farmland as a control. Soil samples were collected from three depth intervals (0–5 cm, 5–20 cm, and 20–40 cm) to capture both surface and subsoil dynamics. We hypothesize that: (1) vegetation restoration significantly improves soil aggregate stability compared to farmland, with mixed forests exhibiting the greatest enhancement due to complementary root systems and higher organic matter input; (2) the effects of vegetation restoration on aggregate stability extend below 20 cm, particularly for deep-rooted species; and (3) soil mechanical composition, specifically clay and silt content, interacts with organic matter to regulate aggregate stability, and vegetation restoration can actively modify textural properties over time.

By addressing these hypotheses, this study aims to: (1) compare the distribution and stability of soil aggregates among different restored vegetation types across soil depths; (2) quantify the relationships among root biomass, soil organic matter, soil texture, and aggregate stability indices; and (3) elucidate the mechanistic pathways through which vegetation restoration influences soil structure. The novelty of this study lies in its systematic, multi-species comparison across depths, its integration of root biomass fractions with textural analysis, and its explicit testing of whether vegetation-induced textural changes contribute to aggregate stability. The findings will provide a scientific basis for optimizing vegetation restoration strategies to enhance soil structure, control erosion, and promote ecological sustainability in the East Qinling Mountains.

2. Materials and Methods

2.1. Study Area

The study area is situated in the southeastern part of Shaanxi Province (E 108°34′20″–111°1′25″, N 33°2′30″–34°24′40″), at the southern foothills of the Qinling Mountains. This region serves as a critical water source protection zone for China’s South-to-North Water Diversion Project. The topography is complex, with higher elevations in the west and lower elevations in the east, dominated by mountains and hills. Most of the area has a warm temperate semi-humid monsoon climate, while the southern portion exhibits a subtropical monsoon climate. The region experiences four distinct seasons, with an average annual temperature ranging from 7.8 to 13.9 °C and annual precipitation between 700 mm and 900 mm. Water resources are abundant, with major rivers including the Danjiang River, Luohe River, and Jinqianhe River. Shangluo City has a high forest coverage rate, reaching 70.0% by 2024. From 2000 to 2015, vegetation coverage increased significantly at a rate of 2.77% per decade [15], particularly in areas with medium to high vegetation coverage. This improvement can be largely attributed to national ecological restoration projects, such as the Grain for Green Program and the Natural Forest Protection Program. Dominant tree species planted during vegetation restoration include Pinus tabuliformis Carrière, Platycladus orientalis (L.) Franco, and Robinia pseudoacacia L., while common herbaceous plants include Potentilla reptans L., Artemisia annua L., and Cynodon dactylon (L.) Pers. The soils in the study area are diverse, primarily classified according to the Chinese Genetic Classification. Yellow-brown soils, which are prevalent in mountainous and hilly areas, can be broadly correlated with Dystric Cambisols in the WRB classification or Typic Dystrudepts in USDA Soil Taxonomy. In low mountain and hill regions, cinnamon soils are commonly distributed; these roughly correspond to Calcaric Cambisols (WRB) or Haplustepts (USDA). Additionally, red soils, found in localized low-altitude zones, are analogous to Acrisols (WRB) or Typic Hapludults (USDA) [4].

2.2. Plot Setting and Vegetation Community Surveys

This study focuses on the southern foothills of the Eastern Qinling Mountains, a region where the vegetation ecosystem remains relatively undisturbed due to minimal human disturbance. Through systematic field surveys and investigations, and considering environmental factors such as slope aspect, slope position, and elevation, 17 typical vegetation plots were selected, including Pinus tabuliformis Carr. (3 plots), Quercus variabilis Blume (4 plots), Robinia pseudoacacia L. (3 plots), Pinus tabulaeformis-Quercus variabilis mixed forests (2 plots), Platycladus orientalis (L.) Franco (2 plots) and abandoned grassland (2 plots). Farmland was included as a control group for a comprehensive analysis of the influence of different vegetation types on ecological processes in the study area (Figure 1,

Table 1. Basic property of sample sites in the study area.

Stand Types and Sites	Forest Age	Slope Position	Slope (°)	Elevation (m)	Canopy Density (%)	Stand Density (hm−2)	Mean Height (m)	Mean DBH (cm)
Platycladus orientalis (L.) Franco (Two plots)	Middle-aged forest (21~40 ages)	Middle	19~24°	1023~1030	33~47	1740~2041	7.46~8.90	13.18~14.95
Pinus tabuliformis Carrière (Three plots)	Middle-aged forest (21~30 ages)	Middle- upper	35~39°	882~884	82~87	740~1775	3.60~13.90	7.58~17.95
Quercus variabilis Blume (Four plots)	Middle-aged forest (21~30 ages)	Middle- upper	20~40°	853~1019	33~80	714~2577	8.30~13.25	8.47~16.63
Pinus tabulaeformis—Quercus variabilis mixed forests (Two plots)	Mature forest (51~60 ages)	Upper	24~30°	1138~1213	67~80	1198~1785	8.00~8.32	10.63~14.01
Robinia pseudoacacia L. (Three plots)	Over mature forest (>30 ages)	Middle	33~34°	867~903	60~73	663~969	6.50~15.24	8.98~15.15
Grassland (Two plots)	(3~10 ages)	Middle- lower	8~15°	1168	—	—	—	—
Maize (Two plots)	—	Lower	23°	706	—	—	—	—

Notes: DBH stands for diameter at breast height, which is a standard measurement in forestry referring to the diameter of a tree trunk measured at 1.3 m above ground level, and “—” means missing data.

). The uneven distribution of plot numbers reflects the actual spatial heterogeneity and availability of representative vegetation stands in the study area. To ensure statistical robustness, all plots were treated as independent sampling units, and for each vegetation type, three replicate soil samples were collected from each depth interval within each plot. This hierarchical sampling design (plots within vegetation types, samples within plots) accounts for both inter-plot variability and within-plot heterogeneity. Based on soil texture uniformity and soil profile integrity, 16 plots were ultimately selected for further research on the characteristics of water-stable soil aggregates.

In October 2023, systematic surveys of vegetation community characteristics were conducted in standard plots of different typical vegetation types. During the plot surveys, following the plot inventory method and considering the actual conditions of the field sites and the characteristics of the vegetation types, quadrats were established for trees (15 m × 15 m), shrubs (5 m × 5 m), and grassland (1 m × 1 m) [16]. Within each quadrat, key indicators such as plant species (including trees and shrubs), quantity, height, and coverage were recorded in detail. Simultaneously, environmental factors such as topography and soil properties were measured, and the dominant herbaceous species under the forest canopy were surveyed and documented. To ensure comparability of results when analyzing data across different vegetation types, the average values of each indicator were calculated separately.

2.3. Methods

2.3.1. Determination of Vegetation Roots

The roots of the vegetation plots in the field were collected using the fixed-volume soil method. For shrub and forest communities, square plots measuring 20 cm × 20 cm were established, and a digging tool was used to vertically excavate soil blocks of fixed volume (20 cm × 20 cm ×40 cm), preserving the complete root structure. Excavation was performed 0.5 m away from the representative plant trunk to minimize spatial heterogeneity-induced differences. A total sampling depth of 40 cm was divided into three layers: 0–5 cm, 5–20 cm, and 20–40 cm. The soil and roots were gradually separated using the water washing method, debris was removed, and all roots were collected and sorted by diameter class (≤1 mm, 1–2 mm, 2–5 mm, and 5–10 mm) into corresponding paper bags. The roots were dried at 75 °C for 48 h until a constant weight was achieved, and their weights were recorded for subsequent analysis.

2.3.2. Determination of Soil Particle Composition

After air-drying the sample, fine roots and other impurities were manually removed. A representative portion of the sample was then weighed, passed through a 2 mm sieve, and subsequently soaked in an excess of hydrogen peroxide for 48 to 72 h to eliminate organic matter. Following this, the sample was immersed in an excess of dilute hydrochloric acid for 24 h to remove calcium carbonate. Lastly, a sodium hexametaphosphate solution (5 mL/L) was added to disperse the soil particles. Particle size analysis was performed using a MS2000 laser particle size analyzer (Malvern Instruments Co., Ltd., Malvern, UK) [17].

2.3.3. Soil Sample Collection

When conducting field sampling of soil aggregates, areas with uniformly distributed vegetation growth are initially selected to ensure the representativeness of the sampling sites. For grasslands and maize fields, two locations exhibiting consistent vegetation conditions are chosen, and two soil profiles are excavated at each location. In forested areas, two dominant trees are randomly selected, and two soil profiles are dug at a distance of 1 m from the tree trunks. Soil samples are collected from three depth intervals: 0–5 cm, 5–20 cm, and 20–40 cm. For each depth, three replicates are collected to ensure data reliability. Approximately 1.5 kg of loose soil is carefully collected using large aluminum boxes, ensuring minimal compression to preserve the soil structure. The samples are subsequently labeled, transported to the laboratory, and air-dried for further analysis and processing.

2.3.4. Determination of Soil Aggregation Indicators

The WSA of soil was determined using the wet sieving method [18]. After bringing the soil samples (0–5 cm, 5–20 cm, and 20–40 cm) back to the laboratory, they were first air-dried naturally. Subsequently, large soil clumps were gently broken apart along their natural fractures to minimize disruption of the soil structure, and plant roots and other intrusions were carefully removed. The contents of air-dried aggregates and WSA were measured using the dry sieving and wet sieving methods, respectively [19]. The specific procedures were as follows: the air-dried soil samples were passed through soil sieves to separate air-dried aggregates of different particle sizes, which were then weighed to calculate their mass percentages. Next, 50 g of air-dried aggregate samples were placed on soil sieves, evenly soaked with distilled water for 30 min, and then oscillated for 1 min using an aggregate analyzer to separate WSA of different sizes (<0.25 mm, 0.25–0.5 mm, 0.5–1 mm, 1–2 mm, 2–5 mm, and ≥5 mm). The separated aggregates were oven-dried to a constant weight in a sand bath, weighed individually, and the content of WSA for each particle size was calculated.

2.3.5. Calculation of the Stability Index of Soil Aggregates

The Mean Weight Diameter

Mean weight diameter (MWD) is an important index to characterize the stability of soil aggregates. In vegetation restoration and ecological restoration studies, the average weight diameter is used to measure the improvement of soil quality. The MWD is determined as follows:

$$MWD={\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}(W_i{X}_i)}$$

(1)

where $W_i$ represents the mass percentage, %, of aggregates of grade i particle size; n represents the number of particle size classifications; $X_i$ represents the average particle size of the aggregate of particle i size, mm.

The Fractal Dimension of Aggregates (D)

For soil aggregates, the fractal dimension can reflect the diversity and uniformity of soil particle size. As an important index of soil quality, it is used to evaluate the aggregate structure and stability of soil. The D is calculated using the following formula:

$${\displaystyle \frac{M_i}{M}}={({\displaystyle \frac{X_i}{X_{\max }}})}^{3-D}$$

(2)

where $M_i$ represents the mass of aggregates less than the i particle size, g; $M$ represents the total mass of soil aggregates, g; $X_i$ represents the particle size of the i particle size class aggregate, mm; $X_{\max }$ represents the maximum particle size of the soil aggregate, mm.

The Macroaggregate Content

Soil macroaggregates are soil aggregates with a diameter greater than 0.25 mm. The change in macroaggregate content can reflect the improvement of soil quality and is one of the important indices of soil fertility. The function is:

$$R_{0.25}=1-W_{0.25}$$

(3)

where $R_{0.25}$ represents the water-stable large aggregate content of diameter > 0.25 mm, %; $W_{0.25}$ represents the content of aggregates less than 0.25 mm in diameter, %.

The Percentage of Aggregate Disruption (PAD)

PAD is the percentage of mechanically stable aggregates greater than 0.25 mm in soil that are destroyed after wet screening. It reflects the stability of soil aggregates under wet conditions and can be used to assess the erosion resistance of soil. The PAD is calculated as follows:

$$PAD={\displaystyle \frac{W_{d0.25}-W_{w0.25}}{W_{d0.25}}}\times 100\%$$

(4)

where $W_{d0.25}$ represents the large aggregate content of air-dried soil, %; $W_{w0.25}$ represents the water stability of the large aggregate content, %.

2.3.6. Determination of Soil Organic Matter

All the invasive substances in the soil samples collected from the field were carefully removed to ensure thorough and homogeneous mixing of the samples. The samples were subsequently placed in a well-ventilated indoor area for air drying (approximately 7 days). Following grinding and sieving (2 mm aperture), the samples were weighed, and their organic carbon content was determined using the potassium dichromate external heating method [20]. The soil organic matter content was then calculated using the formula:

Soil organic matter (g·kg⁻¹) = soil organic carbon (g·kg⁻¹) × 1.724

2.3.7. Data Processing and Analysis

In this paper, Excel 2021 was employed to organize and categorize the 17 fundamental datasets related to vegetation root biomass, soil organic matter, and soil WSA obtained from quadrat surveys. Subsequently, SPSS (Statistics 26) software was utilized to identify outliers, assess normality, and evaluate variance homogeneity, with certain outliers being excluded. The results confirmed that the data met the assumptions of normality (p > 0.05) and variance homogeneity (p > 0.05) for all variables, justifying the use of parametric tests. Based on the root biomass and soil organic matter indices of various restored vegetation types, as well as the stability indices of soil aggregates across different soil layers within the same vegetation type, a One-Way ANOVA was conducted, followed by multiple comparisons using the HSD method. Differences were considered significant at α = 0.05. Additionally, to further elucidate the mechanistic pathways linking vegetation characteristics, the correlations between root biomass and soil organic matter with water-stable aggregate content, MWD, D, R_0.25, and PAD for different particle sizes (<0.25 mm, 0.25–0.5 mm, 0.5–1 mm, 1–2 mm, 2–5 mm, and ≥5 mm) were analyzed using the Spearman rank correlation method. Finally, Origin 2021 software was used to generate relevant charts.

3. Results

3.1. Distribution Characteristics of Soil Water Stability Aggregates of Different Vegetation Types

As shown in Figure 2a, the composition of WSA in the 0–5 cm soil layer varies significantly among different vegetation types. Specifically, forested areas (e.g., Platycladus orientalis (L.) Franco, Robinia pseudoacacia L., Pinus tabulaeformis Carr., Quercus variabilis Blume, and Pinus tabulaeformis-Quercus variabilis mixed forests) and grasslands exhibit higher proportions of WSA, particularly macroaggregates (>5 mm and 2–5 mm), which are significantly higher than those in farmland. This finding suggests that forest and grassland soils possess more stable structures, likely due to their well-developed root systems and higher organic matter content. Among forested areas, Platycladus orientalis (L.) Franco stands display the highest macroaggregate content, while mixed forests generally promote greater enrichment of macroaggregates compared to monoculture forests.

Compared to Figure 2a, the composition of WSA in Figure 2b shows some differences among vegetation types. Farmland exhibits a lower proportion of macroaggregates, whereas forested areas and grasslands maintain higher proportions. This indicates that forest and grassland soils demonstrate stronger structural stability in deeper soil layers. Notably, Pinus tabulaeformis-Quercus variabilis mixed forests have the highest proportion of macroaggregates, likely attributed to their complex root systems and greater organic matter input.

Figure 2c demonstrates that forested areas have significantly higher proportions of larger-sized aggregates (>5 mm and 5–2 mm) compared to farmland. This further highlights the superior stability of forest soils, which can be ascribed to their rich organic matter content and extensive root networks that facilitate the formation and stabilization of macroaggregates. As soil depth increases, grasslands show relatively higher proportions of smaller-sized aggregates (<0.25 mm), possibly due to their shallower root systems and lower organic matter input.

Overall, across the 0–40 cm soil layer, forested areas and grasslands exhibit higher proportions of larger-sized aggregates (>5 mm and 5–2 mm), while farmland shows a relatively higher proportion of smaller-sized aggregates (<0.25 mm) (Figure 2d). This reinforces the structural stability of forest and grassland soils. Mixed forests exhibit the highest proportion of macroaggregates, with soil structural stability markedly superior to that of monoculture forests.

3.2. Stability Analysis of Soil Aggregates of Different Vegetation Types

Figure 3a illustrates the variations in MWD across three soil layers (0–5 cm, 5–20 cm, and 20–40 cm) under different vegetation types. The results indicate that the Pinus tabulaeformis-Quercus variabilis mixed forests consistently exhibit the highest MWD values across all soil layers, reflecting the most stable soil structure. Additionally, Platycladus orientalis (L.) Franco and Quercus variabilis Blume forests demonstrate relatively high MWD values in deeper soil layers (20–40 cm), whereas farmland and grassland exhibit lower MWD values, particularly in the surface soil layer (0–5 cm) of farmland. These findings suggest that vegetation type significantly influences soil structure, with forest and grassland soils generally exhibiting greater stability than farmland (p < 0.05). Across different soil layers, except for farmland, forest, and grassland soils, there is a decreasing trend in MWD with increasing soil depth.

As shown in Figure 3b, the Pinus tabulaeformis-Quercus variabilis mixed forests demonstrate the highest R_0.25 values across all soil layers, further confirming its superior soil structure, followed by Platycladus orientalis (L.) Franco and Quercus variabilis Blume and grasslands also exhibit relatively high R_0.25 values in deeper soil layers (20–40 cm), indicating better aggregate stability in these vegetation types. In contrast, farmland shows lower R_0.25 values, especially in the surface soil layer (0–5 cm), reflecting poorer soil structural stability. Significant differences in R_0.25 values are also observed within the same vegetation type across different soil layers, with Pinus tabulaeformis Carr. and Robinia pseudoacacia L. forests showing significantly higher R_0.25 values in the 0–5 cm layer compared to the 5–20 cm and 20–40 cm layers (p < 0.05).

Figure 3c reveals that farmland has higher D values across all soil layers, suggesting a more fragile soil structure. Conversely, Platycladus orientalis (L.) Franco, Robinia pseudoacacia L., Pinus tabulaeformis Carr., Quercus variabilis Blume, and Pinus tabulaeformis-Quercus variabilis mixed forests exhibit lower D values in the 0–5 cm layer but higher D values in deeper soil layers (20–40 cm), indicating a more complex soil structure. The Pinus tabulaeformis-Quercus variabilis mixed forests show the lowest D values across all soil layers, likely due to their complex root system and higher organic matter content. Differences in D values are also observed within the same vegetation type across soil layers, with Robinia pseudoacacia L. showing significantly higher D values in the 20–40 cm layer compared to the 0–5 cm layer.

Figure 3d indicates that farmland has the highest percentage of aggregate destruction (PAD) values in the 0–5 cm layer, with PAD values decreasing as soil depth increases. Grasslands exhibit the highest PAD values in the 20–40 cm layer, indicating better aggregate stability in this layer. Grasslands, Platycladus orientalis (L.) Franco, Robinia pseudoacacia L., and Pinus tabulaeformis-Quercus variabilis mixed forests show significantly lower PAD values in the 0–5 cm layer compared to other layers (p < 0.05). The PAD values of Pinus tabulaeformis Carr. and Quercus variabilis Blume vary little across soil layers but are generally lower than those of farmland, suggesting that these vegetation types contribute to maintaining aggregate stability. Notably, the Pinus tabulaeformis-Quercus variabilis mixed forests and Platycladus orientalis (L.) Franco stands exhibit relatively low PAD values in the 20–40 cm layer, highlighting their significant role in stabilizing deep soil structures.

3.3. Analysis of Root Biomass, Soil Organic Matter Content, and Soil Particle Composition of Different Vegetation Types

As shown in

Table 2. Root biomass (RB) characteristics of different restored vegetation types g/m².

Soil Layer/cm	Vegetation Type	≤1 mm RB	1~≤2 mm RB	2~≤5 mm RB	5~≤10 mm RB
0~5 cm	Farmland	86.18	72.72	—	—
	Abandoned grassland	43.16	35.28	0.19	—
	Platycladus orientalis (L.) Franco	141.98	19.74	68.99	10.96
	Robinia pseudoacacia L.	33.96	22.65	36.06	91.65
	Pinus tabuliformis Carrière	83.88	22.64	21.04	102.49
	Quercus variabilis Blume	61.85	16.88	17.98	197.08
	Pinus tabuliformis-Quercus variabilis mixed forests	40.52	18.24	9.67	79.38
5~20 cm	Farmland	80.21	77.80	—	—
	Abandoned grassland	24.74	15.75	3.56	—
	Platycladus orientalis (L.) Franco	111.17	23.14	12.39	72.62
	Robinia pseudoacacia L.	37.79	23.42	65.00	322.72
	Pinus tabuliformis Carrière	40.71	23.41	32.57	123.43
	Quercus variabilis Blume	60.40	31.58	24.67	34.78
	Pinus tabuliformis-Quercus variabilis mixed forests	47.08	32.46	26.02	—
20~40 cm	Farmland	85.71	34.09	—	—
	Abandoned grassland	8.48	4.28	5.93	—
	Platycladus orientalis (L.) Franco	67.50	23.87	15.25	112.04
	Robinia pseudoacacia L.	34.01	19.11	41.68	473.41
	Pinus tabuliformis Carrière	28.28	24.18	19.32	33.79
	Quercus variabilis Blume	48.81	21.21	42.92	98.54
	Pinus tabuliformis-Quercus variabilis mixed forests	56.29	34.85	44.31	20.72

Notes: “—” means missing data.

and Figure 4, significant differences (p < 0.05) were observed in root biomass (RB) and soil organic matter (SOM) content among vegetation types and soil depths. Table 2 demonstrates that Robinia pseudoacacia L.exhibited the highest root biomass (1201.45 g/m²) in the 0–40 cm soil layer, followed by Platycladus orientalis (L.) Franco (679.66 g/m²), Quercus variabilis Blume (659.69 g/m²), Pinus tabuliformis Carrière (555.73 g/m²), Pinus tabuliformis-Quercus variabilis mixed forests (409.54 g/m²), and abandoned grassland (138.67 g/m²). Compared with agricultural land, restored vegetation types showed a 70.33–1375.81% increase in ≤10 mm RB. The ≤10 mm RB per unit volume of soil decreased with deeper soil layers.

Figure 4 shows that different vegetation types had significant effects on soil organic carbon content. Compared with agricultural land, restored vegetation types demonstrated a 123.70–279.33% increase in SOM. Notably, Platycladus orientalis (L.) Franco exhibited significantly higher SOM levels than other restored vegetation types (p < 0.05). In the 0–5 cm soil layer (Figure 4a), both Platycladus orientalis (L.) Franco and Robinia pseudoacacia L. displayed the highest SOM concentrations, which were significantly higher than those of farmland and grassland, and they made a significant contribution to the accumulation of SOM. In the 5–20 cm soil layer (Figure 4b), the organic carbon content decreased overall, but Platycladus orientalis (L.) Franco maintained a high level, which was significantly higher than that of other vegetation types. As the soil layer became deeper, the difference in SOM between vegetation types decreased (Figure 4c).

The mechanical composition of soil governs key edaphic properties, including water retention, nutrient availability, aeration, and thermal conductivity, as well as soil aggregate stability and susceptibility to erosion. As illustrated in Figure 5, the dominant particle-size fractions across the control farmland and six restored vegetation types were consistently ranked as follows: silt > clay > sand. Specifically, clay content ranged from 10.79% to 38.56%, silt content from 32.85% to 61.17%, and sand content from 4.11% to 55.76%, collectively classifying the soil texture as clay loam according to the USDA soil taxonomy. Significant differences in all three particle-size fractions were observed among land-use types, indicating that vegetation restoration exerts a pronounced influence on soil mechanical composition. Notably, clay content in the 0–5 cm layer was lower under all restored vegetation types than in the control farmland, and there were certain differences among different vegetation types with soil depth. One-way ANOVA revealed statistically significant divergence in mechanical composition between Quercus variabilis Blume and all other restored vegetation types, and the differences in the mechanical composition of each layer of soil in the abandoned grassland and the Platycladus orientalis (L.) Franco was significant (p < 0.05).

3.4. Relationship Between Soil Aggregate Stability and Organic Matter and Root Biomass in Different Vegetation Types

Figure 6 illustrates the correlation patterns among litter biomass, root biomass across diameter classes, soil physicochemical properties, and soil aggregate stability following vegetation restoration in the Eastern Qinling Mountains. Litter accumulation and root enhanced the formation and persistence of large soil aggregates. Specifically, large WSAs (SA1 and SA2) exhibited significant positive correlations with RB2 and RB3 (p < 0.05), suggesting that carbon inputs from coarse roots facilitate macroaggregate stabilization. The mean weight diameter (MWD) and R_0.25 were positively correlated with small-sized aggregates (SA6), RB3, and RB4, and negatively correlated with PAD and D, reflecting the key role of small-sized aggregates and vegetation roots in enhancing soil structure stability. Soil clay and silt content also contributed positively to MWD and R_0.25, thereby reinforcing aggregate stability (Figure 6). Notably, soil organic matter (SOM) showed strong positive correlations with large aggregates (SA1, SA2), MWD, and R_0.25, and a significant negative correlation with PAD, confirming SOM as a central binding agent for aggregate formation and resilience. By comparison, fine-root biomass (RB1) displayed non-significant correlations with most measured indicators, implying its comparatively limited influence on aggregate dynamics relative to coarse roots and litter.

4. Discussion

4.1. Distribution Characteristics and Stability Evaluation of Soil Aggregates of Different Restored Vegetation Types

This study revealed significant differences in the distribution characteristics of soil aggregates under various restored vegetation types (Figure 2). The proportion of large aggregates (>0.25 mm) in forest land was significantly higher than that in grassland and agricultural land, whereas the proportion of micro-aggregates (<0.25 mm) exhibited an opposite trend. This finding is closely associated with the differences in soil organic matter input and root activities caused by varying vegetation types [12,21]. Forest land, characterized by high vegetation diversity and a complex root structure, provides abundant sources of organic matter and promotes the formation of large aggregates through root secretions and mycorrhizal fungi [22]. In contrast, grassland exhibits relatively low vegetation diversity, shallower roots, and simpler structures, leading to reduced organic matter input and consequently a lower proportion of large aggregates and a higher proportion of micro-aggregates. These findings align well with the research results reported by [2] on the Loess Plateau [23] in the Ziwuling area of northern Shaanxi, and [24] in the Danjiang River Basin.

Furthermore, the stability of soil aggregates in forest land is significantly greater than that in grassland and farmland. This outcome is strongly influenced by factors such as soil organic matter content, microbial activity, and root secretions [11]. The higher levels of organic matter and microbial activity in forest land soils enhance aggregate stability [25,26], whereas the lower organic matter content and microbial activity in grassland and agricultural land soils result in reduced aggregate stability. Physical processes, such as dry-wet alternations, play a dominant role in the formation and stabilization of micro-aggregates. Root secretions in forest land, including polysaccharides and proteins, act as binding agents that promote soil particle aggregation and further enhance aggregate stability [22,26]. In Pinus tabuliformis-Quercus variabilis mixed forests, there is a high diversity of litter. The presence of diverse plant species with interwoven roots creates complex structures that provide rich habitats for soil organisms, thereby increasing the synergistic effects of microbes, animals, and plants in the soil. Consequently, this effectively promotes the increase in the proportion of large particle aggregates (≥1 mm), MWD, and R_0.25 values in the 0–40 cm soil layer, resulting in significantly better soil structure stability compared to single-species forests and agricultural land (p < 0.05) [27,28]. The surface soil of forest land and grassland (0–5 cm) accumulates substantial amounts of plant litter (e.g., dead branches, leaves, and grass root residues), which is rich in organic matter and has dense plant roots. Root penetration and secretions facilitate soil particle aggregation. Additionally, the rich and highly active microbial community in surface soil increases the number and stability of aggregates, improving soil erosion resistance [22,29,30,31,32]. As soil depth increases, soil structure stability deteriorates [12]. Therefore, selecting vegetation types that enhance soil aggregate stability during restoration is crucial for improving soil structure and controlling soil erosion.

Vegetation litter inputs and roots significantly influence the formation and stability of soil aggregates by altering the physical, chemical, and biological properties of the soil. Studies have demonstrated that changes in litter and soil nutrient characteristics during vegetation restoration and their driving mechanisms for plant diversity are critical for maintaining regional biodiversity conservation and ecological stability [33,34]. Moreover, vegetation restoration plays a pivotal role in the formation and stability of soil aggregates by modifying litter quality, microbial community structure, increasing soil organic matter content, and improving soil structure, thereby enhancing soil erosion resistance. This holds great significance for ecological protection and restoration in the Eastern Qinling region [35,36]. The stability of soil aggregates is significantly positively correlated with litter and plant roots (Figure 6). High-stability soil aggregates provide a stable growth environment for plant roots, promoting plant growth and reproduction.

4.2. Change Characteristics of Root Biomass and Soil Organic Matter Content of Different Restored Vegetation Types

Vegetation restoration markedly increased both root biomass and soil organic matter content. However, the magnitude and vertical distribution of these increases were strongly contingent upon vegetation type [37,38]. Among the restored stands, Pinus tabuliformis-Quercus variabilis mixed forests and deciduous broad-leaved forests (Robinia pseudoacacia L.) exhibited the greatest enhancements. Notably, Robinia pseudoacacia L. possessed the highest total RB (1201.45 g/m²) in the 0–40 cm soil profile, followed by Platycladus orientalis (L.) Franco (679.66 g/m²) (Table 2). This substantial belowground carbon input, particularly from fine roots (<2 mm diameter), directly promoted the formation of particulate organic matter and sustained microbial metabolic activity by supplying labile carbon substrates, thereby accelerating SOM accumulation [22,32]. Consequently, these forested areas showed significantly higher SOM concentrations, especially within the surface layer (0–5 cm) (Figure 4a).

The underlying mechanisms differ fundamentally between forest and grassland systems. The deep root systems of forests, especially Robinia pseudoacacia L., effectively transport carbon to deeper soil layers (>20 cm), mitigating the typical surface accumulation found in grasslands [39,40]. Furthermore, the lower root turnover rates in forests prolong carbon residence time, allowing for greater stabilization of SOM through organo-mineral associations [31,39]. In contrast, grasslands stored 71% of their total RB in the uppermost 10 cm of soil. This “shallow-root pump” mechanism results in rapid SOM accumulation in the surface horizon but limited enrichment below 20 cm (Figure 4b,c), making the SOM pool more vulnerable to disturbance and mineralization [41,42].

A key finding was the asymptotic trend of SOM increase after approximately 25 years of restoration (Figure 4d), suggesting a potential C-saturation threshold in these ecosystems [43]. However, even after 30 years, mixed forests maintained a 15% higher stable carbon fraction (associated with silt and clay particles) compared to monoculture plantations, underscoring the long-term advantage of diverse plant communities in creating a more resilient soil carbon pool [44,45]. This diversity fosters a wider range of root litter chemistries and rhizodeposits, supporting a more complex microbial community that enhances carbon sequestration efficiency [29,39].

The pronounced textural differentiation observed among vegetation restoration types, particularly the reduced clay content in surface soils (0–5 cm) under restored vegetation compared to farmland (Figure 5), can be attributed to several interacting mechanisms. First, vegetation restoration enhances biological mixing and bioturbation through root growth and soil fauna activity, which physically translocates fine particles downward along root channels and macropores. Second, the accumulation of organic matter in surface soils under forests (Figure 4) promotes the formation of stable organo-mineral complexes, particularly with silt and sand fractions, effectively ‘diluting’ the relative proportion of clay in the mechanical composition analysis following organic matter removal. Third, root water uptake and canopy interception alter soil hydrological regimes, reducing the intensity of eluviation processes that typically concentrate clays in surface horizons under agricultural tillage. The significant differences in mechanical composition between Quercus variabilis Blume and other restored types, as well as between grassland and Platycladus orientalis (L.) Franco (p < 0.05) suggests that vegetation-specific traits, such as root architecture, litter quality, and associated microbial communities, modulate the rate and magnitude of these textural reorganization processes.

In conclusion, vegetation restoration effectively enhances RB and SOM as well as affects the composition of soil texture, with mixed-species forests offering the most robust and sustainable strategy for long-term soil carbon sequestration and ecosystem functioning in the East Qinling Mountains.

4.3. Analysis of Influencing Factors on Soil Aggregate Stability of Different Restored Vegetation Types

The stability of soil aggregates under different restored vegetation types is regulated by a complex interplay of biotic and abiotic factors, with vegetation type acting as a primary driver by modifying root systems, organic matter input, and associated microbial activity (Figure 6) [30,46].

First, plant roots are fundamental engineers of soil structure. Our results suggested that Robinia pseudoacacia L. and Platycladus orientalis (L.) Franco possessed higher root biomass (Table 2), which may significantly contribute to aggregate stability. Fine roots (<2 mm) physically enmesh soil particles, while root exudates (e.g., polysaccharides and organic acids) act as binding agents that promote particle cohesion [22,32]. This mechanism was particularly evident in forested lands, where developed root systems led to higher proportions of macroaggregates (>0.25 mm) compared to grassland and farmland (Figure 2).

Second, SOM is a crucial cementing agent for aggregate formation and stabilization. The significantly higher SOM content in forest soils, especially in surface layers (Figure 4), provides abundant binding materials that may enhance aggregate resistance to disruptive forces [46,47]. The Pinus tabuliformis-Quercus variabilis mixed forests exhibited superior stability (Figure 3), which could be likely due to its diverse litter input, which increases the quantity and chemical diversity of organic matter, fostering the formation of persistent organic-mineral associations [31,48].

Third, soil microbial activity, especially fungal contributions, is thought to play an indispensable role. Fungi, especially mycorrhizal hyphae, are widely reported to enmesh microaggregates into macroaggregates and produce glomalin-related soil protein, a highly effective organic binding agent [27,30]. The higher organic matter content and diverse root exudates in forest soils likely support a more abundant and active microbial community, which facilitates the decomposition of organic residues and the production of adhesive substances, thereby enhancing aggregate stability [29,45]. In contrast, the lower microbial activity in agricultural soils, due to disturbance and limited organic input, may lead to weaker aggregation. Notably, microbial properties were not directly measured in the present study; thus, the above inferences are based on existing literature and indirect evidence.

Furthermore, the positive contribution of soil clay and silt content to MWD and R_0.25, as illustrated in Figure 6, can be attributed to the fundamental role of fine mineral particles in aggregate formation and stabilization. Clay and silt particles serve as essential binding agents within soil aggregates. Due to their high specific surface area and surface charge, clay minerals facilitate strong cohesion through physicochemical interactions, such as van der Waals forces and cation bridging. Additionally, clay particles can form organo-mineral complexes by adsorbing organic matter, which acts as a persistent cementing agent, further enhancing aggregate stability. Silt particles, while less chemically active than clay, contribute to the structural framework of aggregates and improve internal packing density. Together, these fine fractions promote the formation of macroaggregates (>0.25 mm) and increase resistance to slaking, thereby leading to higher MWD and R_0.25 values. These findings underscore the importance of soil texture as an inherent controlling factor in soil structural stability.

Despite the consistent trends observed in this study, several limitations should be acknowledged. Soil texture, microclimate variability, and previous land-use history may act as confounding factors that influence aggregate stability across sites [10,38]. In addition, the lack of direct measurements of microbial biomass, community composition, and enzyme activities limits definitive causal interpretations. Future studies combining long-term monitoring and direct microbial assays would help to clarify the underlying mechanisms more rigorously.

Overall, the present results suggest that interactions between vegetation-derived carbon inputs (roots and litter) and subsequent microbial processing may create a positive feedback loop that promotes the formation and stabilization of soil aggregates [45,48]. Mixed forests, with complementary root structures and diverse litter quality, appear more effective in generating such synergistic conditions, supporting more stable soil structures than monocultures or non-forested ecosystems.

5. Conclusions

This study systematically evaluated the effects of six typical restored vegetation types on soil aggregate stability across different soil depths in the East Qinling Mountains. The main findings and their implications are summarized as follows:

(1)

Vegetation restoration significantly enhances soil aggregate stability compared to farmland, with forest ecosystems exhibiting superior performance. Among the studied vegetation types, Pinus tabulaeformis-Quercus variabilis mixed forests demonstrated the highest proportion of macroaggregates (>0.25 mm) and the most stable soil structure across all soil depths (0–40 cm), highlighting the advantage of mixed-species systems over monoculture plantations in restoring soil structural integrity.

(2)

Depth-dependent patterns reveal that while surface soils (0–5 cm) show the most pronounced improvement in aggregate stability due to litter accumulation and dense fine-root networks, certain vegetation types—particularly mixed forests and Platycladus orientalis (L.) Franco—also exhibit significant capacity to enhance subsoil structure (20–40 cm). This underscores the importance of considering full-profile effects when evaluating restoration outcomes.

(3)

Vegetation restoration drives substantial changes in soil properties, increasing root biomass by 70–1375% and soil organic matter by 124–279% relative to farmland. Platycladus orientalis (L.) Franco and Robinia pseudoacacia L. are particularly effective in accumulating SOM, especially in surface layers. Notably, restoration also induces textural differentiation, reducing surface clay content and altering particle-size distribution with depth, which contributes to improved soil physical structure.

(4)

Mechanistic analysis suggests that aggregate stability is primarily governed by coarse root biomass (RB2, RB3) and SOM, which together explain 58% of the variance in MWD. Fine roots (RB1) play a comparatively limited role, while clay and silt particles provide foundational support by facilitating organo-mineral associations. These findings elucidate that organic inputs from aboveground litter and coarse belowground tissues, and their transformation into persistent SOM, are the key drivers of soil structural improvement under vegetation restoration.

Our findings provide a scientific basis for optimizing vegetation restoration strategies in the East Qinling Mountains and similar ecologically sensitive regions. We recommend that future restoration efforts prioritize the establishment of mixed forests, particularly Pinus tabulaeformis-Quercus variabilis mixed forests, to maximize both surface and subsoil structural benefits. For areas where monospecific plantations already exist, introducing complementary species to increase structural and functional diversity may enhance long-term soil stability. From a management perspective, protecting surface litter and promoting deep-rooting species should be integral to restoration planning. Future research should investigate the long-term dynamics of soil aggregate stability beyond 30 years, explore the microbial mechanisms underlying SOM transformation and aggregate formation, and assess how climate change scenarios may alter the restoration trajectories identified in this study. Such knowledge will be essential for developing adaptive management strategies that ensure the sustainability of restored ecosystems in a changing environment.