Study on the Mechanical Properties of Roots and Friction Characteristics of the Root–Soil Interface of Two Tree Species in the Coastal Region of Southeastern China

Abstract

The tensile strength of roots and the friction characteristics of the root–soil interface of tree species are the indicators that play a crucial role in understanding the mechanism of soil reinforcement by roots. To calculate the effectiveness of the reinforcement of soil by tree roots based on essential influencing parameters, typical trees in the coastal region of southeastern China selected for this study were subjected to tests of the tensile mechanical properties of their roots, as well as studies on the friction characteristics of the root–soil interface and the microscopic interfaces. The results indicated that in the 1–7 diameter classes, the root tensile strength of both Pinus massoniana and Cunninghamia lanceolata was negatively correlated with the root diameter in accordance with the power function. The root tensile strength of these two trees, however, was positively correlated with the lignin content but negatively correlated with cellulose and hemicellulose contents. The shear strength at the root–soil interface and the vertical load exhibited a constitutive relationship, which followed the Mohr–Coulomb criterion. As the root diameter increased, both the cohesion and the friction coefficients at the root–soil interface gradually increased, but the growth rate stood at around 15%. The cohesion value of the root–soil interface of the two trees decreased linearly with the increase in soil moisture content within the range of 25 to 45%. At the microinterface, the root surface of C. lanceolata exhibited concave grooves and convex ridges that extended along the axial direction of roots, with their height differences increasing with the enlargement of the root diameter. The rough surface of P. massoniana roots had areas composed of polygonal meshes, with an increase observed in the mesh density with increasing root diameter.

1. Introduction

Compared to traditional reinforcement methods like using anchor rods, soil reinforcement with vegetation is an environmentally friendly slope stabilization technique [1]. Plant roots play a significant role in both slope stabilization and erosion control by various mechanical and hydrological processes [2,3]. They can reinforce soil by reducing the evaporation of soil water and maintaining soil moisture content, as well as increasing soil shear strength [4].

The shear stress of the soil can be transferred to the root system through friction and the biochemical interaction between soil particles and roots during shear [5]. Both the soil reinforcement and slope stabilization are significantly influenced by root tensile strength [3]. Moreover, plant species, root diameter, length [6], root water content [7], structural chemical composition (cellulose, hemicellulose, and lignin contents) of root tissues [4,8], and root survival and decay time after plant death [9] are the factors that have an impact on root tensile strength. Studies have shown a decrease in tensile strength of plant roots with an increase in root diameter in a power law curve [7]. This phenomenon can happen due to the differences in the chemical components of root tissues. However, there are relatively few studies on the relationship between the lignin and cellulose contents and the tensile strength of roots [4]. Moreover, these studies have been carried out on only specific varieties of plants, and thus, the existing relationship between these chemical components and root tensile strength cannot be applied to all plants [10]. Indeed, studies on the tensile strength and chemical components of plant roots have been conducted at a regional level, while they have rarely been studied in the roots of some tree species widely distributed in the hills and mountains in southeast China.

It is noteworthy that landslides still occur in areas with high vegetation cover [2,11,12]. Related studies have reported that the occurrence of these landslides in vegetated areas is highly correlated with heavy rainfall. Furthermore, numerous studies have confirmed the significant impact of soil moisture content on the shear strength of the root–soil composite [13,14]. The effects of the reinforcement and anchoring of roots on soil have been found to depend on root–soil friction [15,16]. Therefore, the friction characteristics of the root–soil interface are the key to understanding the mechanism of the root–soil consolidation, and the quantification of the root–soil friction helps to quantitatively evaluate the mechanical reinforcement ability of roots [17,18]. Scholars have conducted numerous experimental studies on the friction characteristics of the root–soil interface. Through in situ and laboratory pullout tests, direct shear friction tests of the root–soil interface, and other experimental methods, the friction between roots and the soil has been proposed to occur as a result of the joint performance of adhesive friction, non-adhesive friction, and shear friction [15,19]. Some scholars have also studied the friction force at the root–soil interface under different moisture conditions using direct shear test equipment [17,20] and explored the microscopic characteristics of the root–soil interface by a scanning electron microscope (SEM) [21]. It is worth noting that the friction at the root–soil interface is different between various plant species, which may also exhibit differences in the roughness of their root epidermis at the same interface. In addition, at different soil moisture contents, the friction properties of the root–soil interface also significantly differ [22,23]. In vegetated coastal regions of southeast China, where shallow landslides are more frequent, there has been limited research on the friction and microstructural characteristics of the root–soil interface of widely distributed trees such as Pinus massoniana and Cunninghamia lanceolata with different root diameters and under various soil moisture conditions.

Landslides in vegetation-covered regions occur frequently in the coastal areas of southeastern China (Figure 1) and are mostly shallow landslides caused by rainfall in areas covered with vegetation [24]. The tensile strength of the roots and the friction at the root–soil interface are the key factors affecting the shear strength of the root–soil composite, but the mechanical properties of tree roots in these areas where landslides are widely distributed have been hardly investigated. Therefore, this paper took Pinus massoniana and Cunninghamia lanceolata, which are commonly found in this area, as the research objects. By the root tensile test, analysis of the structural chemical composition of roots, and an examination of the friction characteristics of the root–soil interface with varying root diameters and at different moisture contents, as well as scanning electron microscopy (SEM), we evaluated the mechanical properties of individual roots, the friction properties of the root–soil interface, and microscopic characteristics of two typical tree species in vegetated regions in southeastern China to obtain the tensile strength of roots and the root–soil interface friction strength of two typical trees in the study area, in order to provide calculation parameters for the subsequent calculation of the shear strength of tree–root–soil composites, and then to evaluate the stability of landslides.

2. Materials and Methods

2.1. Study Area

The study area is located in the southwest Fujian Province, China, between 24°47′ and 25°29′ N latitude and 115°51′ and 116°23′ E longitude. It has a subtropical maritime monsoon climate and an average annual rainfall of 1768.6 mm. The rainy season in this area occurs from March to June, during which the average rainfall is 927.51 mm, accounting for 52.44% of the total annual rainfall amount. The location of the landslide point selected for this study is Shifang Town, Wuping County, Longyan Municipality, Fujian Province, China (Figure 2: the data are sourced from the local government). The overall height and degree of the slope are 43−49 m and 35−68°, respectively. The main types of soil on the slope surface are residual sandy clay soil, fully worn granite, and sand-like strongly worn granite. The vegetation types developed on the slope are mainly Pinus massoniana and Cunninghamia lanceolata, along with a small mass of Phyllostachys edulis, and the vegetation coverage is 90%.

2.2. Root Tensile Tests

This study selected Pinus massoniana and Cunninghamia lanceolata trees with a diameter at breast height (DBH) of 95−110 mm, height of 4−6 m, and age of 9−12 years for root collection. Six trees of each species were sampled, which grew naturally in the mountains without artificial irrigation. Roots were separated from the soil (0−1.2 m deep), grouped into different diameter classes, separately placed in plastic bags, and stored at 4 °C [4,8]. The root systems of P. massoniana and C. lanceolata with diameters ranging from 0.5 to 7.5 mm were taken, and about 200 roots per plant were sampled. The analysis of the mechanical properties of roots was completed within 7 days of sampling (this study sampled roots in July 2023 and completed related tests within the same month). Measurements were taken at both ends and the midpoint of the root using an electronic vernier caliper and averaged to obtain the diameter of the root. The root of each tree species was divided into seven diameter classes: diameter grade 1 (0.5 < D ≤ 1.5 mm), 2 (1.5 < D ≤ 2.5 mm), 3 (2.5 < D ≤ 3.5 mm), 4 (3.5 < D ≤ 4.5 mm), 5 (4.5 < D ≤ 5.5 mm), 6 (5.5 < D ≤ 6.5 mm), and 7 (6.5 < D ≤ 7.5 mm). The DR-503A microcomputer-controlled electronic universal testing machine (Dongri Electrical Equipment Co., Ltd., Dongguan, China), with a maximum load capacity of 10 kN, a resolution of 0.004%, a strain rate control ranging from 0.01 to 500 mm/min, and a displacement measurement with a resolution of 0.01 mm, was employed to perform a tensile test. Currently, there is no standard gauge length of the root for root tensile tests. In this experiment, the gauge length and the loading rate were set to 100 mm and 10 mm/min, respectively, according to the methods of Zhang [4,8]. During the experiment, if the breakage of the root samples occurred at or near either end of the clamps, data were considered invalid [8]. However, data were deemed valid when the root samples were broken in the middle or near the middle of the clamps to ensure that the root fracture was caused by the tensile force rather than by any other damage types during the testing process. A total of 408 root samples were subjected to root tensile tests in this study, of which 140 were successfully tested. The universal testing machine was used to directly determine the tensile force of individual roots when they broke. Based on the root diameter, the tensile strength of the root could be calculated according to Equation (1) as follows:

$$T={\displaystyle \frac{4F}{\pi D^2}}$$

(1)

where T is the root tensile strength (MPa); F is the root tensile force (N); and D is the average root diameter (mm)

2.3. The Analysis of the Structural Chemical Composition of Root Tissues

Upon completion of the root tensile strength test, the roots of P. massoniana and C. lanceolata were classified into diameter classes 1–7, and then, the structural chemical composition of root tissues, including cellulose, hemicellulose, and lignin contents, was evaluated. First, the root samples were crushed into fine powder, and thereafter, the contents of their chemical constituents were measured using the method described by Su [8] and Lv [10] based on the Van Soest procedure.

2.4. Direct Shear Friction Test

The samples were prepared by stripping the root bark [17,20]. Before sample preparation, root vitality was assessed based on root peel color and abscission, and dead roots were removed. Since peeling the epidermis of the roots with a diameter of 0.5−2.5 mm is difficult, that of the roots with three diameter classes, including 4 (3.5 < D ≤ 4.5 mm), 6 (5.5 < D ≤ 6.5 mm), and 8 (>7.5 mm), was removed with a blade. The skin of the root with the same diameter class was attached to the customized round acrylic test block (a diameter of 61.8 mm) using a special wood adhesive so that the upper surface of the test block was positioned on the root surface (Figure 3). A strain-controlled direct shear apparatus was used to conduct the direct shear friction tests on the samples of roots of different diameters. The test procedure involved the embedding of the surface-adhered root bark specimens in the lower box of the direct shear apparatus, with the upper part of the shear box being placed on the root surface. The root axis was positioned parallel to the shear direction to mimic the stress of root–soil separation. The upper shear box was filled with soil with a specific moisture content, collected from the study area; the evaluated volumetric moisture content was within the range of 25% to 45%, and the physical properties of the soil are shown in

Table 1. Physical properties of the residual sandy clay soil.

Soil Type	Density/g·cm−3	Dry Density/g·cm−3	Saturated Hydraulic Conductivity/m·s−1
Residual sandy clay	1.63	1.38	1.75 × 10−6

. Since the root distribution of the two species of trees in the study area is mostly at the soil depth of less than 3 m, the determination of vertical load was performed on the soil layer in which the roots of these trees are mostly concentrated. Based on the formula for the calculation of soil self-weight stress, the self-weight stress of the root distribution layer at the soil depth of 3 m was about 55 kPa. According to the requirements for the direct shear test, the vertical load was set at 25, 50, 75, and 100 kPa. The shear test of the root–soil interface was performed following the GB/T 50123-2019 Standard [25] for Geotechnical Testing Methods, with a shear rate of 0.8 mm/min [18], and the strength parameters obtained from a direct shear test, including the cohesion value and internal friction angle, were calculated according to the Mohr–Coulomb failure criterion [26].

Due to the subtropical maritime monsoon climate of the study area, which is characterized by hot and humid weather, the volumetric soil moisture content remained above 25% throughout the year. Moreover, the instruments currently used for landslide monitoring by determining the soil moisture content mostly employ the principle of time domain reflectometry (TDR), and the moisture content measured using these devices is the volumetric moisture content [27,28]. The conversion of the volumetric moisture content to the mass of water per unit mass of soil (gravimetric water content) can be achieved according to Equation (2). The present study mainly explored the changes in the friction properties of the landslide at the root–soil interface in the study area experiencing heavy rainfall. Therefore, the soil volumetric moisture content selected and evaluated in this study was within the range of 25% (in a natural state) to 45% (at saturation).

$${\theta }_w=w{\rho }_d/{\rho }_w$$

(2)

where ${\theta }_w$ denotes the volumetric moisture content (%); $w$ indicates the mass of water per unit mass of soil (moisture content) (%); ${\rho }_d$ is the dry density of soil (g/cm³); and ${\rho }_w$ is the pore water density (g/cm³).

2.5. Scanning Electron Microscopic Evaluation of Roots with Different Diameters

The root–soil interface is composed of both the root surface and the soil surface in contact with the root [17,29]. Therefore, the friction characteristics of the root–soil interface and their variations are closely related to the root surface microstructure. Roots of different diameters were taken, cut into segments shorter than 10 mm, and then freeze-dried. After drying, they were subjected to micro-morphological analysis [21] using the NOVA NanoSEM 230 field-emission scanning electron microscope (FE-SEM) (manufacturer: Fei CZECH Republic S.R.O., Hillsboro, OR, USA) (Figure 4).

3. Results

3.1. Root Tensile Strength of the Two Tree Species

The root tensile test was successfully performed on 70 samples of P. massoniana, and the root diameter ranged from 0.72 to 7.47 mm, with the tensile strength being within the range of 7.16–33.48 MPa and an average of 15.47 MPa. Moreover, successful testing of 70 samples of C. lanceolata roots was performed by the tensile test, with the tensile strength ranging from 7.26 to 60.95 MPa and an average value of 19.72 MPa. C. lanceolata roots have greater tensile strength than P. massoniana roots. As shown in Figure 5a, the tensile strength of both the tree species tended to rise with the increase in the root diameter, demonstrating a positive power–function correlation. The relationship between the tensile force (F) of Masson pine roots and the root diameter (D) was represented as follows: F = 22.08D^1.4 (R² = 0.93, p < 0.01), and for C. lanceolata, the same relationship of F = 26.75D^1.48 (R² = 0.91, p < 0.01) was obtained. Figure 5b shows that the tensile strength of tree roots tended to decrease with increasing diameter, indicating a negative power–function correlation. The relationship stated below was observed between the tensile strength (T) and diameter (D) of P. massoniana roots: T = 32.46D^−0.66 (R² = 0.86, p < 0.01), while that between the tensile strength and diameter of C. lanceolata roots was as follows: T = 36.55D^−0.58 (R² = 0.81, p < 0.01). Compared to coarse roots, fine roots have a lower tensile force but a higher ultimate tensile strength, which can be explained by the differences in the root chemical components.

3.2. Variations in Relationships between Root Diameter Classes and Contents of the Structural Chemical Composition in Root Tissues

The changes in contents of the structural chemical composition in root tissues of both P. massoniana and C. lanceolata in the 1–7 diameter classes are presented in Figure 6. In the roots of both trees, the cellulose, lignin, and hemicellulose contents together accounted for 60%–70% of the total root biomass. The relationship between the chemical composition of root tissues and diameter varied between species. As shown in Figure 6a, the cellulose and hemicellulose contents in the roots of the former increased with the increase in root diameter, while the lignin content exhibited an opposite trend. The root diameter class (D_c) exhibited a significant positive power–function correlation with the cellulose content (C_e):(C_e = 20.14D_c^0.29, R² = 0.93, p < 0.01) but a significant negative power–function correlation with the lignin content (L_i): (L_i = 36.84D_c^−0.14, R² = 0.69, p < 0.05). According to Figure 6b, unlike the increase achieved in the cellulose and hemicellulose contents in the roots of the latter with the increase in root diameter, the lignin content decreased with increasing root diameter. A significant positive power–function correlation was recorded between the root diameter class (D_c) and the cellulose content (C_e) as follows: (C_e = 17.45D_c^0.20, R² = 0.85, p < 0.01), whereas (D_c) established a significant negative power–function relationship with the lignin content (L_i): (L_i = 37.80D_c^−0.19, R² = 0.80, p < 0.01). The increase in the hemicellulose content in the roots of both tree species with the increase in root diameter, however, was limited.

Hemicellulose and cellulose are together called holocellulose, which is a term that constitutes all the carbohydrates in root fibers. The mass fraction range of holocellulose in P. massoniana was from 28.40% to 46.71%, while in C. lanceolata, it ranged from 26.27% to 38.91%. As presented in Figure 7a, the holocellulose content of both trees showed a significant positive power–function correlation with root diameter, which for P. massoniana was represented as follows: H_o = 28.79D_c^0.25, R² = 0.91, p < 0.01, while for C. lanceolata, Ho = 25.73D_c ^0.18, R² = 0.81, p < 0.01 was obtained. Lignin and cellulose are both the most abundant components in the cell walls of tree roots. The space between cellulose and hemicellulose is filled with lignin, which provides resistance to compression and increases the mechanical strength of the cell wall. The ratio of lignin to cellulose reflects the influence of their relative contents on the mechanical properties of tree roots. The lignin/cellulose of P. massoniana ranged from 0.70 to 1.66, with an average value of 1.12 ± 0.35. In C. lanceolata, however, this ratio was within the range of 0.97–2.03 (an average of 1.38 ± 0.39). It can be seen from Figure 7b that the lignin/cellulose (LC) in the roots of both tree species was negatively correlated with the root diameter class (Dc), which in P. massoniana roots was demonstrated as follows: LC = 1.74D_c^−0.39, R² = 0.90, p < 0.01, with C. lanceolata exhibiting a significant correlation: LC = 2.08D^−0.36, R² = 0.93, p < 0.01.

3.3. Variations in Relationships between the Root Tensile Strength and Contents of the Structural Chemical Composition in Root Tissues

The relationships between the average root tensile strength and the chemical composition of both P. massoniana and C. lanceolata roots with different diameter classes are represented in Figure 8. The cellulose content in the roots of the former had a negative power–function correlation with tensile strength (C_e = 90.24T^−0.43, R² = 0.95, p < 0.01), whereas the lignin content exhibited a positive power–function correlation with tensile strength (L_i = 17.26T^0.22, R² = 0.74, p < 0.05). The latter showed a similar trend, with the relationships obtained for the cellulose and lignin contents being as follows: (C_e = 60.94T^−0.35, R² = 0.82, p < 0.01 and L_i = 11.54T^0.33, R² = 0.79, p < 0.01), respectively. Both types of tree showed a negative but less significant correlation between the hemicellulose content and root tensile strength. The ratio of lignin to cellulose (LC) is a comprehensive indicator of lignin and cellulose contents, which, as shown in Figure 9, was positively correlated with the root tensile strength in both tree species (in P. massoniana, this correlation was as follows: LC = 0.22T^0.61, R² = 0.93, p < 0.01, while for C. lanceola, the following relationship was observed: LC = 0.23T^0.61, R² = 0.93, p < 0.01).

3.4. Relationships between the Vertical Load and Shear Strength of the Root–Soil Interface

The theory of classical soil mechanics suggests that the relationship between the soil shear strength and vertical load should satisfy the Mohr–Coulomb criterion [30,31,32]. At different root diameters and soil moisture contents, the relationship between the shear strength of the root–soil interface and the vertical load also conformed to the Mohr–Coulomb criterion (Figure 10). The shear strength was significantly positively correlated with the vertical load (p < 0.01), with a coefficient of determination (R²) of greater than 0.90. Therefore, both the cohesion and friction coefficients of the root–soil interface were calculated based on the Mohr–Coulomb criterion.

3.5. Friction Characteristics of the Root–Soil Interface Obtained from the Direct Shear Test of the Tree Species Roots with Different Diameter Size Classes

Different plant species and root diameters may have different friction characteristics at the root–soil interface. The coefficients of friction and cohesion of the root–soil interface for P. massoniana and C. lanceolata roots under the same soil moisture conditions (a soil volumetric water content of 25%) and with three different root diameter classes (3.5 < D ≤ 4.5 mm, 5.5 < D ≤ 6.5 mm, and >7.5 mm) are shown in Figure 11. In terms of the natural moisture content, there was a significant difference in the cohesion and friction coefficients of the root–soil interface between the roots of the two tree species. The average cohesion and friction coefficients of the root–soil interface of C. lanceolata in the three diameter classes were found to be 17.78 kPa and 0.54, respectively. The average cohesion value of the root–soil interface of P. massoniana was 12.57 kPa, while the friction coefficient was 0.65. Moreover, both the cohesion and friction coefficients of the root–soil interface for both the tree species gradually increased with the increase in the diameter of the roots, which, however, differed to a limited extent. For both P. massoniana and C. lanceolata, the difference in cohesion and friction coefficients of the root–soil interface between roots with a diameter range of 3.5–4.5 mm and those with a diameter greater than 7.5 mm was around 15%.

3.6. Friction Characteristics of the Root–Soil Interface Obtained from the Direct Shear Test at Different Soil Moisture Contents

The friction characteristics of the root–soil interface of the two tree species and the soil–soil interface at five different soil volumetric water contents, including 25%, 30%, 35%, 40%, and 45% are provided in Figure 12. The cohesion and friction coefficients of the three types of interfaces indicated that the soil moisture content significantly affected the interface friction characteristics (p < 0.01). For both the soil–soil interface and the root–soil interface, the cohesion value decreased linearly with the increase in soil moisture content, with the value of the former decreasing from 24.02 kPa to 5.59 kPa. The relationship between the soil volumetric moisture content (θ_w) and cohesion value (c) was as follows: c = 50.75–1.02θ_w (R² = 0.96, p < 0.01). The cohesion value of the root–soil interface of P. massoniana decreased from 11.74 kPa to 8.33 kPa, and it was related to the soil volumetric moisture content: c = 15.73–0.17θ_w (R² = 0.89, p < 0.01). There was a decrease in the cohesion value of C. lanceolata from 17.06 kPa to 13.24 kPa, with the relationship between soil volumetric moisture content and root–soil interface cohesion being as follows: c = 22.49–0.21θ_w (R² = 0.96, p < 0.01).

For both the soil–soil interface and the root–soil interface of the two species of trees, the friction coefficient decreased with the increase in the soil moisture content, with decreases from 0.67 to 0.58 in the soil–soil interface. The soil volumetric moisture content (θ_w) was correlated with the friction coefficient (f), as follows: f = 1.54θ_w^−0.26 (R² = 0.91, p < 0.01). A decrease within the range of 0.63–0.57 was recorded for the friction coefficient of the root–soil interface of P. massoniana, whose relationship with the soil volumetric moisture content was found to be as follows: f = 1.19θ_w^−0.19(R² = 0.97, p < 0.01). For C. lanceolata, however, there was a reduction in the friction coefficient from 0.53 to 0.43, and the soil volumetric moisture content established a correlation with the friction coefficient of the root–soil interface as follows: f = 1.77θ_w^−0.37 (R² = 0.94, p < 0.01). Under lower vertical pressure (25 kPa), the root–soil interface of C. lanceolata exhibited greater friction. According to the quartile range of the box plot (Figure 13), the variation range of the soil–soil interface cohesion was the widest within the 25%–45% range of the soil volumetric moisture content. Furthermore, the variation range of the root–soil interface cohesion of P. massoniana was the narrowest. In contrast, the friction coefficient of the root–soil interface of C. lanceolata exhibited the greatest variation, while the root–soil interface friction coefficient of P. massoniana varied to the lowest extent.

4. Discussion

4.1. Effects of the Structural Chemical Composition of Root Tissues on Parameters of the Root Tensile Strength

The cell wall of the roots plays a key role in resisting tensile forces. Its main components have a close relationship with the mechanical properties of roots, with cellulose being the main component. Polysaccharide, consisting of D-glucose linked by β–1,4–glycosidic bonds, is the basic ring of cellulose molecules. Cellulose is essentially a linear polymer composed of glucose, insoluble in water and common organic solvents, and also one of the most abundant polysaccharides in plant cells [4,10]. In the cell wall, it exists in the form of molecular chain aggregate, forms both bundles and orderly arranged microfibrils, and acts as a skeletal material similar to steel bars in elements of reinforced concrete [8,33]. Lignin occupies a significant portion of the plant cell wall and its synthesis takes place in the final stages of cell differentiation during the lignification process. It is an organic polymer that comprises phenylpropane units linked by the carbon–carbon and ether bonds. In addition, lignin interacts both covalently and non-covalently with the polysaccharides in the cell wall, through which lignin–polysaccharide complexes that are crucial for maintaining the structural integrity of the cell wall are created. As a result of this phenomenon, an enhancement of the mechanical robustness of the cell wall occurs [4], akin to the cement in reinforced concrete structures due to its strengthening properties. Hemicellulose is a complex polymer made up of various monosaccharides, and in an amorphous state, it infiltrates the structural material. It has a low degree of polymerization and instability, as well as a low molecular weight. Moreover, hemicellulose serves as a matrix bonding agent, and thus, it is called the matrix material, analogous to the fine iron wire binding steel bars embedded in reinforced concrete structures [4,33]. Cellulose, hemicellulose, and lignin, spatially arranged in cell walls, contribute to the strength and resilience of plant roots.

There were notable variations in cellulose, hemicellulose, and lignin contents within the root system of both tree species, demonstrating varying correlations with the root diameter. This phenomenon can be attributed to the lower lignin content in the xylem compared to that in the phloem, whereas the cellulose and hemicellulose contents in the same root tissue exceeded those in the phloem. Throughout the growth cycle, the root cambium cells underwent an outward differentiation into the secondary phloem but differentiated inward into the secondary xylem, with the latter demonstrating higher rates of differentiation and growth. As the xylem proportion increased concurrently with the widening of the root diameter, there was an increase in cellulose and hemicellulose contents within the roots, while the lignin content decreased (Figure 6) [10,34].

The three chemical components, including cellulose, hemicellulose, and lignin, are significantly associated with the mechanical properties of roots. Although cellulose plays a crucial role in supporting the structure of the cell wall, when exceeding a certain level, it may not necessarily promote an increase in root tensile strength. In both Pinus massoniana and Cunninghamia lanceolata, the content of cellulose was observed to be negatively correlated with tensile strength (Figure 8), which has also been observed in the roots of certain tree species such as Pinus tabuliformis and Larix principis-rupprechtii, reported by several studies [10]. There was a positive correlation between the lignin content in the roots of both Pinus massoniana and Cunninghamia lanceolata and root tensile strength. Functioning as a robust material in the cell wall, lignin performs a similar action to “cement” in reinforced concrete structures. When the lignin content is elevated, an enhancement of the overall stability of the cell wall occurs [4]. Hence, a notable positive correlation existed between the lignin/cellulose and tensile strength in the roots of these two tree species.

4.2. The Influence of the Microstructural Characteristics of the Root Surface of the Two Tree Species on the Frictional Performance of the Root–Soil Interface

The results indicated that the friction properties of the interface between the soil and roots of P. massoniana and C. lanceolata varied with different root diameters (Figure 11). This variation occurred primarily due to the inconsistent patterns of the growth of roots with different diameters, leading to differences in the roughness of the root epidermis. Studies have shown an increase in the roughness of the root epidermis of the same plant species with increasing root diameter [35], and a higher roughness often causes a higher frictional resistance at the root–soil interface. The root–soil interface is a zone composed of the root epidermis and the soil surface, which come into contact. There is a close relationship between the friction characteristics of this interface and the microstructure of the root epidermis [21]. The microstructure of C. lanceolata roots is shown in Figure 14. The extension of concave grooves and convex ridges along the axial direction of the root could be observed on the surface of C. lanceolata roots at 400× magnification. The alternating arrangement of the concave and convex features makes the undulating and unequal epidermal cells align perpendicular to the axial direction of the root. With the increase in the root diameter, the difference in the height of the convex ridges and concave grooves gradually becomes larger, causing an enhancement of the root surface roughness. This is also the reason for the increase in friction of the root–soil interface of C. lanceolata with increasing root diameter [35]. At 2000× magnification, the epidermis of a Chinese fir root with a diameter of 5.5 mm can be observed to be composed of multiple irregular grids.

The microscopic structure of the P. massoniana root surface is shown in Figure 15. Observation of the rough root surface of P. massoniana composed of polygonal grids was carried out at 400x magnification. As the root diameter increased, increases in the grid density on the root surface occurred, which consequently enhanced the roughness of the root epidermis. When the roughness of the root epidermis is more significant, not only the frictional strength of the root–soil interface is improved but the root–soil contact area is also increased [21,35], which is also the reason why an increase in the root–soil interface friction occurs with the increase in root diameter. At 2000× magnification, remarkable undulations could be detected in the connections between the grids. The surface of the 3.5 mm diameter roots was seen to consist of elongated quadrilateral grids. The short and long axes of the grid units were about 15–30 µm and 40–100 µm, respectively, with the short axis directed perpendicular to the axial direction of the root system. However, a series of hexagonal, pentagonal, and quadrilateral grid elements made up the surface of roots with a diameter of 5.5 mm, with a short axis of about 20–50 µm and a long axis of about 30–80 µm. Furthermore, the arrangements of these grid elements were more irregular and oscillating, and when the rough surface of the root elongates along the axial direction, a greater friction is exerted than by the fine roots.

4.3. Mechanism of the Influence of Soil Moisture on the Friction Characteristics of the Root–Soil Interface and Landslide Occurrence

The tensile strength of tree roots and the friction force of the root–soil interface are two critical factors influencing the shear strength of the root–soil composite [4,36], which can be enhanced by a higher frictional resistance at the root–soil interface that also significantly restricts soil deformation. In this study, the cohesion and friction coefficients of the root–soil interface underwent a decrease to a certain extent with increasing soil water content, which is consistent with the findings of direct shear, triaxial, and pullout tests reported by other researchers [13,30,37]. The primary reason is the creation of interface suction by pore water existing between the roots and soil particles [23]. If the soil particles are simplified as spheres and the roots are simplified as an infinitely extended plane, the mechanism through which water content affects the root–soil interface can be explained as follows:

When the soil has a certain value of moisture content, a certain amount of capillary water is retained in the area of contact between soil particles and plant roots. The wetting action at this interface generates surface tension in the water–air interface, which is directed inward along the tangent of the curved surface, resulting in the creation of capillary pressure at the contact surface, which consequently increases the root–soil cohesion. However, too high or too low soil water content diminishes this cohesion [23].

There is a layer of weakly bound water film on the surfaces of soil particles. With a low soil moisture content, this water film becomes thin, creating low cohesion at the root–soil interface. With an increase in the soil moisture content, the weakly bound water film becomes thicker, which enhances viscosity and, consequently, improves the cohesion at the root–soil interface. However, when a certain value of the moisture content is exceeded, the loosely bound water film reaches a critical degree of thickness, causing the presence of the pore–water solution at the interface between the roots and soil particles, thereby reducing the friction between the soil particles and the matrix. Additionally, changes in soil moisture can disrupt the action of some cementing substances and cause an increase in the space between soil particles and roots, reducing the bonding strength of the root–soil interface, upon which the gradual weakening of cohesion at the root–soil inter-face ensues [20,38,39].

These characteristics have led some scholars to study the root–soil interface, through which they found that certain plants exhibited an initial increase followed by a decrease in cohesion and friction coefficients with increasing soil water content, with peak values achieved when the gravimetric water content ranged from 10% to 15% [17,20]. However, this phenomenon did not happen in this study because the range of variation in soil gravimetric moisture content at the root–soil interface in these studies was between 5% and 25%, while the range of the volumetric soil moisture content at the root–soil interface in this paper was 25%–45%, e.g, 17.86%–32.14% gravimetric water content. This study focused on the changes in the friction characteristics of the root–soil interface as the soil water content reached saturation from its natural state, which was different from the water content range evaluated in other studies.

Although the soil shear strength can be enhanced by the root system through mechanical enforcement [1,40], this contribution depends significantly on both the distribution patterns and the reinforcement efficiency of plant roots [41]. When shallow landslides occur, heavy rainfall accelerates water infiltration into the surface soil, which is facilitated by predominant root channels [2,42]. This is followed by the downward permeation of rainwater along the soil–soil and root–soil interfaces, which influences both the soil shear strength and the cohesion and friction coefficients of the root–soil interface, subsequently diminishing the efficacy of soil reinforcement by roots, resulting in compromised slope stability.

This study investigated the frictional characteristics of the root–soil interface under different root diameter classes and varying soil moisture contents, revealing significant influences of soil moisture content on the root–soil interface friction for two tree species. In subsequent research, certain physical and chemical properties influencing the frictional characteristics of the root–soil interface warrant further investigation. Moreover, applying parameters of root–soil interface frictional characteristics to calculate the shear strength of root–soil composites for assessing landslide stability requires further exploration to enhance practical applications.

5. Conclusions

This study selected two typical tree species in the coastal region of southeastern China and conducted tests on the tensile mechanical properties of their roots. Through analysis of the structural chemical composition of root tissues, the reasons for the differences observed in the mechanical behavior of roots with different diameters were explained. The friction characteristics of the root–soil interface of Pinus massoniana and Cunninghamia lanceolata with different root diameter classes and at different soil moisture contents were studied using direct shear tests and scanning electron microscopy (SEM) to calculate the effectiveness of the reinforcement of soil by tree roots in the study area based on essential influencing parameters. The conclusions obtained are as follows:

In the 1–7 diameter classes, the root tensile force of P. massoniana and C. lanceolata, ranging from 12.45 to 673.09 N, showed a positive power–function correlation with the root diameter. The root tensile strength was found to be within the range of 7.16–60.95 MPa, demonstrating a negative correlation with the root diameter. Both the average tensile strength and the tensile strength of C. lanceolata roots were higher than those of P. Massoniana roots.

In the 1–7 diameter classes, the cellulose mass fraction of P. massoniana and C. lanceolata ranged between 18.63% and 36.85%, while that of lignin was within the range of 23.53%–37.86%. Moreover, the ranges of hemicellulose and holocellulose contents were from 6.60% to 14.06% and from 26.27% to 46.71%, respectively, with the ratio of lignin to cellulose being 0.70–2.03. The root tensile strength of both tree species was positively correlated with the lignin content but negatively correlated with cellulose and hemicellulose contents.

As the root diameter was enlarged, both the cohesion and friction coefficients of the root–soil interface gradually increased, remaining around 15%. This phenomenon can be associated with the rougher microscopic interface. The soil volumetric moisture contents within the range of 25%–45% significantly influenced the interfacial friction characteristics, leading to a linear reduction in the cohesion of the root–soil interface with the increase in soil moisture content.

C. lanceolata has greater tensile strength in its roots compared to P. massoniana. Under lower vertical pressure (25 kPa), the root–soil interface of C. lanceolata exhibits greater friction. Therefore, for surface soils, the reinforcement effect of Cunninghamia lanceolata roots is better than that of P. massoniana.

At the microinterface, the surface of C. lanceolata roots exhibited concave grooves and convex ridges that extended along the axial direction of the root, with increases in the differences in their height with the enlargement of root diameter. The P. massoniana roots had rough surfaces with areas composed of polygonal meshes, where mesh density increased with increasing root diameter, which had a positive correlation with the root–soil interface friction.