Polyacrylamide-Induced Trade-Offs in Soil Stability and Ecological Function: A Multifunctional Assessment in Granite-Derived Sandy Material

Abstract

Soil erosion in granite-derived weathering mantles poses serious threats to slope stability and ecological sustainability in subtropical regions. While polyacrylamide (PAM) is widely used to improve soil structure, its concentration-dependent effects on multiple soil functions remain unclear. This study developed a multifunctional Soil Function Index (SFI) framework integrating erosion resistance (SFI₁), water regulation (SFI₂), and ecological function (SFI₃) to evaluate the effects of PAM application (0‰, 1‰, 3‰, 5‰, 7‰) on gully-prone sandy material. Herein, SFI₁ was quantified through shear strength (τ) and soil erodibility (K_r); SFI₂ was assessed using soil hydraulic parameters (saturated hydraulic conductivity and water retention curves) and SFI₃ was derived from the grass root system analysis. The results showed that SFI₁ and SFI₂ increased nonlinearly with PAM concentration, reaching maximum values of 0.983 and 0.980 at 7‰, with K_r reduced by 77.3% and non-capillary porosity (NAP) increased by 8.1%. In contrast, SFI₃ peaked at 0.858 under 3‰ and declined sharply to 0.000 at 7‰, due to micropore over-compaction, reduced aeration, and limited plant-available water. The total SFI exhibited a unimodal trend, with a maximum of 0.755 at 3‰, beyond which ecological suppression offset physical improvements. These findings demonstrate that PAM modifies soil multifunctionality through pore-scale restructuring, inducing function-specific thresholds and trade-offs. A PAM concentration of 3‰ is identified as optimal, achieving a balance between erosion control, hydrological performance, and ecological viability in the management of subtropical granite-derived sandy slopes.

1. Introduction

Soil erosion, a globally pervasive issue that chronically contributes to land degradation and agricultural productivity loss, was among the most critical limitations to ecosystem sustainability and socio-economic development worldwide [1]. It is estimated that global economic losses due to soil erosion exceed USD 400 billion annually [2], with subtropical regions being particularly susceptible. Subtropical areas, such as southern China, experience severe soil erosion characterized by extensive gully development, high erosion rates, and pronounced economic consequences [3]. According to the 2005 national soil erosion survey, from 1949 to 2005, the direct economic losses caused by these gullies reached CNY 550 million [4]. Therefore, these gullies are considered to be the main cause of soil erosion and ecosystem deterioration in southern China and are listed as a national concern [5].

Existing studies have shown that gully erosion in subtropical China is predominantly associated with granite-derived weathering mantles [6]. These mantles, formed under hot and humid climatic conditions, were mineralogically dominated by kaolinite and illite, which are clay minerals with low shrink–swell potential but high sensitivity to physical and chemical weathering [7]. The parent granite, characterized by pervasive joints and secondary fractures, weathers into a deep residual profile with a weakly cemented, highly heterogeneous horizon [8]. Among these, the granite-derived sandy material, typically corresponding to the lower sections of the weathering profile, represent the most structurally unstable and erosion-prone horizon [9]. These coarse-textured materials exhibit a disordered, flocculent microstructure, high anisotropy, and low cohesion, and are particularly vulnerable to rapid disintegration under rainfall or runoff impact [10]. Their high porosity and poor aggregate stability make them highly sensitive to moisture variation and preferential flow [11]. Crucially, the collapse and detachment of this sandy horizon directly control gully headcut retreat, slope failure, and sediment yield, particularly at the catchment scale [12].

The mechanical fragility of granite-derived sandy material is further exacerbated by the instability of interparticle binding agents such as free iron oxides, which readily dissociate under fluctuating redox and moisture conditions [13]. Under intense monsoonal rainfall, these soils frequently undergo infiltration-induced failure, shallow landslides, and severe rill-to-gully transitions [14]. Repeated wetting–drying cycles lead to aggregate breakdown, crack formation, and structural collapse, all of which accelerate soil detachment and sediment transport [15]. These intrinsic vulnerabilities underscore the critical need for structural enhancement strategies specifically targeting granite-derived sandy material. Improving their erosion resistance and resistance to moisture-induced degradation is essential not only for stabilizing slopes and reducing sediment loads, but also for enabling ecological restoration in highly erodible, subtropical gully landscapes.

Over the past decades, numerous physical and vegetative measures—such as contour tillage, engineering stabilization, and reforestation—have been applied to mitigate gully development [16]. However, their effectiveness remains limited in granite-derived soils due to the inherently weak structure and shallow rooting environment (Figure S1) [17]. Most conventional measures focus primarily on surface stabilization, while subsurface degradation processes—such as aggregate breakdown, porosity collapse, and loss of water-holding capacity—are often overlooked, even though these factors directly affect soil resilience and vegetation recovery [18]. In response, chemical soil conditioners, particularly polymer-based amendments, have emerged as promising alternatives to improve soil structural integrity and reduce erosion [19]. Among them, polyacrylamide (PAM), a long-chain synthetic polymer, was initially developed for use in industrial wastewater treatment and later adopted in agricultural systems to reduce irrigation-induced soil erosion and improve infiltration [20]. Its high molecular weight, strong flocculation capacity, and surface-binding affinity allow it to enhance interparticle cohesion and stabilize soil aggregates [21]. These characteristics have led to its expanded use in erosion-prone environments such as sloping farmland, post-fire landscapes, and degraded gullies, particularly under rainfall-dominated hydrological regimes [22].

In recent years, PAM has demonstrated particular promise in stabilizing coarse-textured, structurally weak soils, such as granite-derived sandy material, that are highly vulnerable to infiltration-driven collapse and detachment [23]. These soils, common in subtropical China, are characterized by low cohesion, poor aggregate structure, and high macroporosity, all of which contribute to severe gully erosion and slope instability during storm events [24]. By promoting particle bridging, reducing preferential flow, and enhancing micropore water retention, PAM can effectively mitigate both mechanical failure and runoff-induced sediment transport in such fragile substrates [25]. Its low cost, environmental safety, and operational simplicity further support its practical relevance in large-scale ecological slope restoration [26]. In addition to the well-documented stabilizing and water-retention effects, it is important to acknowledge that PAM may also pose potential risks. Residual monomers, incomplete degradation products, or by-products of PAM can potentially exert chemical toxicity in soils, especially under conditions of high application rates [27]. Moreover, PAM is subject to gradual degradation through photolysis, hydrolysis, and microbial activity, which can alter its molecular structure and functionality over time [28]. These processes not only reduce the durability of PAM-induced benefits but may also influence soil–plant–microbe interactions. Despite these predominant advantages, most existing studies have focused on single functional outcomes, such as erosion resistance or water retention, without evaluating how PAM simultaneously affects multiple soil functions, especially ecological performance [29]. Moreover, the potential trade-offs between physical improvements and biological viability have rarely been quantified. High PAM concentrations may lead to micropore over-compaction, reduced aeration, and inhibition of root growth, yet these effects are often overlooked in erosion control strategies. The lack of a unified framework for assessing soil multifunctionality under PAM application restricts its optimized use in fragile subtropical landscapes.

To address these gaps, this study proposes a multifunctional assessment of PAM-treated granite-derived sandy material in a typical subtropical gully system. We hypothesize that PAM enhances soil physical structure and water regulation in a dosage-dependent manner but may suppress ecological functionality beyond a critical threshold. Specifically, we aim to achieve the following:

(i) develop an integrated Soil Function Index (SFI) framework encompassing erosion resistance, water regulation, and ecological function;

(ii) quantify the concentration-dependent effects of PAM (0~7‰) on granite-derived sandy material through a combination of mechanical, hydrological, and microstructural indicators;

(iii) identify functional thresholds and trade-offs across physical and biological domains.

By linking pore-scale restructuring with multifunctional outcomes, this study provides a theoretical and practical basis for calibrating PAM application strategies in erosion-prone subtropical systems.

2. Materials and Methods

2.1. Study Area

The study area is located in Wuhua District (WH), Meizhou City, Guangdong Province (24°6′ N,115°37′ E~24°13′ N,115°34′ E) (Figure 1), and the gully density ranges in this region were classified using the natural breaks method to ensure an objective division of spatial heterogeneity. This region is located in the southern subtropical monsoon climate zone of China, experiencing an average annual temperature of 21.2 °C and an annual precipitation of approximately 1514 mm. The area is characterized by intense seasonal rainfall, steep slopes, and widespread granite-derived soils, which collectively contribute to severe gully erosion and slope instability.

Wuhua was selected due to its representativeness of subtropical granite weathering landscapes, exhibiting high erosion intensity and a dense distribution of collapsing gullies, with a total of 59 gullies identified within the investigated area. Long-term monitoring data from the Wupihe watershed in Wuhua have recorded soil erosion moduli ranging from 5 × 10⁴ to 1 × 10⁵ t km⁻² yr⁻¹, contributing over 85% of the sediment yield in local catchments [30]. The land use type is natural secondary forest mainly with Pinus massoniana, Eriachne pallescens, Phragmites australis, Rosa laevigata, Smilax china, and Dicranopteris linearis [31,32].

2.2. Soil Sampling and Experiment Design

Based on the IUSS Working Group WRB classification principles [32], a typical granite-derived weathering profile can be divided into three main horizons: the humus horizon (A), the illuvial horizon with clay accumulation (B), and the parent material horizon (C1 and C2) [33]. Previous investigations have shown that the sandy material horizon, due to its low cementation and weak structural stability, is the most erosion-prone horizon and plays a critical role in slope failure and gully formation [34]. Therefore, this study selected the C1 and C2 sandy material horizon as the primary source of experimental materials.

Soil samples were collected at the lower part of the erosion gully using the S-shaped sampling method. A soil drilling machine was employed to obtain samples at a depth of 5–7 m. In total, 30 intact soil cores (approximately 500 cm³ each) were collected, with each core having a diameter of ~6 cm and a height of 18 cm. Among these, 15 cores were air-dried and sieved to prepare bulk soil samples (approximately 22 kg) for physical and chemical analyses. Both the C1 and C2 horizons belong to the sandy material horizon of the C horizon and differ only in depth. During sampling, materials from the C1 and C2 horizons were combined at a 1:1 ratio to generate composite samples. Detailed physical and chemical characteristics of the collected soil are provided in

Table 1. Basic physicochemical properties of the experimental granite-derived sandy material.

Soil texture	Sand (%)	Clay (%)	Silt (%)	BD (g cm−3)	SOM (g kg−1)	TP (%)	LL (%)
Alumic Acrisols	74.31 ± 1.440	9.00 ± 0.270	16.69 ± 0.910	1.391 ± 0.122	3.631 ± 0.284	49.581 ± 2.877	38.730 ± 2.423
Soil texture	PL (%)	Ks (cm min−1)	Fed (g kg−1)	Ald (g kg−1)	Feo (g kg−1)	Alo (g kg−1)	w (%)
Alumic Acrisols	24.722 ± 1.747	0.307 ± 0.001	3.786 ± 0.424	1.498 ± 0.181	0.156 ± 0.031	0.491 ± 0.080	17.864 ± 1.072

Note: BD is soil bulk density; SOM is soil organic matter content; TP is total porosity; LL is liquid limit; PL is plastic limit; K_s is saturated hydraulic conductivity; Fe_d is amorphous iron oxide; Al_d is amorphous aluminum oxide; Fe_o is free iron oxide; Al_o is free aluminum oxide; and w is natural water content.

. After air-drying, the soil samples were passed through a 2.0 mm sieve to remove coarse debris, gravel, and plant residues, thereby ensuring sample homogeneity and consistency across test replicates. According to the USDA soil classification system, the tested material is classified as Alumic Acrisols [32,35].

PAM (Beijing Chemical Ltd., Beijing, China) used in this study is composed of white powdered particles with a diameter < 0.02 mm, a molecular weight of 1.2 × 10⁶ g/mol, a solid content of 93.3%, a hydrolysis degree of 25.2%, and a water-insoluble content of 0.2%. Based on pre-experimental trials, PAM was applied to soil samples at five different mass concentration ratios: 0‰ (CK), 1‰ (P1), 3‰ (P2), 5‰ (P3), and 7‰ (P4) [36,37]. To ensure uniform application, the granular PAM was slowly added into deionized water under continuous stirring until completely dissolved. The PAM solution was evenly sprayed onto the sieved, air-dried soil while continuously stirring to promote homogeneous distribution (Figure 2).

Following application, the PAM-treated soil was sealed with plastic film and allowed to equilibrate at ambient temperature for 48 h to facilitate full interaction between PAM molecules and soil particles [38]. The treated soils were then compacted in three layers using a standardized procedure to achieve a target bulk density of 1.35 g/cm³, with each layer filled into a ring cutter (diameter 6.18 cm, height 2 cm), thereby simulating in situ field compaction conditions and all experiments were conducted with three replicates to ensure the reliability and reproducibility of the results. The soil samples were divided into two categories: treatments without grass and treatments with grass. In the treatments without grass, only different concentrations of PAM were applied. In the treatments with grass, in addition to PAM application, 15 Vetiveria zizanioides seeds were evenly placed in each soil sample to investigate the effect of PAM on seed germination.

2.3. Hydraulic–Mechanical Characteristics

The soil water characteristic curve (SWCC) was determined using a 15-bar pressure plate apparatus by incrementally applying suction levels ranging from 0 to 1500 kPa until equilibrium was achieved at each step. The test was initiated with soil samples in a saturated state, and suction was applied in a sequence of 0, 0.1, 0.3, 0.5, 1, 3, 5, 10, and 15 bar, corresponding to pressures of 0, 10, 30, 50, 100, 300, 500, 1000, and 1500 kPa. The resulting volumetric water content data were fitted to the van Genuchten model [39] to characterize the soil moisture retention behavior:

$$\theta (\psi )={\theta }_r+{\displaystyle \frac{{\theta }_s-{\theta }_r}{{(1+{\left|\alpha \psi \right|}^n)}^{1-1/n}}}$$

(1)

$${\psi }_{air-entry}={\displaystyle \frac{1}{a}}$$

(2)

where θ_s and θ_r are the saturated and residual volumetric water content, respectively; $\psi$ (kPa) is the matric suction; a and n are curve parameters.

Soil shear strength, defined as the maximum shear stress that soil can withstand before failure, was measured using a direct shear apparatus [40]. Saturated soil samples were subjected to normal stresses of varying magnitudes, and the peak shear stress at failure was recorded. Mohr–Coulomb failure criterion was used to estimate the shear strength parameters [41]:

$$\tau =c+{\sigma }_n\mathrm{t}\mathrm{a}\mathrm{n}\phi$$

(3)

where τ (kPa) is the shear strength; c (kPa) is cohesion; σ_n (kPa) is the normal stress acting on the failure surface; and φ (°) is the angle of internal friction.

2.4. Soil Microstructure and Surface Area Analysis

To investigate microstructural changes induced by PAM treatment, scanning electron microscopy (SEM) was used to qualitatively and quantitatively analyze soil aggregate morphology, pore structure, and particle contact modes. SEM-based image analysis enabled the estimation of water- and air-filled pore spaces, offering visual evidence of pore continuity and connectivity. These microstructural attributes were further linked to macroscopic hydraulic properties, such as unsaturated hydraulic conductivity, facilitating mechanistic interpretation of PAM-induced structural modifications [42,43].

Nitrogen adsorption isotherms were measured using a Micromeritics ASAP2020M (Micromeritics Instrument Corporation, Norcross, GA, USA) volumetric adsorption apparatus. Specific surface area (SBET) was calculated based on the isotherm data using the Brunauer–Emmett–Teller (BET) equation [44]:

$${\displaystyle \frac{q}{q_m}}={\displaystyle \frac{cp}{(p_0-p)\left[1+(c-1)\left(p/p_0\right)\right]}}$$

(4)

where q is the amount adsorbed (mmol·g⁻¹); q_m is the monolayer adsorption capacity (mmol·g⁻¹); p is the equilibrium gas pressure (MPa); p₀ is the saturated vapor pressure (MPa); and c is a constant related to the heat of adsorption.

The BET model describes the relationship between gas pressure and adsorbed mass, enabling determination of surface area and adsorbed film thickness. The latter was estimated using the following scaling relation [45]:

$$m=S\times h\times {\rho }^{ave}$$

(5)

where m is the adsorbed mass (g·g⁻¹); S is the surface area (m² g⁻¹); h is the adsorbed film (nm); and ρ^ave is the average density of the adsorbed film (g/cm³).

Thermogravimetric analysis (TGA) was conducted using a Netzsch STA-2500 instrument to quantify bound water content. Samples were heated at a constant rate of 5 °C/min, and weight loss was recorded to differentiate water types based on thermal transition behavior. Compared to traditional techniques, TGA offers high sensitivity and allows the identification of water release patterns via inflection points, peak shifts, and other kinetic features [46,47].

Based on the integration of BET surface area data and TGA-derived bound water content, the thickness of the bound water film (t) was calculated using an empirical polynomial relation:

$$y=0.001x^2-0.027x+1.801 (R^2=0.99)$$

(6)

$$t={\displaystyle \frac{w_a}{A_{S}y}}$$

(7)

where t is the thickness of bound water (nm); w_a is the moisture content of maximum adsorbed water (%); A_s is the specific surface area of soil particles (m²/g); x is the water content (%); and y is the density of bound water (g/cm³). The microscopic pore structure indicators are summarized in

Table 2. Micropore structure parameters derived from BET and TGA analysis.

PAM (‰)	MSAi (m2 g−1)	MSAe (m2 g−1)	APD (nm)	SSA (m2 g−1)	MSSai (m2 g−1)
0	2.114 ± 0.001	7.857 ± 0.001	5.919 ± 0.002	9.980 ± 0.008	2.130 ± 0.015
1	0.514 ± 0.001	6.329 ± 0.001	6.824 ± 0.002	6.864 ± 0.030	0.517 ± 0.003
3	0.678 ± 0.001	6.146 ± 0.001	6.738 ± 0.001	6.826 ± 0.004	0.446 ± 0.003
5	0.387 ± 0.001	5.455 ± 0.002	7.123 ± 0.002	5.844 ± 0.004	0.390 ± 0.005
7	0.213 ± 0.002	5.341 ± 0.001	7.427 ± 0.001	5.569 ± 0.026	0.213 ± 0.001
PAM (‰)	MSSae (m2 g−1)	AAPw (nm)	TPV (cm3 g−1)	AWD (g cm−3)	WFT (nm)
0	7.873 ± 0.018	5.949 ± 0.030	0.013 ± 0.001	0.744 ± 0.042	0.710 ± 0.009
1	6.330 ± 0.003	6.842 ± 0.019	0.012 ± 0.001	0.893 ± 0.002	1.834 ± 0.004
3	6.170 ± 0.022	6.772 ± 0.031	0.011 ± 0.001	1.030 ± 0.009	2.154 ± 0.004
5	5.472 ± 0.019	7.150 ± 0.034	0.010 ± 0.001	1.119 ± 0.015	3.234 ± 0.004
7	5.343 ± 0.002	7.462 ± 0.030	0.011 ± 0.001	3.560 ± 0.015	6.515 ± 0.004

Note: MSA_i is the micropore surface area; MSA_e is the mesopore surface area; APD is the average pore diameter; SSA is the specific surface area; MSS_ai is the micropore-specific surface area; MSS_ae is the mesopore-specific surface area; AAP_w is the average adsorption pore width; TPV is the total pore volume; AWD is the adsorbed water density; and WFT is the water film thickness.

.

2.5. Simulated Scouring Experiment

To assess the erosion resistance of PAM-treated soils under surface runoff, a custom-designed flume system was employed. The setup consisted of a water distributor with adjustable slope, a rectangular soil flume (380 cm × 20 cm × 20 cm), a diversion channel, and a terminal runoff-infiltration collection system comprising a V-shaped trough and a storage reservoir (Figure S2) [48]. The flume slope was adjustable between 0° and 70° using a mechanical lifting device and stabilized using support rods. Local tap water was used as the flow source.

Before the experiment, soil samples were air-dried, gently crushed to break up large aggregates, and passed through a 2 mm sieve to facilitate subsequent treatments. The flume was filled with the treated soil, and samples were pre-saturated using ring knife cylinders, then equilibrated at room temperature for 24 h. The experiment was conducted at a fixed slope of 20°, under five different flow rates: 0.2 L/s (0.27 m/s), 0.3 L/s (0.36 m/s), 0.5 L/s (0.38 m/s), 0.7 L/s (0.45 m/s), and 0.9 L/s (0.54 m/s). Water depth and surface velocity were determined using the weighing method and the potassium permanganate tracer technique, respectively. Each condition was replicated three times for consistency. Runoff and infiltrated water were collected separately using the V-shaped trough and storage slots, ensuring complete capture of eroded material.

Sediment samples were collected at 10 s intervals during the first 30 s and at 30 s intervals from 0.5 to 4 min. The experiment was terminated either when the erosion depth reached 2 cm or when the structural integrity of the soil was visibly compromised. Sediment mass was determined via oven-drying, and detachment rate was calculated using the following formulas:

$$\tau =\rho gHS$$

(8)

$$D_c={\displaystyle \frac{M_c}{AT}}$$

(9)

where τ is the shear force of water flow (Pa); ρ is the water density (kg m⁻³); H is the water depth (mm); S is the sine value of slope; D_c is the soil separation capacity [kg (m²s)⁻¹]; M_c refers to the drying weight of the received sediment in unit time (kg); A is the sampler area (m²); and T is the scouring duration (s).

To further estimate soil erodibility under hydraulic stress, the Water Erosion Prediction Project (WEEP) model framework was employed, relating detachment to flow shear stress (Figure S2) [49]:

$$D_{\mathrm{c}}=K_r(\tau -{\tau }_c)$$

(10)

where K_r is the soil erodibility (s m⁻¹) and τ_c is the critical shear stress (Pa).

2.6. Integrated Soil Function Index (SFI) Construction

To quantitatively assess the integrated effects of PAM application on soil function, a composite Soil Function Index (SFI) was constructed, comprising three sub-indices representing distinct functional domains: (1) SFI₁ (Erosion Resistance): based on indicators related to mechanical strength and microstructural stability; (2) SFI₂ (Water Regulation): reflecting soil moisture availability and retention capacity; (3) SFI₃ (Ecological Function): capturing root development potential and biological compatibility.

(1) Indicator selection and dimension reduction

Candidate indicators for each sub-index were selected based on relevance to their functional domain and sensitivity to PAM treatment. Principal component analysis (PCA) was employed to reduce dimensionality and eliminate multicollinearity. Only components with eigenvalues > 1 were retained, and indicators with high loadings and low intercorrelations (|R| < 0.9) were used to represent each principal component. The optimal indicators selected for each sub-index are summarized in Table S1.

(2) Scoring and normalization

To ensure comparability across different units and scales, each selected indicator was normalized using a nonlinear scoring function as recommended by Andrews et al. [50]. Indicators were transformed so that higher values uniformly reflected better soil function. The scoring functions used for each indicator are detailed in

Table 3. The scoring function methods of soil quality indicators.

Score	Scoring Curve	Standard Scoring Functions
SFI	More is better	${\displaystyle \frac{X}{X_{max}}}$
	Less is better	${\displaystyle \frac{X_{min}}{X}}$

Note: X is the measured value of soil indicator; X_max is the observed maximum value; X_min is the observed minimum value.

and Table S2.

(3) Weighting and sub-index calculation

The relative contribution (weight) of each indicator to its corresponding sub-index was determined based on its communality in PCA. Each sub-index (SFI₁, SFI₂, SFI₃) was calculated as a weighted sum of normalized indicator scores:

$${\mathrm{S}\mathrm{F}\mathrm{I}}_k={\sum }_{i=1}^{nk}{w}_i{\times S}_i$$

(11)

where w_i is the weight of the i-th indicator in the sub-index k, and S_i is its normalized score.

(4) Integration and Statistical Analysis

The overall Soil Function Index (SFI) was computed as the arithmetic sum of the three equally weighted sub-indices:

SFI_total = 1/3SFI₁ + 1/3SFI₂ + 1/3SFI₃

(12)

Assuming equal importance of physical, hydraulic, and ecological functions in slope restoration, no further weighting was applied at the integrative level. One-way ANOVA was used to test for significant differences in SFI components among PAM treatments, followed by LSD post hoc comparisons at a 95% confidence level (p < 0.05). All statistical analyses were performed using SPSS v26.0.

3. Results

3.1. Effects of PAM on Soil Erosion Resistance Function Indicators

To assess the influence of PAM on erosion resistance, a set of mechanical and microstructural indicators were evaluated, including shear strength parameters, pore metrics, and detachment properties. Principal component analysis (PCA) was performed on the erosion-related variables (Table S1 and Figure 3a), and two representative indicators were retained based on high factor loadings and low collinearity (|R| < 0.9): the water-induced detachment coefficient (K_r) and the critical shear stress (τ). These indicators were used to construct the erosion resistance sub-index (SFI₁) via nonlinear scoring and PCA-derived weighting.

The final SFI₁ was calculated as a weighted sum of the normalized scores of these two indicators, with weights proportional to their communality in PCA (Table S2). Among them, the contributions of K_r and τ are basically the same. This weighting reflects the dominant role of hydraulic detachment resistance and shear strength in controlling soil erodibility.

As shown in Figure 4a, SFI₁ exhibited a nonlinear, dose-dependent increase with PAM concentration. The index rose from 0.420 at 0‰ to 0.983 at 7‰, with incremental gains at 1‰ (0.174), 3‰ (0.279), 5‰ (0.335), and 7‰ (0.563). The relationship followed a quadratic trend (R² = 0.958, p < 0.001), with the threshold near 7‰, which indicates that PAM concentration is positively correlated with SFI₁, and the polymer–soil bonding ability is the strongest at this time, implying that PAM application markedly enhances soil resistance to detachment and erosion.

To further explore the underlying mechanisms of this improvement, Figure 3b,c depict the changes in two key parameters: the soil erodibility factor (K_r) and the critical shear stress (τ). K_r showed a rapid decrease with increasing PAM concentration. At 0‰, K_r was 0.032 s·m⁻¹, reflecting high susceptibility to erosion. Upon PAM addition, Kr declined sharply, reaching at 3‰ (0.026 s·m⁻¹), and continued to decrease gradually at 7‰ (0.007 s·m⁻¹) (Table S3). This trend suggests that PAM substantially reduces soil erodibility. Conversely, the critical shear stress (τ), exhibited an overall increasing trend with PAM concentration. Starting from the control (7.317 kPa), τ rose steadily to 7‰ (11.830 kPa). The increase in τ is indicative of enhanced soil cohesion and mechanical resistance under PAM treatment and increasing inter-particle friction, thereby elevating the energy threshold for erosion initiation.

The synergistic effects of decreasing K_r and increasing τ collectively contributed to the consistent elevation of SFI₁. Reduced K_r implies improved structural integrity and lower erodibility, while increased τ reflects stronger mechanical resistance to hydraulic forces. Together, they provide mechanistic support for the observed improvements in SFI₁ with PAM application.

Overall, these results suggest that PAM strengthens erosion resistance primarily through reducing pore size and connectivity, thereby increasing critical shear stress and limiting detachment potential (Figure 4b). The dose-dependent response highlights PAM’s capacity to reinforce soil microstructure, especially under rainfall-induced shear conditions.

3.2. Effects of PAM on Soil Water Regulation Function Indicators

The soil water regulation sub-index (SFI₂) was developed to evaluate the influence of PAM on the soil’s ability to retain and supply plant-available water. Based on the PCA of hydraulic-related variables (Table S1 and Figure 5a), three representative indicators were selected: saturated hydraulic conductivity (K_s), volumetric water content (θ_v), and non-active pores (NAPs). These indicators showed high PCA loading values, low collinearity (|R| < 0.9), and strong responsiveness to PAM treatments. Following nonlinear scoring and PCA-derived weighting, θ_v, Ks, and NAPs were assigned weights of 0.354, 0.293, and 0.353, respectively, reflecting the dominant role of capillary-bound water in determining short-term water regulation capacity (Table S2). The final SFI₂ was calculated as a weighted sum of the normalized scores of these three indicators.

As shown in Figure 6a, SFI₂ exhibited a strong positive response to increasing PAM concentrations, with values steadily rising from 0.516 (0‰) to 0.980 (7‰). The relationship was well-described by a second-order polynomial regression (R² = 0.997, p < 0.001), suggesting a consistent and robust enhancement of soil water retention functionality with PAM addition. This improvement likely results from PAM-induced changes in soil physical structure, including reduced macro-pore prevalence and increased matrix stability.

Figure 5b–d reveal the variation trends of three critical influencing factors: saturated hydraulic conductivity (K_s), volumetric water content (θ_v), and the proportion of non-active pores (NAPs). Ks showed a steep decline with PAM applications. From a control value of 0.307 cm·min⁻¹, it dropped below 0.013 cm·min⁻¹ at 5‰ PAM and continued to decrease slightly at 7‰. This reduction indicates lower infiltration rates, likely due to PAM-induced surface sealing, pore blockage, and enhanced capillarity. θ_v increased continuously with PAM concentration, rising from 16.7% (0‰) to over 30% at 7‰ (Table S4). This trend reflects improved soil water-holding capacity under PAM-induced aggregation and micropore development. NAPs decreased with increasing PAM, suggesting a pore-size distribution shift toward water-retentive pores. These combined changes directly support the increase in SFI₂, illustrating how PAM enhances soil water retention through both hydraulic and structural modifications. Mantel tests (Figure 6b) indicated that Ks and θ_v were significantly correlated with soil microstructural parameters, including specific surface area (SSA_i), average pore diameter (APD), and total pore volume (TPV). These results confirm that the improvement in soil water regulation is the product of integrated changes in soil hydraulic properties and pore structure induced by PAM.

Mechanistically, the increase in SFI₂ was supported by water retention curve data and BET-derived pore structure analysis. These improvements reflect the enhanced moisture retention capacity under unsaturated conditions, as confirmed by the shape of the water retention curves fitted to the van Genuchten model (Figure 7a and Table S5). Specifically, key fitting parameters across treatments are listed in Table S5. The α parameter declined from 17.069 at 0‰ to 3.222 at 7‰, indicating increased resistance to desaturation and a higher water-retention threshold. The n parameter also slightly decreased, implying narrower pore size variability. The residual water content (θ_r) increased steadily, rising from 5.092% to 8.131%, suggesting improved moisture retention even under high suction. These findings indicate that PAM reduces macroporosity while enhancing microporosity, increasing the soil’s ability to retain water under drought or suction stress. These changes collectively enhanced the soil’s capacity to retain plant-available water and delay desaturation, thereby improving its short-term water regulation function.

3.3. Effects of PAM on Soil Ecological Function Indicators

The ecological function sub-index (SFI₃) was constructed to evaluate the effect of PAM on root development and plant–soil compatibility. Based on PCA of ecological performance metrics (Table S1 and Figure 8a), five indicators were selected: germination rate (G_r), root length (R_l), root count (R_cn), root average diameter (ARD), and root area (R_an). These were grouped into two principal components explaining 93.5% of total variance, with indicator weights derived from PCA communality values (Table S2). Among the selected indicators, R_cn and R_l showed the strongest loading values and highest sensitivity to PAM concentration, contributing 20.6% and 20.3% to SFI₃, respectively. The final SFI₃ was calculated as a weighted sum of normalized scores, following the same procedure described in Section 2.6.

As shown in Figure 9a, SFI₃ exhibited a unimodal response to PAM concentration. The ecological index increased from 0.446 (0‰) to 0.768 (1‰) and peaked at 0.858 (3‰), indicating enhanced conditions for seed emergence and root elongation at low to moderate PAM levels. However, a sharp decline was observed at higher concentrations: SFI₃ dropped to 0.235 at 5‰ and further declined to near zero (0.00) at 7‰. One-way ANOVA (IBM SPSS Statistics 27) confirmed significant differences between treatments (p < 0.05), with the 5‰ and 7‰ groups performing significantly worse than the control. As shown in Figure 8b–f, core plant-related indicators followed a trend congruent with SFI₃. Specifically, Gr significantly increased under 1–3‰ PAM, reaching maximum levels at 3‰, then sharply decreased with higher concentrations. This implies that moderate PAM improves seed–soil contact and water availability, facilitating seedling establishment. Rl, Rcn, Ran, and ARD were highest at 3‰, indicating vigorous root development and resource uptake capacity (Table S6). The pattern across multiple root morphological parameters supports the conclusion that PAM at 3‰ optimally stimulates root system complexity and ecological function.

The sharp decline in SFI₃ at higher PAM concentrations is likely associated with a combination of microstructural and physiological limitations. Figure 9b presents the relationships between SFI₃ and three key micropore parameters: MSA_e, SSA, and MSS_ae. All indicators showed significant positive correlations with SFI₃ in a parabolic pattern. Regression models demonstrated strong fits (R² = 0.934, R² = 0.958, and R² = 0.936). Low-concentration PAM enhanced pore surface properties, potentially improving microbial habitat space, aeration pathways, and rhizosphere development. A subsequent decline at higher PAM levels suggests pore blockage or aggregation collapse, which likely impairs microbial function and root penetration. These relationships emphasize the central role of soil pore microstructure in mediating the beneficial ecological outcomes of PAM amendment. BET analyses (Table 2) revealed that excessive PAM application induced a dense accumulation of micropores and reduced overall pore connectivity. This structural compaction may have restricted oxygen diffusion and increased the proportion of tightly bound water, thereby limiting the availability of plant-accessible moisture. While a moderate increase in microporosity (<0.2 μm) can enhance water retention and support root hair proliferation, overdevelopment of micropores reduces gas exchange and may suppress aerobic respiration in root tissues. Such conditions can impair energy production and nutrient uptake, ultimately resulting in shorter root length, lower emergence rates, and reduced root density, as observed in the 5‰ and 7‰ treatments. Additionally, surface polymer films formed at high PAM levels may increase soil hardness and introduce mechanical resistance to root penetration, further hindering root expansion and function.

These findings highlight a clear ecological threshold in PAM application: while low to moderate concentrations (1~3‰) improve root growth by enhancing water availability and aggregate stability, excessive application (>5‰) impairs ecological function by over-restricting gas and water exchange. The inverted-U pattern of SFI₃ suggests a strong functional trade-off and emphasizes the need for dosage optimization when applying PAM in ecological restoration contexts.

3.4. Effects of PAM on the Integrated Soil Multifunctionality Index

To evaluate the net effect of PAM across multiple soil functions, the total Soil Function Index (SFI) was calculated as the unweighted sum of the three sub-indices: erosion resistance (SFI₁), water regulation (SFI₂), and ecological function (SFI₃). This equal-weight approach reflects the importance of balancing structural stability, hydrological performance, and ecological viability in slope restoration contexts.

As shown in Figure 10a, the integrated SFI exhibited a unimodal response to increasing PAM concentration. The SFI_total increased significantly from the control to 3‰ PAM, reaching the maximum (0.755), and then showed a trend of first decreasing and then increasing with the further increase in PAM concentration. Although the overall SFI_total value peaks at 3‰ PAM, a notable recovery is observed at 7‰ compared to 5‰. This reversal is primarily attributed to the following mechanisms: at 7‰ PAM, the soil erosion resistance function (SFI₁) reaches its maximum value (Figure 4a). The dense polymer network formed at higher concentrations enhances particle binding and surface sealing, thereby improving soil structural integrity and substantially suppressing soil loss under erosive forces. Indicators such as critical shear stress (τ) increase markedly, while soil erodibility (K_r) decreases, reflecting enhanced structural protection against erosion. These effects have a strong positive influence on SFI_total. Simultaneously, SFI₂ remains high at 7‰ (Figure 6a), continuing the increasing trend observed across lower concentrations. This is attributed to improved water-holding capacity (higher θ_r and θ_v) and greater prevalence of NAPs, which slow drainage and increase moisture retention, supporting the elevated SFI₂ score.

The ecological function score (SFI₃) drops considerably at 5‰ PAM due to inhibited root development and poor aeration and deteriorates further at 7‰ (Figure 9a). Despite SFI₃ declining to its lowest value at 7‰, the cumulative improvements in SFI₁ and SFI₂ are sufficient to overcompensate for the ecological loss. This indicates a functional trade-off: further gains in erosion resistance and water retention were achieved at the expense of ecological compatibility. Given that SFI_total is a composite index, its magnitude depends on the collective. Thus, the rebound of SFI_total at 7‰ PAM is primarily driven by the synergistic improvement in soil physical stability and moisture retention capacity, which together outweigh the suppressed ecological performance. This emphasizes the importance of considering multi-functionality trade-offs when evaluating soil amendment strategies. This pattern underscores the existence of an optimal PAM threshold (3‰) for maximizing multifunctional soil benefits.

Figure 10b shows the pairwise relationships among SFI₁, SFI₂, and SFI₃ to SFI_total. The regression models exhibited strong significance: Among them, SFI₃ had the strongest correlation with SFI_total, suggesting that ecological function is the most sensitive and integrative indicator of soil multifunctionality. This also highlights the vulnerability of SFI_total to ecological disturbances when PAM is over-applied. In contrast, while SFI₁ and SFI₂ exhibited a monotonic increase, its correlation with SFI_total was slightly weaker, indicating that erosion resistance and water regulation function cannot fully explain multifunctional soil health without concurrent improvements in ecological function.

Mantel test results (Figure 10c) evaluated the correlation between SFI_total and a suite of soil pore structure parameters. The results reveal strong positive correlations (R > 0.7, p < 0.001) between SFI_total and mesopore surface area (MSA_e), specific surface area (SSA), micropore-specific surface area (MSS_ai), average adsorption pore width (AAP_w), and total pore volume (TPV). This suggests that improvements in both the quantity and quality of pore architecture, particularly those associated with fine-scale pore surfaces and connectivity, play a critical role in enhancing soil multifunctionality under PAM treatment. A higher MSA_e and MSS_ai implies greater exposure of functional pore interfaces, which enhances root–soil contact, aeration, and microbial colonization. These pores also facilitate preferential flow and moderate water storage, directly supporting SFI₁ by stabilizing aggregates and SFI₃ through better root development conditions. A larger TPV reflects enhanced total storage capacity, improved infiltration, and root penetration space. The increase in TPV under PAM application contributes to all three functional dimensions, particularly SFI₁ by facilitating rapid infiltration, and ecological function SFI₃ via improved root exploration zones. In contrast, SFI_total showed a negative correlation with WFT. An increase in water film thickness is often indicative of reduced pore connectivity or over-compaction, which may hinder root respiration, microbial mobility, and capillary water movement. This underscores that excessive water retention can detract from multifunctionality despite improving isolated water-holding indicators.

These results highlight the importance of maintaining a well-balanced pore architecture to support the synergistic expression of physical, hydraulic, and biological soil functions under PAM treatment, demonstrating PAM induces divergent responses across soil functional domains, leading to a concentration-dependent threshold effect. While high concentrations (>5‰) may be appropriate for short-term erosion control in severely degraded slopes, they are ecologically unsustainable. In contrast, a PAM application rate of 3‰ provides the optimal balance among competing functions, maximizing structural and hydraulic improvements without compromising ecological performance. Therefore, precise management of the PAM dosage is essential to maintain or enhance comprehensive soil health and functionality in subtropical granite-derived sandy slopes.

4. Discussion

4.1. Effects of PAM on Soil Anti-Erosion, Water Retention, and Ecological Functions: Microstructural Mechanisms and Threshold Responses

Soil degradation in granite-derived sandy landscapes is characterized by the simultaneous deterioration in physical structure, moisture buffering capacity, and ecological viability [51]. Although polyacrylamide (PAM) has been widely applied to improve soil aggregate stability and infiltration, previous studies have largely focused on isolated parameters, with limited insight into how PAM affects multiple soil functions concurrently, especially across a gradient of application rates [52]. The ecological risks of over-application also remain poorly understood. Addressing these gaps, this study adopts an integrated multifunctionality framework based on three soil function indices, erosion resistance (SFI₁), water regulation (SFI₂), and ecological function (SFI₃), and combines it with SEM imaging and BET pore structure analysis to elucidate how microstructural changes regulate these functions across varying PAM concentrations.

The results revealed a strong concentration-dependent differentiation among soil functions. Low to moderate PAM additions (1~3‰) consistently improved all three SFI components (Figure 10a). However, higher concentrations (≥5‰) led to marginal hydraulic gains and ecological collapse, indicating the presence of functional thresholds driven by PAM-induced structural saturation and overcompaction (Figure 10b). These thresholds can be mechanistically explained by divergent pore-scale responses.

Erosion resistance (SFI₁) increased monotonically across the concentration gradient, reaching its peak at 7‰. This improvement is primarily driven by progressive aggregate densification and enhanced interparticle cohesion [53]. PAM’s long molecular chains act as flocculants and bridges, physically linking soil particles and reinforcing aggregate boundaries [54]. SEM micrographs demonstrated a clear transition from loose, skeletal soil fabrics to compacted, polymer-bound microaggregates (Figure S3). This structural reorganization reduces macropore connectivity and pore-wall exposure, thereby minimizing detachment zones under surface runoff [55] (

Table 4. Quantitative micropore structure data derived from SEM analysis.

PAM (‰)	MPD (μm)	APA (μm2)	P (%)	PND	FDps
0	33.780 ± 0.589	1376.616 ± 2.847	9.581 ± 1.539	765.573 ± 110.382	1.389 ± 0.036
1	31.784 ± 0.014	1283.664 ± 0.635	11.545 ± 0.242	997.581 ± 6.423	1.446 ± 0.006
3	33.247 ± 0.520	1337.445 ± 25.627	11.675 ± 0.137	1005.088 ± 26.968	1.430 ± 0.009
5	33.724 ± 0.098	1383.286 ± 2.520	11.136 ± 0.313	950.193 ± 1.565	1.452 ± 0.003
7	32.698 ± 0.755	1272.991 ± 4.001	11.694 ± 0.003	948.164 ± 1.067	1.459 ± 0.004

Note: MPD is the mean pore diameter; APA is the average pore area; P is the porosity; PND is the pore number density; and FD_ps is the fractal dimension of pore structure.

). As pore spaces contract and aggregate integrity improves, the critical shear threshold increases, leading to reduced sediment detachment (Figure 3b,c). Although the internal friction angle (φ) showed only minor fluctuations, the increased aggregate integrity and decreased shear vulnerability underpinned the consistent rise in SFI₁.

Water regulation (SFI₂) exhibited a nonlinear response (Figure 6a). At 7‰ PAM, SFI₂ reached 0.980, an increase of 89.8% over the control, supported by an 84.4% increase in volumetric water content (θ_v) and an 8.1% rise in non-active pores (NAPs). These improvements are attributed to PAM-induced micropore (<0.2 μm) formation that plays a central role in enhancing the soil’s short-term water retention capacity [56]. Moderate PAM concentrations promote the development of capillary-sized pores and increase specific surface area, as evidenced by BET analysis (Figure 7c,d). These micropores store water via film adsorption and matric suction, stabilizing water content during desaturation [57]. This shift is not merely volumetric but functional: it increases the plant-available water without compromising infiltration [58].

In contrast, ecological function (SFI₃) displayed a unimodal trend with a well-defined threshold. SFI₃ peaked at 3‰ (0.858) (Figure 9a), reflecting optimal root emergence and elongation under structurally stable, moderately porous conditions. While low-to-moderate concentrations support root elongation by providing stable water and physical anchorage, high concentrations generate microstructural constraints [59]. Overcompaction of micropores reduces oxygen diffusivity and shifts the water status toward bound water forms, neither hydraulically mobile nor biologically accessible [60]. The root system, highly sensitive to redox and mechanical cues, responds with shortened root length, reduced density, and poor emergence. In addition, PAM-induced surface films may increase soil hardness, adding mechanical resistance to root penetration [61]. These effects, acting synergistically, explain the ecological collapse observed at ≥5‰ PAM.

Taken together, these findings confirm that PAM regulates soil multifunctionality through microstructural reorganization. Its efficacy is function-specific and highly dosage-sensitive. The positive responses of SFI₁ and SFI₂ up to 5~7‰ contrast with the ecological threshold at 3‰, beyond which further compaction disrupts biological compatibility. The divergence of sub-index trends above this point illustrates a clear structure-driven trade-off: what improves physical functions may compromise biological ones. Therefore, PAM functions as a dual-edged agent, capable of restoring degraded soil functions when applied at optimized levels but posing ecological risks when overdosed. Restoration efforts in subtropical erosion-prone systems must thus calibrate PAM concentration not only for physical stabilization but also for long-term ecological sustainability.

4.2. PAM-Induced Soil Multifunctionality: Optimal Concentration and Practical Implications

Previous research has consistently demonstrated the beneficial effects of polyacrylamide (PAM) on individual soil physical properties such as aggregate stability, infiltration capacity, and erosion resistance [62]. However, these studies have primarily focused on single-function outcomes, offering limited insight into the compound effects of PAM on subtropical granite-derived sandy slopes, particularly under variable concentration regimes [63]. Moreover, the potential ecological risks associated with PAM overuse remain underexplored in ecologically fragile or restoration-sensitive landscapes. The present study advances this field by adopting a multifunctionality-based evaluation framework that integrates erosion resistance (SFI₁), water regulation (SFI₂), and ecological function (SFI₃). This approach combines principal component analysis, nonlinear indicator scoring, and microstructural characterization via SEM and BET analysis, enabling a mechanistic interpretation of how PAM influences soil function through structural modification. The observed non-monotonic responses, specifically, the divergence between physical gains and ecological suppression at concentrations ≥5‰, highlight the importance of identifying functional thresholds and trade-offs that are rarely captured in conventional studies.

From a practical standpoint, these findings suggest that PAM acts as a concentration-sensitive soil amendment, with distinct functional outcomes across application levels. At low to moderate concentrations (1~3‰), PAM simultaneously improves erosion resistance and water retention while preserving ecological compatibility. This makes it particularly suitable for use in ecological restoration projects and agricultural systems, where biological function must be sustained. In such contexts, complementary practices such as shallow tillage may further enhance PAM distribution and prevent aggregate over-compaction. In contrast, higher concentrations (5~7‰) offer rapid structural reinforcement and are more appropriate for emergency erosion control on severely degraded or exposed slopes, particularly during extreme rainfall events. However, these short-term benefits come at the expense of ecological function, as evidenced by a complete collapse of SFI₃ at 7‰, and thus necessitate follow-up ecological restoration measures, such as vegetation re-establishment or microbial inoculation. For granite-derived sandy material in subtropical monsoon regions, a PAM concentration of 3‰ is identified as the optimal threshold for multifunctional restoration. At this dosage, soil structural stability, hydrological behavior, and ecological performance are all simultaneously enhanced, providing a balanced and field-applicable strategy for integrated slope management. Beyond the biophysical trade-offs, the economic feasibility of PAM application is a key consideration for large-scale ecological restoration. Based on current market prices of PAM (approximately CNY 3.5–4.5 per kg), applying a 3‰ concentration to a soil mass equivalent to 1 ha·m depth would cost in the range of CNY 10,500–13,500 per hectare [64]. This investment is substantially lower than the costs associated with repeated mechanical stabilization measures or the use of higher PAM concentrations that may compromise ecological functions. Moreover, because the 3‰ dosage balances erosion resistance with vegetation establishment, it can reduce long-term maintenance costs by promoting natural recovery processes. Therefore, the 3‰ application rate represents a cost-effective compromise for sustainable soil stabilization and ecological restoration at the landscape scale.

Despite these contributions, several limitations warrant further investigation. This study was conducted under controlled conditions; thus, the long-term ecological impacts of PAM, particularly in response to natural rainfall variability, biological activity, and land-use dynamics, remain to be validated under field settings. The long-term persistence of PAM in soil systems remains a critical factor for the durability of its beneficial effects. Although PAM molecules can strongly interact with soil particles in the short term, they gradually degrade through photolysis, hydrolysis, and microbial activity under field conditions. Consequently, the stabilization and water-retention effects may weaken with time, particularly after multiple wetting–drying cycles or seasonal variations. This implies that the optimal concentration of PAM is not fixed but may require periodic adjustment or replenishment in multi-seasonal applications. It is also important to note that the ecological function index (SFI₃) developed in this study is based on grass root traits and is therefore valid only for comparing PAM treatments within the same soil type. Root system development is influenced not only by PAM application but also by the inherent nutrient status and fertility of the soil. Accordingly, applying SFI₃—and the overall SFI—across soils with different types or fertility levels should be interpreted with caution. This limitation emphasizes that the findings of this study are strictly applicable to C-horizon granite-derived sandy soils, and further validation is required before extending the index to other soil systems.

Therefore, precise management of PAM dosage is essential to maintain or enhance soil health and functionality, particularly given the variability in soil properties and environmental conditions. However, current understanding of the long-term effectiveness and ecological consequences of PAM application remains limited. Future studies should thus evaluate the temporal dynamics of PAM performance under field conditions with repeated application cycles, in order to provide reliable guidance for long-term ecological restoration strategies. Moreover, investigations should extend to different soil types, focusing on their physicochemical properties, and explore interactions between PAM and native soil microbial communities, including the potential synergistic or antagonistic effects when combined with organic or biological amendments. These efforts will help refine PAM-based management strategies, ensuring both their scientific rigor and broader applicability in sustainable land restoration practices.

5. Conclusions

This study evaluated the effects of polyacrylamide (PAM) on the multifunctionality of granite-derived sandy material by integrating erosion resistance (SFI₁), water regulation (SFI₂), and ecological performance (SFI₃) into a composite index. Results revealed clear concentration-dependent trends: SFI₁ and SFI₂ increased steadily with PAM dosage, peaking at 7‰, while SFI₃ followed a unimodal response, reaching its maximum at 3‰ and collapsing at higher concentrations. These patterns indicate distinct functional thresholds and highlight trade-offs induced by over-application of PAM. At 1~3‰, PAM enhanced aggregate stability and capillary pore development, improving cohesion, moisture retention, and root environment, leading to coordinated gains in SFI₁, SFI₂, and SFI₃. However, at ≥5‰, excessive micropore accumulation and reduced pore connectivity constrained aeration and increased bound water, impairing root growth and causing ecological function to decline sharply. This microstructural over-compaction explains the observed divergence in SFI trends and delineates a critical concentration threshold beyond which ecological compatibility is compromised despite continued physical benefits. Collectively, a PAM concentration of 3‰ is identified as optimal for multifunctional soil restoration, offering a balanced improvement in structural integrity, hydrological performance, and ecological viability. Higher concentrations may be appropriate for short-term erosion control but should be accompanied by ecological remediation strategies. These findings provide a mechanistic basis and practical guidance for concentration-calibrated use of PAM in sustainable slope and soil management. To better evaluate long-term ecological outcomes and soil biological dynamics, further research should be conducted under field conditions, with particular attention to the interactions between PAM, microbial communities, and organic amendments.