Investigating Soil Water Retention and Water Content in Retrogressive Thaw Slumps in the Qinghai-Tibet Plateau, China

Abstract

Retrogressive thaw slumps (RTSs) are becoming more common on the Qinghai-Tibet Plateau as permafrost thaws, but the hydraulic properties of thaw slumps have not been extensively studied. To fill this knowledge gap, we used the “space-for-time substitution method” to differentiate three stages of RTSs: original grassland, collapsing, and collapsed. Our study included on-site investigations, measurements in the laboratory, and measured and simulated analyses of soil water retention curves and estimated hydrological properties. Our findings show that the measurements and simulated analyses of soil water retention were highly consistent across RTSs, indicating the accuracy of the Van Genuchten model in reproducing soil hydraulic parameters for different stages of RTSs. The original grassland stage had the highest soil water retention and content due to its high soil organic carbon (SOC) content and fine-textured micropores. In contrast, the collapsed stage had higher soil water retention and content compared to the collapsing stage, primarily due to increased proportions of soil micropores, SOC content, and lower bulk density (BD). From original grassland stage to collapsed stage, there were significant changes on the structure of each RTS site, which resulted in a decrease in SOC content and an increase in BD in general. However, the absence of soil structure and compaction led to the subsequent accumulation of organic matter, increasing SOC content. Changes in field capacity, permanent wilting point, and soil micropore distribution aligned with variations in SOC content and water content. These findings highlight the importance of managing SOC content and water content to mitigate the adverse effects of freeze-thaw cycles on soil structure and stability at different thaw collapse stages of RTSs. Effective management strategies may include incorporating organic matter, reducing soil compaction, and maintaining optimal water content. Further research is needed to determine the most suitable management practices for different soil types and environmental conditions.

1. Introduction

The thawing of permafrost, including seasonal permafrost, due to melting ice has significant implications for Arctic landscapes and ecosystems [1,2,3]. This phenomenon results in the formation of thermokarst landforms known as retrogressive thaw slumps (RTSs) [4,5,6,7], which are prevalent in permafrost regions globally, including the Qinghai-Tibet Plateau (QTP) in China [8,9,10]. RTSs stabilize gradually as air temperatures drop below 0 °C, and ground ice ceases to melt. Mechanical or thermal erosion triggers these RTSs, often manifesting as minor disturbances in the base or slope breaks, aligning with water flow trajectories or shallow landslides that remove insulating materials from the surface and expose the ice-rich permafrost [11,12]. Thawed materials are seasonally transferred downslope through gradual creep or episodic surface or deeper flows, resulting in periodic activity and instability in slopes affected by thaw slides [13,14,15,16,17]. Freeze-thaw cycles influence the rates of ice melting and thaw slumping in RTSs, leading to the destruction of the original soil structure. In the summer, RTSs reactivate, causing vegetation and soil on slopes to flow away and re-expose ground ice. Climate warming can alter the rate of ice ablation and degradation by modifying the net radiative flux or sensible heat flux, thereby impacting the RTS rate of slopes [3]. Generally, the thawing of ice-rich permafrost on sloping terrain can result in the thermokarst development of RTS, initiating in the middle of a slope and advancing until stabilization [4,7,18,19,20]. RTSs significantly impact alpine meadow grasslands by modifying soil structure, vegetation coverage, soil hydraulic properties, and inducing soil erosions [21,22]. Studying soil water retention and hydraulic properties in alpine meadows at different thaw slump stages of RTSs is vital for understanding the mechanisms affecting water availability in these fragile ecosystems, assessing the impacts of climate change, and informing management and conservation strategies. However, the widely investigated hydraulic characteristics of RTSs and comprehensive field surveys are still lacking.

Previous investigations have revealed certain trends in alpine grassland degradation, such as an increase in soil bulk density and a decrease in soil moisture content and soil organic matter content [23]. However, the Qinghai-Tibet Plateau is characterized by temporal and spatial heterogeneity, leading to inconsistent findings regarding the factors influencing soil water retention in alpine ecosystems [24,25,26]. For instance, grassland degradation in the eastern Tibetan Plateau has been associated with a decrease in field capacity and soil water content [27]. In contrast, in Maqu County of the Qinghai-Tibet Plateau, alpine meadow degradation initially results in an increase and then a decrease in field capacity [28]. Moreover, the effects of soil texture on soil water retention have been inconsistent across different studies [24,29]. Previous research has predominantly concentrated on the effects of human-induced degradation on alpine meadow grasslands and alterations in soil water retention. However, limited field investigation has been conducted on the characteristics of degradation and changes in soil water retention resulting from natural factors such as the freeze-thaw process observed in retrogressive thaw slumps [30,31,32]. These retrogressive thaw slumps have a significant impact on land surface properties in the Qinghai-Tibet Plateau, leading to the redistribution and decrease of soil organic matter content [31,32]. Soil water retention curves (SWRCs) offer valuable insights into the water-holding capacity of soil and can be utilized to analyze the response of soil water retention to freeze-thaw cycles [33,34,35]. These curves provide a comprehensive understanding of how soil moisture content changes with varying soil water potentials, allowing for a more precise evaluation of the impact of freeze-thaw processes on soil water retention. The “space-for-time” substitution method enables a quantitative analysis of the impacts of freeze-thaw cycles and retrogressive thaw slumps on soil water retention and hydraulic properties [36]. This method involves studying sites with different stages of freeze-thaw processes, ranging from intact areas to fully degraded slumps, and inferring the temporal changes in soil water retention based on the spatial variations observed. By employing this approach, researchers can gain a better understanding of how freeze-thaw cycles and retrogressive thaw slumps influence soil water retention and the associated hydraulic characteristics. To address the gaps in knowledge, this study conducted field sampling and laboratory experiments to quantify soil water retention curves, soil water content, and hydraulic properties at different stages of RTSs. The Van Genuchten (VG) model was used to fit SWRCs based on experimental data and simulations [37]. Pedotransfer functions were utilized to determine the field capacity and the permanent wilting coefficient, which serve as indicators for available water in plant-soil systems [38,39,40]. The study aimed to explore the variation characteristics of soil water retention and hydraulic properties in different thaw slump stages of RTSs and analyze the factors and mechanisms through which thaw slumps influence soil water retention and hydraulic properties.

2. Methods

2.1. Study Area and Field Investigation

The study was conducted in typical RTSs located at various latitudes and longitudes on the northeastern QTP (Figure 1). Four representative RTS sites were selected for the study: Maduo, Xinghai, Dari, and Gangcha. The vegetation type is Kobresia pygmaea, and the soil type is subalpine meadow soil. These sites are considered typical examples of RTSs on the QTP, but they differ in their collapsing stages. While the Maduo site is experiencing slumping, with simultaneous slumping of soil and vegetation resulting in an increase in bare soil and a decrease in vegetation, the collapsing stage of the remaining three sites has gradually become stable, characterized by the absence of vegetation and mainly bare soil.

Table 1. Information of selected study sites of Retrogressive thaw slumps (RTSs).

Study Site	Latitude	Longitude	Elevation (m)	Mean Annual Air Temperature (°C)	Highest Temperature (°C)	Lowest Temperature (°C)	Annual Precipitation (mm)
Maduo	34°29′ N	98°02′ E	4355	−1.45	8.8	−13.8	409
Xinghai	35°49′ N	99°55′ E	3900	0.55	11.5	−11.9	380
Dari	34°08′ N	99°28′ E	4329	−3.3	8.1	−14.9	540
Gangcha	37°53′ N	98°25′ E	3515	−6.2	6.6	−18.9	313

presents the land surface information for each site. With the exception of Xinghai, where the average annual temperature is above 0 °C (0.55 °C), the other three sites have temperatures below 0 °C, with the lowest temperature recorded at the Gangcha site. The lowest and highest temperatures for all four stations were recorded in January and July, respectively. In January, the Gangcha site experienced the lowest temperature (−18.9 °C), while in July, the Xinghai site recorded the highest temperature (11.5 °C). The highest annual precipitation (540 mm) occurs at the Dari site, while the lowest annual precipitation (313 mm) occurs at the Gangcha site. The temperature data was downloaded from ERA5-Land hourly data from 1950 to present (https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-land?tab=overview, accessed on 1 August 2023). The precipitation data was downloaded from terraclimate (https://www.climatologylab.org/terraclimate.html, accessed on 1 August 2023) validated through measurements taken at the atmospheric meteorology site. All four sites have subalpine meadow soil as the predominant soil type, and the dominant species in these sites is the typical Kobresia pygmaea community, which is known for its rich organic matter content [41]. The RTSs of thermokarst landforms were classified into three stages based on field investigations, following the classification for the RTSs of Eboling Mountain on the northeastern QTP [42]. These stages are the original grassland stage, collapsing stage, and collapsed stage. Each stage represents a different degree of damage and collapse of the mattic epipedon and the Kobresia pygmaea community. In the original grassland stage, the Kobresia pygmaea community remains intact. The collapsing stage is characterized by the development of cracks and slumps in the mattic epipedon of the Kobresia pygmaea community, with vegetation starting to decrease and bare soil gradually increasing. Finally, the collapsed stage is characterized by severe damage to the mattic epipedon, resulting in little to no vegetation remaining (Figure 1b–e). To assess the degree of alpine meadow degradation, a space-for-time method was employed. This method relies on observing long-term changes in community succession stages to infer the degradation process [21]. In total, 48 undisturbed ring cutter soil samples were collected at the three stages of the four RTSs, with four replicated soil samples collected at each stage of each site. The ring cutter soil-sampling method was used [33], where each soil sample was placed in a ring cutter and sealed with filter paper to prevent evaporation. The collected undisturbed ring cutter soil samples mainly consisted of the top 5 cm of soil at different thaw slump stages of the RTSs. The four study sites were predominantly affected by natural factors, including freeze-thaw cycles, precipitation patterns, and net solar radiation.

2.2. Measurement of Soil Water Retention Curves and Soil Physical Properties

Four undisturbed soil samples were collected at each stage of RTSs using a ring cutter (height = 5 cm and volume = 100 cm³) at depth of 5 cm. The SWRCs were measured using a centrifuge approach (H-1400, KOKUSAN, Japan) at 22 °C (room temperature). The centrifuge had the capacity to test four soil samples simultaneously. Different centrifugal forces were applied to each soil sample, and soil water content was determined once equilibrium was achieved under the respective centrifugal force. This fast and accurate centrifuge allowed the determination of SWC under multiple centrifugal forces, which could be used to fit SWRCs eventually. Although the method of measuring SWRCS through centrifugation is more precise, it is also time consuming and labor intensive. The rotational speed of the centrifuge was converted to the soil matrix potential [43]. Rotational speeds of 300, 400, 500, and 600 rpm (revolutions per minute) etc., were set according to different time intervals such as 20 min, 30 min, 1 h, and 1.5 h etc., respectively. After reaching equilibrium, the centrifuged soil samples were weighed, and a laboratory oven was used to obtain the mass water content (θ_m) [44]. The mass water content of soil was measured at each rotation speed (suction) and subsequently used to calculate the volumetric water content (θ_v).

The value of the soil volumetric water content θ_v (cm³ cm⁻³) was derived from the mass water content θ_m (%) using the following formula:

$${\theta }_v=\frac{{\theta }_m{\rho }_b}{{\rho }_w}$$

(1)

where ${\rho }_b$ represents the bulk density (g cm⁻³) of the soil and ${\rho }_w$ (g cm⁻³) denotes the density of water.

The rotational speed of the centrifuge is converted into matric soil water potential, and the rotational speed is usually expressed in units of angular velocity, such as radians per second (rad s⁻¹). The equation used to calculate the average distribution of matric soil water potential, $\overline{h}$ (kPa), is as follows [43]:

$$\overline{h}=\frac{k}{6}\frac{{\omega }^{2}L}{g}\left[L-{3r}_e\right]$$

(2)

where g is the acceleration of gravity (9.8 m s⁻²), and k is a constant with a value equal to 0.098 kPa cm⁻¹, L is the length of soil sample (cm), ω is the angular velocity (rad s⁻¹), r_e is the outer radius of the centrifuge (cm).

pF is the abbreviation for soil suction and the use of pF originated from the study of soil water retention curves, and it can be converted from matric soil water potential [43]. The specific formula is as follows:

$$pF=-{log}_{10}\psi$$

(3)

pF is soil suction and ψ is matric soil water potential (kPa).

The soil organic carbon contents were measured with a carbon nitrogen analyser (CN802, VELP, Italy) in this study [45,46].

2.3. Simulation of Soil Water Retention Curves with Associated Soil Hydraulic Properties

The VG model and experimental data were employed to fit SWRCs and effectively depict the relations between matric suction and SWC [37].

$$\frac{{\theta }_v-{\theta }_r}{{\theta }_s-{\theta }_r}={\left(\frac{1}{1+{\left(\alpha \mid h\mid \right)}^n}\right)}^m$$

(4)

where θ_v is the volumetric water content (cm³ cm⁻³); θ_s denotes the saturated water content (cm³ cm⁻³); θ_r is the residual water content (cm³ cm⁻³); h is the pressure head (cm); α (cm³ cm⁻³), n, and m denote the parameters fitted by the VG model and are used to fit the shape of soil water retention curves, and m is obtained by n (m = 1 − 1/n).

The VG model can be utilized to fit SWRCs, which in turn enables the estimation of soil hydraulic parameters [33]. By analyzing and comparing the characteristics of the measured and fitted soil water retention curves, the volumetric water content corresponding to matrix potentials of −337.1 and −15,352.19 represents the field capacity (FC) and permanent wilting point (PWP), respectively. The plant available water capacity (PAWC) is determined as the difference between FC and PWP. To estimate the plant available water (PAW), the difference between measured soil water content (SWC) and PWP. The soil water storage (SWS) and plant available water storage (PAWS) for the 5 cm soil layer at the ground surface can be calculated using the following equations [47]:

$${SWS}_5=SWC\times {\rho }_b\times ∆d\times {{\rho }_W}^{-1}\times UCF$$

(5)

$${PAWS}_5=PAW\times {\rho }_b\times ∆d\times {{\rho }_w}^{-1}\times UCF$$

(6)

where SWS₅ is the soil water storage in the top 5 cm deep layer of the ground surface (mm), ${\rho }_b$ is the BD (g cm⁻³), $∆d$ is the thickness of the soil layer (cm), ${\rho }_W$ is the water density (g cm⁻³), and UCF is a unit conversion factor (10 mm cm⁻¹).

In the study, the soil pore-size distributions were also determined from SWRCs and the pore diameter was calculated using Jurin’s law equation. The specific equation, as provided by Gao et al. (2019) [48], is presented below:

$$d_e=\frac{4v\mathrm{c}\mathrm{o}\mathrm{s}\alpha }{{\rho }_{w}g\mid h\mid }$$

(7)

where d_e is the pore diameter (μm), h is the pressure head (cm), v is the water surface tension (75 × 10⁻³ N m⁻¹), α is the contact angle between the soil and water and set to zero in the study, ${\rho }_W$ is the water density (10³ kg m⁻³), and g is the acceleration of gravity (10 m s⁻²). Using these values, Equation (7) can be simplified as follows:

$$d=\frac{3000}{\left|h\right|}$$

(8)

In this study, soil pores were categorized into three groups according to their diameters: micropores (<30 μm), mesopores (30–100 μm), and macropores (>100 μm) [49]. Consequently, in accordance with Equation (8), mesoporosity was calculated based on the SWC in the range of 30–100 cm in absolute value, whereas microporosity was determined based on the SWC at a pressure head of 100 cm in absolute value. Macroporosity was derived by subtracting the combined values of mesoporosity and microporosity from the saturated water content (θ_s).

2.4. Evaluation of Simulation Performances

SWRCs were simulated using the VG model in MATLAB R2020b (Math Works Inc., Natick, MA, USA). The simulation effect of the VG model was evaluated using the coefficient of determination (R²) and root mean square error (RMSE). The fitting parameters, SWC, proportion of pore size, and SOC in different thaw slump stages of RTSs were compared. Two-way analysis of variance (ANOVA) and least significant difference methods were used to compare the fitting hydraulic parameters, soil water retention, SWC, soil hydraulic properties, soil pore-size distribution, and SOC for different RTSs sites and its thaw slump stages. Pearson correlation was used to quantify the relations between SOC content, soil properties, and soil water retention and content.

3. Results

3.1. Basic Soil Properties in Different Thaw Slump Stages

The soil water content (SWC) was measured at the 5 cm depth from the ground surface during the three stages of thaw collapse at four RTS sites: Maduo, Xinghai, Dari, and Gangcha (

Table 2. The average value of measured soil moisture content, bulk density and soil organic carbon content among three freeze thaw slump stages of four RTSs.

Sites of Thaw Slump	Stages of Thaw Slump	Soil Water Content (cm3 cm−3)	Soil Bulk Density (g cm−3)	Soil Organic Carbon (g kg−1)
Maduo	Collapsed	0.50 ± 0.03	0.62 ± 0.09	119.87
	Collapsing	0.52 ± 0.03	0.49 ± 0.05	129.52
	Original grassland	0.62 ± 0.01	0.40 ± 0.03	224.23
Xinghai	Collapsed	0.17 ± 0.03	1.33 ± 0.04	31.72
	Collapsing	0.12 ± 0.04	1.42 ± 0.12	19.84
	Original grassland	0.26 ± 0.04	0.91 ± 0.16	57.15
Dari	Collapsed	0.49 ± 0.08	0.79 ± 0.23	N/A
	Collapsing	0.33 ± 0.01	1.16 ± 0.01	N/A
	Original grassland	0.61 ± 0.01	0.66 ± 0.01	N/A
Gangcha	Collapsed	0.24 ± 0.02	1.1 ± 0.21	39.24
	Collapsing	0.19 ± 0.04	1.14 ± 0.13	46.89
	Original grassland	0.34 ± 0.01	0.75 ± 0.14	N/A

Note(s): N/A is not available. The soil organic carbon content at each stage has only one value, so the standard deviation cannot be obtained.

). Notably, Except for the lowest value of SWC appearing in the collapsed stage of Maduo site, the lowest SWC values of the remaining three sites occurred during the collapsing stage. The highest SWC was observed during the original grassland stage for all the RTS sites. The decrease in SWC during the collapsing stage or collapsed stage for all the four RTSs was attributed to thaw collapse caused by increased freeze-thaw action, causing a reduction in vegetation cover and destruction of soil structures. The degree of collapse increased with increasing thaw slumps, resulting in an initial decrease and then an increase or decrease from original grassland to collapsed stage in the ground surface SWC. Changes in soil bulk density were observed across the three thaw slump stages at the four RTS sites. The change of soil bulk density is opposite to that of soil moisture content, that is, the greater the soil bulk density, the smaller the corresponding soil moisture content. The SOC content at the Maduo and Xinghai sites of the RTS for the original grassland stage were the maximum value, minimum values occur during the collapsing or collapsed stage. The changing trend of soil organic carbon is consistent with soil moisture content.

3.2. Soil Water Retention Curves and Estimated Soil Hydrological Properties

The simulation results demonstrate excellent performance for SWRCs, exhibiting a high coefficient of determination (R²) of 0.99 and a negligible RMSE of less than 0.01 cm³ cm⁻³ (

Table 3. The fitting parameters of soil water retention curves (SWRCs) and the prediction performance of Van Genuchten (VG) model.

Sites of Thaw Slump	Stages of Thaw Slump	Fitting Parameters of SWRCs				Model Performance
		θr (cm3 cm−3)	θs (cm3 cm−3)	α (°)	n	RMSE (cm3 cm−3)	R2
Maduo	collapsed	0.00 ± 0.00	0.71 ± 0.02	0.02 ± 0.01	1.25 ± 0.03	0.012	0.995
	collapsing	0.07 ± 0.07	0.75 ± 0.04	0.02 ± 0.02	1.22 ± 0.03	0.011	0.993
	original grassland	0.00 ± 0.00	0.78 ± 0.02	0.01 ± 0.01	1.17 ± 0.02	0.013	0.992
Xinghai	collapsed	0.10 ± 0.02	0.48 ± 0.03	0.01 ± 0.00	1.42 ± 0.08	0.005	0.998
	collapsing	0.07 ± 0.04	0.46 ± 0.03	0.01 ± 0.00	1.37 ± 0.08	0.005	0.998
	original grassland	0.13 ± 0.10	0.65 ± 0.05	0.02 ± 0.01	1.28 ± 0.10	0.010	0.994
Dari	collapsed	0.00 ± 0.00	0.67 ± 0.06	0.01 ± 0.01	1.21 ± 0.01	0.009	0.996
	collapsing	0.00 ± 0.00	0.56 ± 0.02	0.00 ± 0.00	1.23 ± 0.03	0.004	0.999
	original grassland	0.00 ± 0.00	0.75 ± 0.01	0.01 ± 0.00	1.20 ± 0.00	0.006	0.998
Gangcha	collapsed	0.01 ± 0.02	0.52 ± 0.03	0.02 ± 0.01	1.18 ± 0.04	0.012	0.985
	collapsing	0.04 ± 0.04	0.53 ± 0.04	0.04 ± 0.04	1.23 ± 0.06	0.007	0.994
	original grassland	0.09 ± 0.10	0.68 ± 0.06	0.03 ± 0.03	1.26 ± 0.13	0.013	0.990

Notes: 4 sites × 3 states × 4 replications = 48 soil samples; the data represent mean ± standard deviation.

). These findings indicate that the VG model possesses the capability to accurately reconstruct SWRCs for soils in varying thaw collapse stages of QTP RTSs (Figure 2). Specifically, it has been observed that at a specific suction level, the higher the volumetric water content, the stronger the soil water retention during the three stages. Based on two-way analysis of variance (ANOVA) with two-factors and subsequent multiple comparison tests reveal significant disparities in the SWRCs for different sites and thaw slump stages, particularly at the Maduo site (Figure 3). The results of ANOVA showed a significant impact of different freezing and thawing stages on soil water retention/content, with different sites also causing a pronounced impact on it. It is worth noting that different sites and stages exerted a substantial interactive effect on soil water retention/content. Soil water capacity indicates the corresponding soil water content under different soil suction forces. Notably, a significant difference in soil water retention/content is observed between the collapsed stage and the original grassland stage at the Maduo site (p < 0.05). Conversely, significant differences are observed between the original grassland stage and the collapsing stages at the other RTSs, with probabilities of 0.029, 0.024, and 0.013 at the Xinghai, Dari, and Gangcha sites, respectively. For the four RTSs, the original grassland stage exhibits the highest soil water retention compared to the other stages. By contrast, the degree of thaw collapse is relatively mild at the Maduo site compared to the other three sites, where the collapsing stages experience severe thaw collapse and lack vegetation cover. Consequently, water retention in the collapsing stage of the Maduo site surpasses that in the collapsed stage, while the opposite is observed in the other three RTSs. Based on the analysis of SWRCs, it can be concluded that the original grassland stage in the four representative RTSs of QTP exhibits the highest soil water retention.

The model has demonstrated that the related parameters (such as θ_s and θ_r) possess reasonable values, accurately representing soil hydraulic properties. The fitting parameters of SWRCs are explicitly presented in Table 3. Generally, there is a gradual decrease in saturated water content from the original grassland to the collapsed stage. Compared to the collapsed stage, the original grassland exhibits a 10% higher saturated moisture content at the Maduo site, a 26% higher content at the Xinghai site, an 11% higher content at the Dari site, and a 23% higher content at the Gangcha site. The FC, PWP, and PAWC at the three different stages of thaw slump at the RTS sites were determined using SWRCs (Figure 4). From original grassland stage to the collapsed stage, the FC shows a decreasing trend, indicating a 29% decrease at the Maduo site. At the Xinghai, Dari and Gangcha sites, FC all first increase and then decrease, resulting in overall reductions of 15%, 13% and 20%, respectively. The changes in PWP from the original grassland stage to the collapsed stage follow the same trend as FC. PWP exhibited a gradually decreasing trend, resulting in an overall reduction of 47%. In the remaining three sites, it shows a trend of first decreasing and then increasing, leading to overall declines of 33%, 17% and 24% at Xinghai, Dari and Gangcha sites, respectively. PAWC first decreased and then increased (except for the Xinghai site, where it steadily decreased) from the original grassland stage to the collapsed stage in all four RTS sites. In the Xinghai site, PAWC continuously decreased, experiencing an 11% decrease overall. At Maduo, Dari and Gangcha sites, PAWC first decreased then increased, resulting in overall decreases of 7%, 10% and 14%, respectively. Overall, FC, PWP, and PAWC values in the original grassland stage of each RTS site were greater than those in the collapsing and collapsed stages. Consequently, the values of FC, PWP, and PAWC decreased as the intensity of freeze-thaw slumping increased, particularly in the collapsing stage.

3.3. Plant Available Water, Soil Water Storage and Plant Available Water Storage Changes across Thaw Slump Stages

Table 4. Soil water storage (SWS), plant available water storage (PAWS) and plant available water (PAW) at the different stages of thaw slump in the four RTS sites.

Sites of Thaw Slump	States of Thaw Slump	SWS (mm)	PAWS (mm)	PAW (mm)
		Total	Total	Total
Maduo	collapsed	15.44 ± 0.85	4.96 ± 0.63	0.16 ± 0.03
	collapsing	12.8 ± 0.67	5.63 ± 0.88	0.23 ± 0.04
	originalgrassland	12.48 ± 0.3	8.8 ± 0.76	0.44 ± 0.02
Xinghai	collapsed	11.35 ± 1.71	0.40 ± 0.11	0.01 ± 0.00
	collapsing	8.74 ± 2.99	0.00 ± 0.00	0.00 ± 0.00
	Original grassland	11.70 ± 1.76	0.94 ± 0.59	0.02 ± 0.01
Dari	collapsed	19.22 ± 3.22	9.09 ± 3.70	0.23 ± 0.09
	collapsing	19.17 ± 0.70	5.8 ± 0.15	0.10 ± 0.00
	originalgrassland	15.67 ± 0.28	7.65 ± 0.23	0.30 ± 0.01
Gangcha	collapsed	13.21 ± 1.06	2.73 ± 0.89	0.05 ± 0.03
	collapsing	10.20 ± 2.08	0.57 ± 0.03	0.01 ± 0.00
	originalgrassland	12.57 ± 0.48	3.00 ± 0.56	0.08 ± 0.03

Notes: Total represents the total water storage of the 5 cm soil profile. The data represent mean ± standard deviation.

summarized the variation characteristics of plant available water (PAW) observed during the three different thaw slump stages of RTS sites. At the Maduo site, the PAW values for the original grassland, collapsing, and collapsed stages were estimated as 0.44, 0.23, and 0.16 cm³ cm⁻³, respectively. At the Xinghai site, the PAW values were estimated as 0.02, 0.00, and 0.01 cm³ cm⁻³, respectively. At the Dari site, the PAW values were estimated as 0.30, 0.10, and 0.23 cm³ cm⁻³, respectively. At the Gangcha site, the PAW values were estimated as 0.08, 0.01, and 0.05 cm³ cm⁻³, respectively. The results showed that among the four RTS sites, the highest PAW value was observed at the original grassland stage, followed by the collapsed stage, with the smallest value in the collapsing stage. Therefore, compared with the collapsed stage, the original grassland stage exhibited higher PAW values, suggesting that thaw slumps will cause a reduction in PAW during the three stages of four RTSs. Additionally, Table 4 provides information on the changes in soil water storage (SWS) and plant available water storage (PAWS) across the three thaw slump stages at the four RTS sites. At the Maduo site, the SWS values at a soil depth of 5 cm for the original grassland, collapsing, and collapsed stages were estimated as 12.40, 12.74, and 15.50 mm, respectively. At the Xinghai site, the SWS values were estimated as 11.83, 8.52, and 17.96 mm, respectively. At the Dari site, the SWS values were estimated as 15.56, 19.14, and 19.36 mm, respectively. At the Gangcha site, the SWS values were estimated as 12.38, 10.26, and 13.08 mm, respectively. Based on the data, the SWS value at a depth of 5 cm is the highest at the collapsed stage. At the Maduo site, the PAWS values at the 5 cm deep soil for the original grassland, collapsing, and collapsed stages were 8.80, 5.63, and 4.96 mm, respectively. At the Xinghai site, the values were 0.94, 0.00, and 0.40 mm, respectively. At the Dari site, the values were 7.65, 5.80, and 9.90 mm, respectively. At the Gangcha site, the values were 3.00, 0.57, and 2.70 mm, respectively. The values of SWS and PAWS were influenced by the complex freeze-thaw action and soil BD, and they did not exhibit a consistent trend across different thaw slump stages at the four RTS sites.

3.4. Relations of Soil Water Retention/Content with Other Soil Properties

Figure 5 illustrates the results of a linear correlation analysis examining the relationship between soil water retention/content, plant available water capacity, and several influencing factors. The results reveal that, among the several factors examined, soil BD exhibits a negative correlation with both SWC and PAWC, with correlation coefficients (R²) of 0.73 and 0.29, respectively. Conversely, SOC and soil microporosity display a significant positive correlation with SWC and PAWC, with correlation coefficients (R²) of 0.91 and 0.4 for SOC, and 0.99 and 0.76 for soil microporosity, respectively. It is worth noting that the original grassland stage, which maintained a higher SWC, can be attributed to its elevated levels of SOC and the presence of a well-developed mattic epipedon that remained undamaged [50]. Clearly, the content of SOC and the proportion of soil microporosity have a positive impact on both SWC and PAWC, while soil BD demonstrates a negative correlation with these parameters. Therefore, the correlation analysis highlights the crucial role of SOC, soil microporosity, and soil BD in determining soil water retention/content throughout the three thaw slump stages.

4. Discussion

4.1. Difference of Soil Water Retention across Thaw Slump Stages

Soil water retention is affected not only by plant characteristics but also by the physical structure of the soil, which plays an important role in surface processes and hydrological cycles in cold regions [24]. The combination of experiment and model simulations effectively demonstrates soil water retention, soil water content (SWC), and other soil hydraulic parameters under different freeze-thaw intensities and stages at the four RTS (retrogressive thaw slump) sites. Significant differences (p < 0.01) were observed between the soil water retention and content values of the original grassland and the collapsing or collapsed stages. However, no significant difference (p > 0.05) was found between the soil water retention and SWC values of the collapsing and collapsed stages. The collapsing and collapsed stages, identified by the geomorphic features and characteristic land surface, were connected, and their difference was not statistically significant (p > 0.05). For instance, soil water retention, SWC, FC, PAW, and other soil hydraulic parameters exhibited significant differences (p = 0.015) between the original grassland stage and the collapsed stage at the Maduo site, while the soil hydraulic parameter values at the collapsing stage and collapsed stage at the Maduo site did not exhibit significant differences (p = 0.24, i.e., p > 0.05). This study revealed that soil water retention at the original grassland stage was significantly higher than those at the other two thaw slump stages at the four RTS sites (Figure 2). This trend matches with the trend presented in previous research, indicating that as grassland degradation increases, soil water retention decreases [27,51]. However, other studies have shown that soil water retention does not necessarily decrease with vegetation degradation [24]. The influence of thaw slumps on hydraulic characteristics varies owing to the diverse vertical heterogeneity and vegetation distribution on the QTP as well as different causes of alpine meadow degradation, such as human activities, grazing, and freeze-thaw action [23]. Our findings suggest that as freeze-thaw intensity increases, represented by different stages of RTS, and vegetation coverage decreases, there is a decrease in soil water retention. However, in the collapsed stage, soil water retention and water content are higher than in the collapsing stage. These results indicate that the degradation of alpine meadows caused by grazing and human activities differs from thaw slumps induced by natural factors such as, continuous melting of underground ice in the active layer of permafrost and seasonally frozen ground, which can accelerate RTSs and amplify alpine meadow degradation. Negative feedback resulting from these factors can reduce the organic matter content in the soil, hindering plant growth and development in alpine meadow ecosystems, particularly under the backdrop of climate warming. Since 1950, the Tibetan Plateau has been warming at a rate of 0.32 °C per decade, twice the global average temperature increase [52]. The ground surface soil of the QTP exhibited high sensitivity to climate warming and caused permafrost degradation, and also had significant impact on hydrology [53,54]. Although remote sensing satellite technologies have been employed to detect and locate RTSs sites [55], limited research has investigated soil water retention and soil parameters of RTSs in different thaw slump (FTS) stages. Our findings indicate that the negative feedback between FTS and soil water retention can contribute to the long-term degradation of alpine meadow soil, exerting a negative influence on plant growth. Through a comparative analysis of soil water retention across four RTS sites, particularly during the collapsing stage characterized by intense freeze-thaw slumping and severe damage to soil structure, hydraulic parameters, and vegetation, we identified the significance of the mattic epipedon in retaining soil water in alpine meadows. The intensification of underground ice melting, ground surface subsidence, and degradation of the mattic epipedon inevitably leads to the destruction of alpine meadow vegetation. Therefore, safeguarding and managing the collapsing stage of RTSs alpine meadow landscapes is of utmost importance. We conducted a comparison of FC, PWP, and PAWC at various stages of FTS across the four RTS sites. In this study, from original grassland stage to collapsed stage, as freeze-thaw action intensified, the characteristics of FC and PWP variations aligned with those of soil water retention, emphasizing the impact of thaw slumps on soil structure and hydraulic properties in RTSs. Previous studies suggest that SOC plays a vital role in soil water retention due to its hydrophilic properties and influence on soil BD. Freeze-thaw collapse also induces changes in SOC, with considerably lower SOC values observed during the collapsing stage compared to the other two thaw slump stages. This phenomenon can be attributed to the destruction of soil structure during the collapsing stage, resulting in the destruction of plants and a gradual reduction in SOC to its minimum value. Furthermore, FC and PAW reached their lowest values during the collapsing stage under the influence of freeze-thaw slumping (Figure 3). This is due to the severe slumping, causing Kobresia pygmaea vegetation to slide along with collapsing earth blocks during freeze-thaw action, leading to their destruction and a significant decline in SOC content over time. Consequently, this process results in increased soil BD and ultimately reduces soil water retention.

4.2. Effects of RTSs on Soil Structure over Thaw Slump Stages

The soil structure consists of solid particles and a three-dimensional arrangement of voids, which have an impact on its capacity to maintain and transport air, water, and nutrients essential for plant growth and development. Additionally, it plays a crucial role in enhancing soil fertility, permeability, and erosion prevention [56]. The freeze-thaw process also influences the soil structure and must be considered when studying the characteristics of frozen soil structures. The proportion of soil pore sizes is a vital property of soil structures that affects their water retention capacity and water content. Previous research has indicated that alterations in soil pore-size distribution can enhance soil water retention and water content during shrub encroachment [33]. In this study, in order to explore the impact of RTSs on soil structure and the response of soil water retention on this impact, soil pores were classified into micropores (<30 μm), mesopores (30–100 μm), and macropores (>100 μm) [49], with micropores dominating in all three stages across the four RTS sites (

Table 5. Soil pore size distributions in three thaw slump stages of four typical RTSs in QTP.

Sites of Thaw Collapse	States of Thaw Slump	Porosity (cm3 cm−3)
		Microporosity (<30 μm)	Mesoporosity (30–100 μm)	Macroporosity (>100 μm)
Maduo	collapsed	0.57 ± 0.03	0.08 ± 0.02	0.05 ± 0.02
	collapsing	0.64 ± 0.03	0.07 ± 0.03	0.05 ± 0.04
	original grassland	0.72 ± 0.03	0.04 ± 0.01	0.02 ± 0.01
Xinghai	collapsed	0.44 ± 0.03	0.03 ± 0.01	0.01 ± 0.00
	collapsing	0.41 ± 0.04	0.04 ± 0.02	0.02 ± 0.01
	original grassland	0.52 ± 0.04	0.08 ± 0.01	0.05 ± 0.01
Dari	collapsed	0.62 ± 0.02	0.04 ± 0.03	0.02 ± 0.02
	collapsing	0.53 ± 0.00	0.02 ± 0.01	0.01 ± 0.00
	original grassland	0.70 ± 0.00	0.04 ± 0.01	0.02 ± 0.00
Gangcha	collapsed	0.44 ± 0.02	0.05 ± 0.09	0.03 ± 0.01
	collapsing	0.41 ± 0.06	0.06 ± 0.02	0.06 ± 0.05
	Original grassland	0.56 ± 0.05	0.07 ± 0.03	0.06 ± 0.05

Note: The data represent mean ± standard deviation.

). The proportion of mesopores and macropores was relatively small, with the proportion of mesopores slightly higher than that of macropores. At the Maduo site, the proportion of micropores gradually decreased from the original grassland to the collapsed stage, whereas the proportion of mesopores and macropores gradually increased, with the highest proportion of mesopores (0.08 cm³ cm⁻³) observed in the collapsed stage. Similarly, at the Dari site, the proportion of micropores also decreased gradually, with the highest proportion (0.66 cm³ cm⁻³) observed in the original grassland. The proportion of mesopores and macropores first increased and then decreased, reaching maximum proportions of 0.04 cm³ cm⁻³ and 0.02 cm³ cm⁻³, respectively, during the collapsing stage. In Xinghai and Gangcha sites, the proportion of micropores first decreased and then increased, with peak values of 0.70 cm³ cm⁻³ and 0.56 cm³ cm⁻³, respectively, occurring in the original grassland stage. The proportion of mesopores and macropores generally decreased from the original grassland to the collapsed stage, with maximum values of 0.04 cm³ cm⁻³ and 0.07 cm³ cm⁻³, and 0.01 and 0.06 cm³ cm⁻³, respectively, in the two RTS sites. Consistent with FC, the proportion of soil micropores was relatively small during the collapsing and collapsed stage, and the maximum values all appear in the original grassland stage, and the lower the proportion of soil micropores, the lower the FC values. Micropores in the soil function as water storage reservoirs, aiding in the retention of water [33], while mesopores act as conduits, providing air and water for plant growth [57]. The proportion of soil micropores was highest in the original grassland stage for the four RTS sites. Freeze-thaw collapse damages the soil structure, particularly by reducing the proportion of soil micropores during the collapsing and collapsed stages, consequently diminishing soil water retention and water content.

4.3. Controls of Differences in Soil Water Retention/Content for RTSs

Various factors, including soil texture, bulk density (BD), soil organic carbon (SOC), soil pore-size distribution, and vegetation coverage, have been identified as influencing soil water retention/content [58,59,60,61]. These factors interact with each other and play important roles in determining the ability of soil to hold and release water. Previous studies indicate that frequent freeze-thaw cycles indirectly influence soil hydrological property by modifying soil texture [34], soil structure [62] and vegetation in the RTS landform of the QTP. Previous studies have indicated that thaw slumps in the Eboling mountain induce soil texture changes in three distinct stages. Specifically, the clay content in the original grassland stage is the highest, followed by the collapsing stage, and the lowest in the collapsed stage. Conversely, the sand content exhibits opposite variation characteristics to the clay content [42]. Clay content is a crucial factor influencing soil water retention, as supported by correlation analysis indicating a strong positive relationship between clay content and soil water retention [63]. This finding aligns with the observation that the clay content is highest in the original grassland stage, which correlates with the highest water retention capacity. Previous studies have demonstrated that the degradation of alpine meadow grasslands brings about alterations in soil properties, significantly affecting soil water retention [24]. In the alpine meadows of the QTP, both RTSs and grassland degradation exert a substantial influence on soil texture, consequently affecting soil hydrological processes, including soil water retention. In comparison to grassland degradation, RTS induces more pronounced changes in soil structure and texture, leading to a substantial impact on soil water retention. The influence of freeze-thaw cycles on the soil texture and structure of RTS is prominently observed through a reduction in soil organic carbon (SOC) content and an increase in bulk density (BD). Previous studies have indicated that SOC plays a crucial role in affecting soil water retention in the Qinghai-Tibet Plateau alpine ecosystem [24,64,65]. The specific effects of SOC on soil water retention can be explained from two perspectives. First, a higher SOC content leads to increased vegetation coverage and plant growth, rendering it more resistant to freeze-thaw cycles while also accelerating thaw collapse due to the amplified presence of meltwater and rainfall. Second, SOC influences soil BD and the mattic epipedon [66], as supported by the correlation between SOC and soil BD in this study (R² = 0.45). SOC alters soil structure by reducing soil solid density and enhancing soil porosity (mesopores and macropores), ultimately affecting soil BD [24,67] and subsequently influencing soil water retention and content. In this study, the original grassland stage exhibited the highest SWC due to its elevated SOC content and the presence of Kobresia pygmaea, which remained undamaged. Previous research on the soil water retention of alpine hillside meadows in Dari County on the QTP indicates that greater surface coverage of alpine meadows corresponds to a stronger soil water conservation capacity [68]. Additionally, root adsorption also plays a vital role in soil water retention in alpine meadows [50,62]. The soil moisture content was also highest in the original grassland stage, with soil water remaining above 25% even under permanent wilting point of vegetation. Consequently, it can be concluded that SOC greatly influenced the highest soil water retention and content in the original grassland stage. The higher soil water retention and content in the collapsed stage, compared to the collapsing stage, may be attributed to lower topography, runoff, the meltwater of the alpine path, and a higher proportion of micropores in the soil. Soil pore size is a critical factor affecting soil water retention/content. The proportion of micropores in the ground surface soil is considerably high, while the proportions of mesopores and macropores are very low. Micropores exhibit a positive correlation with soil water retention/content (R² = 0.99) and PAWC (R² = 0.76). Conversely, mesopores and macropores show a negative correlation with SWC and PAWC, and their proportions are very low in the study. Pearson correlation analysis was also conducted to examine the three influencing factors on SWC and PAWC during the three thaw slump stages. The results indicate that the soil on the QTP is affected differently by freezing and thawing action, resulting in considerable soil heterogeneity. The SOC content at each RTS site exhibits a positive correlation with the Pearson correlation of SWC and PAWC, except for the Xinghai RTS site, where the SOC content is negatively correlated with PAWC owing to its steeper incline compared to the other three sites. In terms of the proportion of micropores, except for individual collapsing stage and collapsed stage, the overall soil micropore proportion exhibits a high correlation with SWC and PAWC, particularly during the collapsing stage. This finding is consistent with the low soil micropore proportion and the corresponding low SWC and FC values in the collapsing stage. Notably, the variation in the characteristics of micropores in the three collapse stages is consistent with the variation laws of SWC and FC, suggesting that the proportion of micropores in the ground surface soil considerably influences SWC and FC. This study examines the characteristics of soil water retention/content in three thaw slump stages of alpine meadows using space analysis instead of time analysis. It includes a comparative analysis and discussion of influencing factors. Understanding the soil hydrological processes and their influencing mechanisms in typical RTSs on the QTP is crucial. The soil organic carbon content in the original grassland stage of RTS is the highest and the mat layer is intact, corresponding to the highest soil water retention. During collapsing stage, the mat layer was damaged and the SOC content was reduced, and the soil water retention decreased. In the collapsed stage, fallen debris begin to accumulate, soil organic carbon content increases, and vegetation begins to appear. Interestingly, our findings show that when the collapsing stage is stable, the soil water retention/content of the soil in the collapsed stage is greater than that in the collapsing stage (Xinghai, Dari and Gangcha sites), and when the collapsing stage is slumping, its soil water retention is greater than the collapsed stage. This shows that the mat layer has an important influence on soil water retention/content.

Although our results are encouraging indicating that RTSs have a negative impact on soil water retention, this study has several limitations. Even though measuring soil water retention curves on 48 samples is expensive and time-consuming, more RTSs should be selected on the Tibetan Plateau for further verification. In addition, this study was determined through a single sampling survey and cannot represent changes in soil water retention in other periods except the growing season. Therefore, in order to more convincingly prove the interaction between RTSs and soil water retention in future studies, more RTS points should be selected to test the universality of these results, along with long-term field sampling experiments.

5. Summary and Conclusions

The retrogressive thaw slump action has a significant impact on soil water retention in alpine meadow soil, leading to changes in soil hydraulic parameters, other soil properties and vegetation coverage. The thaw slumps result in shifts in the proportion of micropores, which in turn affects the soil water retention. During the collapsing stage, there is a decrease in micropore proportion and soil organic carbon content, which subsequently leads to reduced soil water retention. However, it is worth noting that the soil water retention of the original grassland stage surpasses that of the collapsing and collapsed stages. This is due to a higher proportion of micropores and higher soil organic carbon content, which can improve the soil’s ability to retain water. This study emphasizes the negative feedback effect of thaw slumps on soil water retention/content. It provides a new perspective for understanding and monitoring hydrological changes in RTSs, as well as for establishing parameters for hydrological models in cold regions to develop effective strategies for fragile alpine meadow ecosystems management and conservation.