Influence of the Plateau Pika Mound Numbers on Soil Water Erosion Properties in Alpine Meadows of the Yellow River Source Zone, Western China

Abstract

The plateau pika (Ochotona curzoniae) actively contributes to soil erosion and meadow degradation in western China’s Yellow River source zone. This study aimed to elucidate the effects of the pika mound numbers on the hydrodynamic characteristics and soil water erosion through simulated rainfall experiments. The inhibition effects of restored vegetation growth on meadow degradation and soil erosion were explored using a revegetated pika mound as a control. The results showed that at a rainfall intensity of 30 mm/h, the soil loss per unit time increased and then decreased with rainfall time and that 15–20 min of rainfall duration was the sensitive period for soil loss in different pika mound patch lands. The degradation of meadows due to the activities of plateau pika is an essential factor influencing soil erosion, and the soil erosion rate is positively correlated with both the pika mound numbers and slope. The mean flow velocity can better describe the process of the soil erosion, and its value increased with the number of pika mounds and slope. The Reynolds number ranged from 57.85 to 153.63 (Re <500), and it was preliminarily determined that all slope runoff was laminar flow. The Froude number increased linear function with pika mound numbers (p < 0.01), and both the pika mound numbers and slope were significant factors affecting it (p < 0.05). The Darcy–Weisbach resistance coefficient instead decreased with the pika mound numbers and slope, and the inhibitory effect of vegetation on soil erosion was probably limited when the mound numbers reached a certain level. According to the results of the grey correlation and Pearson correlation analysis, changes in the number of pika mounds led to variability in the soil water erosion properties by altering the landscape scale effect. The number of pika mound patches (NP), edge length index (TE), area index (AREA), and volume index (V) were the key influencing factors on the soil erosion rate and hydrodynamic parameters. We conclude that plateau pika activities intensify meadow degradation, and the continuous increase in pika mounds decreases the vegetation cover and flow resistance and increases soil erodibility. Controlling the intensity of plateau pika activities will effectively prevent soil erosion in degraded areas.

1. Introduction

The Yellow River source zone is in the northeast of the Qinghai–Tibet Plateau, the major flow-producing area of the upper Yellow River and an essential water-conserving region in China [1,2]. Alpine meadows, as the most important cover type in the source zone of the Yellow River [3], account for about 80% of the total area, and their environmental changes directly affect the soil and water conservation and ecological security of the entire Yellow River Basin [4]. Recently, influenced by rodent activity, livestock production, and climate change, alpine meadows have been degraded to different degrees [5] and show a trend of increasing degradation yearly [6]. Plateau rodents are the local dominant biological species; the small rodents in the Yellow River source zone mainly include plateau zokor (Eospalax baileyi) and plateau pika (Ochotona curzoniae) [7]. At present, along with the phenomenon of meadow vegetation degradation and soil degradation, which also provides a good vision and site for the activities of plateau rodents, the local government’s early rodent extermination behavior has seriously affected the number of their predators, resulting in a continuous surge in their populations and an expansion of their activity range in recent years. In addition, although the activities of these rodents may have some positive effects on soil air circulation, nitrogen cycling, and vegetation community restoration, their burrowing and gnawing behaviors disrupt the natural vegetation community features and soil structures of alpine meadows [8,9,10], and the accumulated loose mounds indirectly form secondary bald patches of land and supply a material source for soil erosion [11,12]. So far, many scholars have concluded that rodent activity-induced meadow degradation areas stimulate soil erosion under various wind, water, or freeze–thaw dynamics and generate or exacerbate other natural hazards [13,14], negatively affecting the sustainability and security of the whole alpine meadow ecosystem [4,15].

Landscape patterns are the consequence of coupling diverse ecological processes [16], and studying the spatial pattern changes in landscape patches is significant to understand the soil degradation and vegetation degradation mechanisms [17,18]. The source zone of the Yellow River is influenced by rodent activities [19], and the landscape fragmentation is severe, among which the rodent mound patches formed by plateau pika activities lead to the reverse succession of the meadow ecosystem, which has a specific impact on community habitat texture heterogeneity, species diversity, and soil erodibility [9,20,21]. Landscape indices are quantitative indicators that reflect the structural composition and spatial configuration of the landscape, which serve as a bridge between landscape patterns and ecological processes [22], and currently, with the advancement of 3S technology and landscape ecology, studies on the effects of soil erosion based on landscape indices have been widely carried out [23,24,25]. For example, Zhao [26] et al. showed in riparian zone areas that vegetation fragmentation distribution increased soil erodibility; the landscape pattern influence on the soil erosion process depended on the landscape scale. Sun et al. [27] used the Markov model to associate the relationship between landscape indices and soil sand production and suggested that the landscape index could be an essential indicator to influence the soil erosion sand production process. However, due to the soil erosion process’s complexity and the landscape index’s limitations [28,29], selecting an appropriate landscape index will effectively advance the intrinsic association between changes in landscape patterns and soil erosion processes [30]. Furthermore, hydrodynamic characteristics are essential to reflect the soil erosion process on slopes and its kinetic mechanism [31]. However, there are still significant gaps regarding the response mechanisms of the slope surface flow hydrodynamic characteristics to landscape patterns, and few studies have reported the role of landscape patterns on hydrodynamic characteristics. Therefore, the study of the number, shape, volume, and spatial distribution pattern of plateau pika mound patches on soil erosion and hydrodynamic characteristics based on microhabitat scale is vital to investigate in depth the response mechanism of meadow patches to hydraulic soil erosion.

Soil erosion is the process by which surface soil and its parent material are separated, transported by a combination of internal and external forces, closely related to rainfall, topography, soil, and vegetation [32,33,34,35]. Rainfall is a major driver of soil erosion, and many scholars have shown that changes in the rainfall-runoff can directly affect the soil erosion conditions and hydrodynamic characteristics [26,36]; hydrodynamic characteristics serve as visual reflections of soil erosion processes, which are related to landscape patterns, slope, and vegetation cover [37,38]. For example, Chen et al. [39] suggested that landscape fragmentation can increase soil erosion by altering the hydrological connectivity and sediment transport pathways. Li et al. [40] showed that as the vegetation cover decreased from 75% to 25%, the runoff velocity increased, the runoff resistance decreased, and the soil anti-erosion capacity decreased. Sun et al. [41] found that the flow resistance decreased when the slope and precipitation intensity increased, and that the Reynolds number can accurately describe the process of soil erosion. Furthermore, in the context of the continuous deterioration of the ecological environment of alpine meadows, so far, although many scholars have made significant progress in research on soil erosion and hydrodynamic characteristics, few studies have been conducted to quantify the characteristics of soil water erosion of the pika mound patch lands in the Yellow River source zone.

This research attempts to fill the above knowledge gap by conducting simulated rainfall experiments in areas with frequent activities of plateau pika with the aim to clarify the influence of pika mound numbers on the soil erosion and hydrodynamic characteristics. The targets were to (1) clarify the soil erosion regimes under different pika mound numbers; (2) reveal the hydrodynamic characteristics of surface flow under different pika mound numbers; (3) elucidate the influence of pika mound numbers on the soil erosion and hydrodynamic characteristics; (4) explore the inhibitory effect of restored vegetation growth on meadow degradation and soil erosion. The study results help reveal the soil erosion mechanism in the degraded meadow regions by rodent activities and provide a theoretical basis for restoring and managing meadow ecological vegetation.

2. Materials and Methods

2.1. Study Area

This study area is situated in Henan County, Huangnan Prefecture, Qinghai Province (34°47′ N, 101°32′ E), which is part of the alpine meadow region in the source zone of the Yellow River (Figure 1a). The altitude is about 3600 m, the average annual wind speed is 2.6 m s⁻¹, and the annual precipitation is 597.1~615.5 mm, which is a typical continental plateau type climate. Rains are mainly centered in June–September of each year, and heavy rainfall events occur primarily in July–August, with rainfall durations ranging from 30 min to 180 min, characterized by short duration, suddenness, and small occurrence range [42,43]. The type of the meadow is mainly alpine dwarf tarragon meadow; the major plant species are Kobresia capillifolia, Kobresia humilis, Kobresia pygmaea, and Elymus nutans. The type of soil is humic cambisols (WRB), in which the undegraded areas are dominated by chalk and clay grains, while the degraded areas are dominated by sand grains and show significant coarse-graining and sanding. In addition, rodent activity is rampant in the study area, and many mounds have accumulated on the meadow surface (Figure 1b), among which plateau pika mounds are hemispherical and relatively gentle, with an average height of about 8 cm and a diameter of about 45 cm. The density of pika mounds is 0.36 mounds m⁻² (Figure 1c). The soil properties of the pika mounds are highly susceptible to more severe soil erosion, and eventually intensify meadow degradation due to the apparent differences that occur before and after (

Table 1. Basic physio-mechanical characteristics and the soil grain size class composition of pika mounds.

Pika Mounds Type	Moisture Content	Density	Porosity	Cohesion	Firmness	Below a Certain Soil Grain Size (%)
	(%)	(g m−3)	(%)	(kPa)	(kPa)	d < 2 mm	d < 0.5 mm	d < 0.075 mm
Fresh pika mounds	9.35 ± 1.75	0.87 ± 0.11	17.34 ± 0.16	11.56 ± 1.02	10.01 ± 1.95	88.79	59.72	8.88
Revegetated pika mounds	15.16 ± 2.08	1.51 ± 0.15	11.38 ± 0.14	26.61 ± 3.24	59.64 ± 2.38	75.26	48.97	8.15
Healthy meadows	38.25 ± 2.16	1.68 ± 0.08	4.75 ± 0.07	35.86 ± 3.53	153.50 ± 5.18	68.55	45.39	7.70

), having negative impacts on the regional ecological environment and development of livestock [44].

2.2. Experimental Design

This study mainly focused on the erosive effects of rainfall scouring on the different pika mound patch lands. Through the field survey, we found that the plateau pika is widely distributed in the study area, and there is no apparent requirement for slope in its habitat choice, resulting in some of the distribution areas of the pika mounds having slopes reaching 33–35°, and the percentage of the area with slopes exceeding 30° was about 9.24%. Therefore, we chose to set up three removable water erosion test chambers in the field that were 1.5 m × 3 m in area and 0.5 m in depth, while the slope was designed as 10°, 20°, and 30°, respectively. The bottom of the box had natural drainage holes, and the front end was equipped with a “V”-shaped converging sediment deflector hole to collect and test the sediment runoff volume (Figure 2a). The soil in the test box was filled with native meadow soil from the study area that had been closed for more than 4 years and contained a vegetative root–soil structure. In situ meadow squares of 50 cm × 50 cm × 50 cm were cut from healthy meadow areas in June 2020. These squares were then sequentially pieced together and laid into the corresponding water erosion test chamber. The edge joints between the meadow squares were artificially filled and repaired to produce a complete test site. All test boxes were left in the field for 1–2 years after completion of the above preparatory work, and soil water erosion tests were carried out after the seams between the meadow squares were fully healed and the vegetation cover was greater than 90% (Figure 2b).

The mounds in our study were fresh and uncrusted mounds formed within a short period after the soil was scattered and buried by vegetation death due to the mounding behavior of the plateau pika. The number of mounds in the test box was controlled to be 1–5, and the number of mounds was set to 1, 2, 3, 4, and 5 by artificially simulating mounds of the same shape, size, weight, and water content in the test boxes. The soil of the simulated mounds was selected from fresh mound soil accumulated by highland pika activities in the study area, and the more disturbed mounds were artificially restored twice. Since the initial vegetation cover in each test chamber was unequal, the meadow vegetation was artificially treated with equal cover (i.e., the vegetation cover was all reduced to about 70%). Immediately after completing the three sets of replicate tests, the number of mounds was increased to two by artificially simulating additional mound patches and so on, until a whole set of hydraulic erosion tests were completed one after another. In addition, we found during the field survey that the recovery of the dominant plant species always accompanied the succession process of the pika mounds. This study conducted a control test on the old pika mound with about 60% of the surface vegetation restored, which was comparable in terms of vegetation growth, external morphology, and basic soil properties.

The hydraulic erosion test used a field portable artificial rainfall simulator (Figure 3a). The device’s working principle is as follows: when connecting the rainfall device, extend the inlet pipe of the pump into the water tank, set up a layer of filtering net at the water intake to avoid sediment impurities into the pump, and finally, start the pump so that the water extracted by the pump is evenly sprinkled on the ground through the nozzle. The vertical distance of the rainfall nozzle from the center of the meadow surface in the test chamber is 3 m, which distributes rainfall evenly from the center of the 1 mm diameter hole. The effective radius of low rainfall is 1.5 m, and the rainfall intensity controller can regulate the precipitation intensity between 0 and 60 mm h⁻¹ with a simulated degree of ±0.5 mm h⁻¹. Based on the real-time monitoring data of intense rainfall weather in 2018–2021 from small meteorological stations in the test area and the test demand, the average rainfall intensity for all trials was set to 30 mm h⁻¹. In addition, to prevent the wind from affecting the trajectory and direction of the raindrops, the testing area was enclosed by a tarpaulin during the tests (Figure 3b).

This experiment recorded the time when runoff appeared from the inflow hole, the total precipitation duration was set to 1 h, and the runoff sediment samples were taken at 5-min intervals using catchment buckets. Meanwhile, soil erosion due to rain splash was included as a component of the measured sediment flux, considering that soil erosion due to rain splash is not only related to soil redistribution but is also challenging to separate from the lost soil. All of the samples collected in the field were taken back to the laboratory, and the sediment separation process was carried out by sedimentation and filtration methods. Furthermore, the separated samples were dried in an 105 °C oven to measure the amount and the rate of soil loss. In addition, each slope’s water temperature, runoff width, and runoff velocity were measured every 5 min during the sustained rainfall events. The runoff flow rate was determined by the dye (KMnO₄) traced method, and three measurement zones of 1 m, 2 m, and 3 m were set up from the top of the slope downward, with the length of the zones designed to be 0.8 m, and the mean flow rate of each zone was used as the real-time runoff flow rate of the slope.

2.3. Indicator Testing and Analysis

The different hydrodynamic parameters representing the runoff flow process on the slope were taken to accurately describe the effect of changing the number of pika mounds on the hydrodynamic characteristics. The hydrodynamic parameters selected for this study included the runoff depth (H), mean flow velocity (v), Froude number (Fr), Reynolds number (Re), and Darcy–Weisbach resistance coefficient (f); the specific formulae and their applicability are given in the citations [45,46].

The landscape pattern index is an essential means to quantitatively describe the landscape structure, function, and ecological process, which can highly concentrate the landscape pattern information including fundamental landscape elements such as patch number, shape, size, and distribution combination [47]. In this study, a total of six landscape pattern indices including the number index (NP), edge length index (TE), density index (PD), area index (AREA), volume index (V), and separation index (Ni) were selected at the patch, type, and landscape level to accurately describe the influence of the pika mound numbers on the soil water erosion and hydrodynamic characteristics. The calculation and ecological significance of these indices are listed in

Table 2. Landscape pattern indicators of pika mound patches.

Landscape Index	Equations	Description and Meaning of Indicators
NP	$\mathrm{N}\mathrm{P}=\mathrm{N}$	The total number of all pika mound patches (N) to reflect the landscape’s spatial pattern and the whole landscape’s qualitative heterogeneity.
TE	$\mathrm{T}\mathrm{E}=\mathrm{E}$	The sum of the edge lengths (E) of all pika mound patches can describe the degree of agglomeration and extension trends of a given patch type.
PD	$\mathrm{P}\mathrm{D}=\frac{\mathrm{N}}{{\mathrm{A}}_{\mathrm{i}}}$	The overall heterogeneity and fragmentation of the landscape and the degree of fragmentation of a given patch type are examined.
AREA	$\mathrm{A}\mathrm{R}\mathrm{E}\mathrm{A}=2\mathsf{\pi }\mathrm{r}\mathrm{h}$	The area of all pika mound patches is the basis for calculating other landscape indicators with implications for species abundance, energy nutrients, and productivity levels.
V	$\mathrm{V}=\frac{\mathsf{\pi }\mathrm{h}}{6}({3\mathrm{r}}^2+{\mathrm{h}}^2)$	The volume of all pika mound patches can describe the overall spatial characteristics of the landscape and accurately reflect the degree of landscape disturbance by external or human factors.
Ni	$\mathrm{N}\mathrm{i}=\frac{\mathrm{A}}{{2\mathrm{A}}_{\mathrm{i}}}\sqrt{\frac{\mathrm{N}}{\mathrm{A}}}$	The level of dispersion or aggregation of the individual spatial distribution of a specific type of patch in the landscape.

Note: A (m²) is the total landscape area, A_i (m²) is the area of landscape type “i”, r (m) is the pika mound patches radius, and h (m) is the height of the pika mound patches.

[26,48].

This study used Microsoft Excel 2021 software for data reduction and drawing charts, etc. In addition, based on the average value of the hydrodynamic parameters and soil erosion rate, we used IBM SPSS Statistics 26.0 software to analyze the statistical and correlation between the landscape pattern of the plateau pika mound patches and soil erosion and hydrodynamic characteristics.

3. Results

3.1. Soil Erosion Regulation under Different Pika Mound Numbers

3.1.1. Comparison of Total Soil Loss under Different Pika Mound Numbers

During this rainfall event (60 min for example), the total soil loss per unit area increased with the number of pika mounds and slopes (Figure 4). At the 10° slope, the soil loss from degraded patch lands with one, two, three, four, and five mounds was 2.12, 3.58, 5.11, 6.98, and 8.71 times higher than that of the revegetated pika mound, respectively. The increase in the number of mounds had a more significant impact on soil loss, and the total soil loss per unit area adding in a multiplicative power function with the number of pika mounds (p < 0.01). Compared to the fresh mounds under these different quantities, the amount of soil erosion was significantly reduced because the vegetation protected and ameliorated the revegetated mounds, resulting in a fundamental improvement in their soil characteristics. In addition, when the slope was increased from 10° to 30°, the total soil loss per unit area of the revegetated mound and patch lands under one, two, three, four, and five mound numbers increased by 63.97%, 66.41%, 59.70%, 56.44%, 43.93%, and 40.01%, respectively. The statistical analysis results showed significant differences in the total soil loss between different numbers of mounds on the same slope (p < 0.05). Between the same number of mounds on different slopes, except for the revegetated mound, which did not show any significant difference, the total soil loss under the other mounds showed significant differences (p < 0.05). Furthermore, we also found that as the number of pika mounds increased, the average slope of the different test plots increased by 0.3 to 0.6°, indicating that both the slope and number of pika mounds were significant factors influencing soil erosion.

3.1.2. Soil Loss Characteristics under Different Pika Mound Numbers

In Figure 5, the soil loss under the three gradients followed an upward and then downward trend with the time of rainfall erosion and reached a peak at about 20 min. The change in soil loss showed a significant increase (p < 0.05) during the period of rainfall until 15–20 min, and the soil loss generated during this period contributed the most to the total soil loss, indicating that 15–20 min of rainfall erosion was a sensitive period for soil loss in these experiments. It can also be seen from the figure that the magnitude of soil loss variation in the pika mound patch lands at different slopes were significantly different. Compared to the revegetated pika mound, soil loss under the remaining number of mounds increased significantly with the number of mounds (p < 0.05). In addition, combined with the observation results, we found that the height of the mound settled with 1.36 cm h⁻¹, the radius of the mound expanded with 2.45 cm h⁻¹, and the shape of the mound changed from a hemispherical mound before rainfall to an oblate spherical mound after rainfall. The surface layer of the mound was gradually exposed to more sand and gravel during the rainfall process, but the overall content was relatively small. The enrichment of the sand and gravel phenomenon did not appear until the end of the rainfall event, where there were only a few rain traces and imbricated texture by rain splash. The sediment size changed significantly before and after rainfall, for example, on a 20° slope, the average particle size of the sediment particles were 0.31 mm in 20 min before rainfall and 0.18 mm in 20 min after rainfall. Therefore, we suggest that the soil loss status of degraded meadow areas due to plateau pika activities is related to the slope, rainfall duration, and the number of pika mounds. Under the continuous high-intensity rainfall conditions, the soil loss was serious when the slope was steeper and contained more mounds.

3.1.3. Soil Erosion Regimes under Different Pika Mound Numbers

Under the precipitation strength of 30 mm h⁻¹, the soil erosion occurred in all test plots after 6.5 min of sustained rainfall, regardless of the slope and number of mounds. As shown in Figure 6, the soil erosion rates varied significantly between different numbers of pika mounds at the same slope. On a 10° slope, for example, the soil erosion rate varied from 0.61 to 4.68 g m⁻² min⁻¹ in the revegetated mound and the degraded patch lands with one mound; from 2.07 to 9.93 g m⁻² min⁻¹ in the degraded patch lands with two and three mounds; and from 4.87 to15.42 g m⁻² min⁻¹ in the degraded patch lands with four and five mounds. Furthermore, as the gradient reached 20° or 30°, the soil erosion rate increased significantly (p < 0.05), except for the revegetated mounds, and was positively correlated with the slope as a linear function. Overall, when the mounds were about the same size and weight, the soil erosion patterns were composed of two phases: a sharp rise and gentle decline, and the soil erosion rate peaked at about 25 min. It is also worth mentioning that an increase in mound numbers significantly enhanced the growth trend in the first phase and the fluctuation range in the second phase compared to the vegetation restoration of mouse mounds. This effect became more evident with the increasing number of pika mounds and slopes.

3.1.4. Relationship between the Pika Mound Numbers and Soil Erosion in a Degraded Meadow Area

By establishing a functional relationship between the pika mound numbers and vegetation cover with the soil erosion rate (Figure 7), we concluded that the soil erosion rate showed a linear positive correlation with the number of pika mounds (p < 0.01) while there was a linear negative correlation with vegetation coverage (p < 0.01). The results indicate that soil erosion in the pika mound patch lands was influenced by the number of pika mounds and related to the vegetation cover of the meadows destroyed by the activities of plateau pika.

3.2. Hydrodynamic Characteristics of Alpine Meadows under Different Pika Mound Numbers

3.2.1. Average Flow Rate

Flow velocity is a primary parameter to study the hydrodynamic characteristics of slope runoff, which is closely related to slope, vegetation cover, and runoff volume. As shown in

Table 3. Hydrodynamic characteristics at different pika mound numbers.

Slope (°)	The Number of Pika Mounds (pcs)	Runoff Depth (h, mm)	Flow Velocity (v, cm s−1)	Reynolds Number (Re)	Froude Number (Fr)	Darcy–Weisbach Resistance (f)
10°	Revegetated pika mound	3.57 ± 0.74 Aa	2.29 ± 0.27 Cf	57.85 ± 18.45 Cd	0.143 ± 0.009 Be	159.07 ± 31.77 Aa
	1 pika mound	3.25 ± 0.73 Ab	2.61 ± 0.26 Ce	66.08 ± 20.63 Ccd	0.168 ± 0.016 Cd	99.10 ± 13.41 Ab
	2 pika mounds	3.06 ± 0.75 Ac	3.20 ± 0.32 Cd	75.01 ± 25.91 Cbc	0.236 ± 0.024 Bc	71.56 ± 13.54 Ac
	3 pika mounds	2.83 ± 0.76 Ad	3.56 ± 0.33 Cc	80.79 ± 31.53 Cb	0.281 ± 0.037 Bb	51.89 ± 9.71 Ad
	4 pika mounds	2.75 ± 0.74 Ade	3.95 ± 0.34 Cb	76.36 ± 35.42 Cbc	0.289 ± 0.049 Cb	38.86 ± 8.56 Ae
	5 pika mounds	2.85 ± 0.77 Ad	4.26 ± 0.29 Ca	93.15 ± 39.01 Ca	0.322 ± 0.073 Ca	34.89 ± 6.56 Ae
20°	Revegetated pika mound	3.33 ± 0.72 Ba	2.65 ± 0.32 Bf	86.55 ± 25.82 Bd	0.157 ± 0.010 Af	125.80 ± 17.73 Ba
	1 pika mound	3.00 ± 0.72 Bb	3.03 ± 0.35 Be	89.75 ± 27.57 Bd	0.180 ± 0.018 Be	86.71 ± 16.54 Bb
	2 pika mounds	2.82 ± 0.76 Bc	3.56 ± 0.33 Bd	99.19 ± 33.12 Bc	0.271 ± 0.021 Ac	57.98 ± 9.70 Bc
	3 pika mounds	2.66 ± 0.79 Bd	4.05 ± 0.42 Bc	114.18 ± 38.56 Bb	0.253 ± 0.040 Cd	42.56 ± 9.68 Bd
	4 pika mounds	2.45 ± 0.70 Be	4.63 ± 0.37 Bb	105.39 ± 37.84 Bc	0.310 ± 0.045 Bb	29.81 ± 6.33 Be
	5 pika mounds	2.61 ± 0.71 Bd	5.19 ± 0.42 Ba	131.92 ± 42.18 Ba	0.361 ± 0.065 Ba	34.51 ± 5.41 Ade
30°	Revegetated pika mound	2.79 ± 0.61 Ca	2.98 ± 0.40 Af	105.12 ± 28.83 Ad	0.161 ± 0.017 Af	71.47 ± 8.71 Ca
	1 pika mound	2.56 ± 0.63 Cb	3.56 ± 0.38 Ae	116.64 ± 34.92 Acd	0.203 ± 0.016 Ae	50.29 ± 7.68 Cb
	2 pika mounds	2.37 ± 0.66 Cc	4.36 ± 0.39 Ad	134.71 ± 37.62 Abc	0.270 ± 0.026 Ad	30.55 ± 5.33 Cc
	3 pika mounds	2.29 ± 0.73 Ccd	4.57 ± 0.35 Ac	125.36 ± 41.09 Ac	0.293 ± 0.034 Ac	23.66 ± 5.11 Ccd
	4 pika mounds	2.38 ± 0.79 Bc	5.22 ± 0.38 Ab	141.86 ± 45.26 Ab	0.329 ± 0.047 Ab	17.87 ± 4.76 Cd
	5 pika mounds	2.18 ± 0.82 Cd	5.82 ± 0.44 Aa	153.63 ± 51.49 Aa	0.385 ± 0.055 Aa	16.99 ± 4.76 Bd

Note: All tests were repeated thrice, and 12 datasets were taken each time. The values in the table indicate the mean and standard deviation (SD). Different capital and lowercase letters represent significant differences in the individual hydrodynamic parameters at the same level (p < 0.05).

, the mean flow velocity ranged between 2.29–4.26 cm s⁻¹, 2.65–5.19 cm s⁻¹, and 2.98–5.82 cm s⁻¹ for different pika mound patch lands at 10°, 20°, and 30° slopes, respectively, and the mean flow velocity increased linearly with the number of mounds (

Table 4. Regression equation of the number of pika mounds with the hydrodynamic parameters at different slopes.

Hydrodynamic Parameters	The Number of Pika Mounds (pcs)
	10° Slope	20° Slope	30° Slope
v	y = 0.4066x + 2.2952(R2 = 0.9917)	y = 0.514x + 2.5667 (R2 = 0.9963)	y = 0.554x + 3.0333 (R2 = 0.9875)
Re	y = 60.234e0.0826x (R2 = 0.8717)	y = 85.092e0.078x (R2 = 0.8532)	y = 108.2e0.0689x (R2 = 0.8825)
Fr	y = 0.0433x + 0.1663 (R2 = 0.9782)	y = 0.0398x + 0.1559 (R2 = 0.9305)	y = 0.0374x + 0.1474 (R2 = 0.9418)
f	y = 140.93e–0.306x (R2 = 0.9738)	y = 112.19e–0.285x (R2 = 0.9653)	y = 64.547e–0.301x (R2 = 0.9733)

Note: y is the hydrodynamic parameter, x is the number of pika mounds.

). In addition, as the gradient rose from 10° to 30°, the mean flow velocity increased by 30.13%, 36.40%, 36.25%, 28.37%, 32.15%, and 36.62% for the revegetated mound and different mound patch lands, respectively. The statistical analysis showed that the mean flow velocities differed significantly (p < 0.05) for all slope and pika mound number combination patterns. We suggest that an increase in the number of mounds mainly increases the mean flow velocity by weakening the vegetation’s inhibitory and interception effect on surface runoff, and the mean flow velocity tends to be higher on steeper slopes due to inertia forces.

3.2.2. Flow Regime

The criterion value of the Reynolds number (Re = 500) is taken to determine whether the slope flow is laminar or turbulent. [27]. The experiment values of Re at 10°, 20°, and 30° slopes ranged between 57.85–93.15, 86.55–131.92, and 105.12–153.63, respectively (Table 3), and were positively correlated with the pika mound numbers as an exponential function (Table 4). Depending on the basis for the discrimination of open channel flows, we tentatively determined that all slope runoff was laminar. Furthermore, from the results of the statistical analysis, the variability in Re was not significant between different numbers of mounds at the same slope, but the difference in Re between different slopes at the same number of mounds was significant (p < 0.05), concluding that the effect of slope on Re was much more significant than the change in the number of pika mounds.

The Froude number (Fr) represents the flow state of the runoff. As shown in Table 3, the values of Fr were 0.143–0.322, 0.157–0.361, and 0.161–0.385 (none of them exceeded 1) for different slopes, so it was determined that the relationship of Fr with slope and the number of mounds was consistent with that of Re (i.e., it increased with the slope and the number of mounds), and Fr increased as a linear function of the pika mound numbers (Table 4). The analytical results showed that Fr had significant differences (p < 0.05) for different combinations of slope and number of mounds; both were significant factors affecting Fr (two-way ANOVA, p < 0.05).

3.2.3. Flow Resistance

The Darcy–Weisbach resistance coefficient (f) characterizes the magnitude of the resistance to runoff generated by the bedding surface as water flows along the slope. In Table 3, f reduced with an increasing number of mounds and slope, where the range of f was 34.89–159.07 for different pika mound patch lands at a 10° slope, and f decreased to 34.51–125.80 and 16.99–71.47 when the slope was increased to 20° and 30°, respectively. Statistical analysis showed that f showed significant differences (p < 0.05) between different numbers of mounds in the same slope excluding revegetated pika mounds, while f showed significant differences (p < 0.05) between different slopes for the same number of mounds. In addition, f was negatively correlated with the number of mounds by an exponential function (Table 4), indirectly reflecting that the steeper topography and the outbreak of pika mounds in the area are very likely to cause a significant reduction in flow resistance, thus increasing the degree of soil erosion.

3.3. Relationship between Landscape Pattern with Soil Water Erosion Properties under Different Pika Mound Numbers

3.3.1. Analysis of Landscape Pattern Indicators of Pika Mound Patches

We analyzed the spatial distribution pattern of the pika patch landscape by selecting the appropriate landscape index (

Table 5. Landscape pattern indicators of the plateau pika mound patches.

Slope (°)	The Number of Pika Mounds (pcs)	NP	TE	AREA	V	PD	Ni	r	h
10°	Revegetated pika mound	1 e	1.130 f	0.0712 f	0.0033 f	14.042 a	14.894 a	0.180 c	0.063 b
	1 pika mound	1 e	1.413 e	0.1130 e	0.0066 e	8.846 c	9.383 b	0.225 ab	0.080 a
	2 pika mounds	2 d	2.700 d	0.2106 d	0.0118 d	9.495 b	7.121 c	0.215 b	0.078 a
	3 pika mounds	3 c	4.145 c	0.3357 c	0.0193 c	8.936 c	5.472 d	0.220 b	0.081 a
	4 pika mounds	4 b	5.903 b	0.4841 b	0.0296 b	8.263 d	4.382 e	0.235 a	0.082 a
	5 pika mounds	5 a	7.065 a	0.5864 a	0.0345 a	8.527 cd	4.045 e	0.225 ab	0.083 a
20°	Revegetated pika mound	1 e	1.036 f	0.0632 f	0.0027 f	15.821 a	16.780 a	0.165 c	0.061 b
	1 pika mound	1 e	1.382 e	0.1147 e	0.0066 e	8.721 c	9.249 b	0.220 b	0.083 a
	2 pika mounds	2 d	2.889 d	0.2311 d	0.0138 d	8.654 cd	6.491 c	0.230 a	0.080 a
	3 pika mounds	3 c	4.239 c	0.3476 c	0.0204 c	8.631 c	5.285 d	0.225 ab	0.082 a
	4 pika mounds	4 b	5.903 b	0.4664 b	0.0284 b	8.577 d	4.549 e	0.235 a	0.079 a
	5 pika mounds	5 a	7.222 a	0.5561 a	0.0332 a	8.991 b	4.265 e	0.230 a	0.077 a
30°	Revegetated pika mound	1 e	1.005 f	0.0553 f	0.0022 f	18.095 a	19.193 a	0.160 c	0.055 b
	1 pika mound	1 e	1.444 e	0.1156 e	0.0069 e	8.654 c	9.179 b	0.230 ab	0.080 a
	2 pika mounds	2 d	2.700 d	0.2133 d	0.0120 d	9.375 b	7.031 c	0.215 b	0.079 a
	3 pika mounds	3 c	4.427 c	0.3675 c	0.0225 c	8.164 d	4.999 d	0.235 a	0.083 a
	4 pika mounds	4 b	6.029 b	0.4582 b	0.0284 b	8.730 c	4.630 de	0.240 a	0.076 a
	5 pika mounds	5 a	6.908 a	0.5388 a	0.0309 a	9.279 b	4.402 e	0.220 b	0.078 a

Note: NP = the number of pika mound patches; TE = edge length index; AREA = area index; V = volume index; PD = patch density index; Ni = separability index; r = radius of pika mound patches; h = height of pika mound patches. Different lowercase letters in the same column indicate significant differences between the same indices (p < 0.05).

). As the pika mounds increased, the TE, AREA, V, and Ni significantly differed (p < 0.05) at different slopes, while the PD, r, and h did not show significant differences. The analysis also showed that TE and Ni increased exponentially (p < 0.01) and AREA and V increased linearly (p < 0.01) as a function of the pika mound numbers, while PD, r, and h did not show significant functions with the pika mound numbers. In addition, we also found that NP, TE, AREA, and V composed the first principal component (PCA), so their interpretation was as high as 81.32%. There was a highly significant positive correlation between NP and TE, AREA, V, and a highly significant negative correlation with Ni, but no significant correlation with r and h.

3.3.2. Relationship between Landscape Pattern and Soil Erosion under Different Pika Mound Patches

The correlation analysis showed that NP, TE, AREA, and V were extremely significantly correlated with RSE at different slopes, with correlation values ranging from 0.883 to 0.958, while PD, Ni, r, and h were less correlated with the soil erosion rate, with correlation values ranging from 0.547 to 0.731 (

Table 6. Correlation analysis of the soil erosion rate with the landscape pattern indicators of pika mound patches.

Slope (°)	Correlation	NP	AREA	TE	V	PD	Ni	r	h
10°	Grey correlation	0.947	0.932	0.938	0.908	0.626	0.547	0.681	0.685
	Pearson correlation	0.992 **	0.997 **	0.995 **	0.995 **	−0.693	−0.896 *	0.691	0.741
	Sig. (two-tailed)	0	0	0	0	0.127	0.016	0.128	0.092
20°	Grey correlation	0.948	0.958	0.948	0.949	0.661	0.584	0.710	0.702
	Pearson correlation	0.990 **	0.998 **	0.995 **	0.997 **	−0.595	−0.843 *	0.706	0.416
	Sig. (two-tailed)	0	0	0	0	0.213	0.035	0.117	0.413
30°	Grey correlation	0.919	0.908	0.921	0.883	0.667	0.594	0.742	0.731
	Pearson correlation	0.984 **	0.994 **	0.990 **	0.987 **	−0.624	−0.840 *	0.629	0.555
	Sig. (two-tailed)	0	0	0	0	0.186	0.036	0.181	0.253

Note: RSE = soil erosion rate. The definition of landscape pattern index in the table is the same as Table 5. * is a significant correlation (0.01 < p < 0.05), and ** is a highly significant correlation (p < 0.01).

). This table also shows a highly significant positive correlation between RSE and NP, TE, AREA, and V, a significant negative correlation with Ni, and no significant correlation with PD, r, and h at different slopes. The correlation analysis results concluded that the interaction effect between the landscape pattern of pika patches and soil erosion was evident, in which NP, TE, AREA, and V were important factors affecting the RSE, and NP could be used as the main control factor to determine the intensity of soil erosion. This result indirectly shows that the spatial configuration of landscape patterns can be effectively optimized by controlling the number of pika mounds and restoring a certain degree of vegetation growth in the plateau pika mounds outbreak area, suppressing soil erosion.

By establishing the regression equation between the landscape index of the pika mound patches and the soil erosion rate at different slopes (Figure 8), we concluded that the soil erosion rate was positively correlated with the edge length index (TE) and volume index (V) as a linear function (p < 0.01), positively correlated with the area index (AREA) as a power function (p < 0.01), and negatively correlated with the separation index (Ni) as a power function (p < 0.01), while there was no significant regular variation with the density (PD), radius (r), and height (h) of the pika mound patches.

3.3.3. Relationship between Landscape Pattern and Hydrodynamic Characteristics under Different Pika Mound Patches

Based on the results of the correlation analysis at different slopes (

Table 7. Correlation analysis of the hydrodynamic parameters with the landscape pattern indicators of pika mound patches.

Slope (°)		NP	AREA	TE	V	PD	Ni	r	h
10°	v	0.977 **	0.991 **	0.986 **	0.987 **	−0.634	−0.853 *	0.638	0.566
	H	−0.841 *	−0.908 *	−0.884 *	−0.928 **	0.817 *	0.952 **	−0.842 *	−0.758
	Re	0.902 *	0.916 *	0.893 *	0.895 *	−0.658	−0.841 *	0.591	0.650
	Fr	0.943 **	0.972 **	0.955 **	0.970 **	−0.663	−0.873 *	0.656	0.622
	f	−0.851 *	−0.910 *	−0.885 *	−0.920 **	0.854 *	0.979 **	−0.846 *	−0.802
20°	v	0.991 **	0.998 **	0.996 **	0.998 **	−0.592	−0.842 *	0.706	0.408
	H	−0.848 *	−0.899 *	−0.873 *	−0.910 *	0.805	0.954 **	−0.885 *	−0.672
	Re	0.907 *	0.903 *	0.899 *	0.892 *	−0.509	−0.759	0.583	0.368
	Fr	0.954 **	0.959 **	0.961 **	0.960 **	−0.564	−0.819 *	0.699	0.356
	f	−0.854 *	−0.904 *	−0.878 *	−0.914 *	0.829 *	0.979 **	−0.904 *	−0.695
30°	v	0.970 **	0.977 **	0.973 **	0.968 **	−0.643	−0.855 *	0.631	0.576
	H	−0.833 *	−0.877 *	−0.844 *	−0.869 *	0.796	0.934 **	−0.730	−0.792
	Re	0.907 *	0.902 *	0.903 *	0.884 *	−0.626	−0.823 *	0.582	0.557
	Fr	0.958 **	0.957 **	0.958 **	0.944 **	−0.617	−0.831 *	0.599	0.543
	f	−0.862 *	−0.915 *	−0.892 *	−0.923 **	0.810	0.960 **	−0.799	−0.76

Note: The hydrodynamic parameters in this table are as defined in Table 3, and the landscape pattern indices in this table are as defined in Table 5. * is a significant correlation (0.01 < p < 0.05), and ** is a highly significant correlation (p < 0.01).

), there were highly significant positive correlations between NP, TE, AREA, and V with v and Fr, significant negative correlations with H and f, and significant positive correlations with Re. Ni showed a highly significant positive correlation with H and f, and a significant negative correlation with v, Re, and Fr. No significant correlations were found between PD, r, and h and the hydrodynamic parameters. It shows that there is indeed an interaction between the landscape pattern and hydrodynamic characteristics of pika mound patches, in which NP, TE, AREA, V, and Ni affect the connectivity and mobility of the slope surface flow by changing the landscape scale effect, which leads to significant variability in the hydrodynamic characteristics, and they can all be significant factors affecting the hydrodynamic characteristics of slope runoff.

3.3.4. Relationship between Soil Erosion and Hydrodynamic Characteristics under the Different Pika Mound Patches

Based on the results of the correlation analysis (

Table 8. The correlation analysis between soil erosion and the hydrodynamic parameters at different slopes.

Slope (°)	Correlation	H	v	Re	Fr	f
10°	Grey correlation	0.672	0.798	0.760	0.817	0.571
	Pearson correlation	−0.887 *	0.991 **	0.919 **	0.957 **	−0.905 *
	Sig. (two-tailed)	0.018	0	0.011	0.003	0.013
20°	Grey correlation	0.690	0.836	0.793	0.834	0.587
	Pearson correlation	−0.907 *	0.998 **	0.919 **	0.953 **	−0.914 *
	Sig. (two-tailed)	0.013	0	0.012	0.004	0.011
30°	Grey correlation	0.674	0.786	0.729	0.817	0.569
	Pearson correlation	−0.910 *	0.993 **	0.938 **	0.994 **	−0.933 **
	Sig. (two-tailed)	0.012	0	0.006	0.001	0.007

Note: The abbreviations representing the hydrodynamic parameters are the same as in Table 7. * is a significant correlation (0.01 < p < 0.05), and ** is a highly significant correlation (p < 0.01).

), v, Re, and Fr were most significantly associated with the RSE at different slopes, with correlation values around 0.8; whereas H and f were relatively less associated with the RSE, and the correlation values ranged from 0.569 to 0.690. In addition, a highly significant positive correlation was found between v, Re, and Fr with the RSE, while there was a significant negative correlation between H and f with the RSE. We initially judged that the average flow rate could be better to describe the soil erosion process.

4. Discussion

4.1. Comparison with Other Similar Studies

The power source of soil water erosion is mainly rainfall and slope runoff, and the process is influenced by many factors such as the vegetation cover, slope, and soil properties [11,32,33,49]. In order to further characterize the soil water erosion of the pika mounds, the results of the present study were compared with other similar reports. We found that the amount of soil loss per unit time produced showed an increase and then a decrease with rainfall duration (Figure 5), consistent with the findings of Li et al. [12] on plateau zokor mounds. However, from the information on localized changes in soil loss, the peak points of soil loss in the two studies showed apparent differences, which were considered likely to be related to the rainfall intensity and number of rodent mounds. In addition, our research concluded that the soil erosion rate of the vegetation restoration pika mound was reduced by 1.03 to 6.74 times compared with the different numbers of mounds. However, the soil erosion rate increased with the number of mounds and slope (Figure 6). The analyzed reason mainly lies in the fact that the increase in the number of mounds also reduces the vegetation cover, weakening the inhibition and interception function of meadow vegetation [50] and ultimately exacerbating soil erosion under the superimposed effect with slope. Our results agreed with the results of Zhao et al. [26] and Chen et al. [39] in the degraded vegetation area, and we believe that the vegetation restoration growth in rodent-infested degraded areas can effectively prevent soil erosion. At the same time, it is essential to reasonably control the activity range of plateau pika and the density of pika mounds. Marzen et al. [51] suggested that the influence of rainfall on slope erosion is largely the result of the potential erosive force of rain splash and runoff on loosening topsoil. We also observed that the rain splash would lead to the rain traces and stacked tile texture on the surface of the pika mounds. Rain splash is also an essential element in the study of soil erosion mechanisms, so studying the effect of rain splash on the soil erosion of pika mounds with different rainfall intensities is also a scientific problem that needs to be investigated. Hydrodynamic characteristics are an essential area to reflect the soil erosion process; Zhang et al. [37] and Yuan et al. [52] found that the increase in flow velocity contributed to the runoff connectivity and soil erodibility, which led to an increase in the surface runoff capacity and soil erosion, and this indirectly confirmed our viewpoint. We also found that runoff velocity was the hydrodynamic parameter most closely related to soil erosion through the correlation analysis (Table 8). We found that the runoff resistance decreased with the increase in mound numbers (Table 3), mainly because the increase in mounds significantly reduced the vegetation cover, which in turn led to a decline in runoff resistance, which agrees with the results of Zhao et al. [26] and Sun et al. [49] Furthermore, although increasing mound numbers may disrupt the runoff connectivity and increase landscape roughness, it also provides runoff channels for overland flow, and with an increase in the average slope, the runoff resistance tends to decrease. In addition, we observed that the height of the plateau pika mounds settled by 1.36 cm h⁻¹ during the experiment, but Li et al. [12] found that the height of zokor mounds decreased at a rate of 1.8 cm h⁻¹ in a previous hydraulic erosion test on plateau zokor mounds. We speculate that this might be related to the soil properties, slope, and rainfall devices on the one hand, and on the other hand, it was mainly affected by the shape and size of the two types of mounds. Together with the experimental results and the actual landscape scale, the erosion of the pika mound soil in the degraded meadow area was not only affected by the number of mounds, slope, and vegetation cover, but was also closely related to the landscape pattern of the patches, topographic, and micro-geomorphic features as well as the adaptability of vegetation replacement, etc. The erosion mechanism and ecological succession of the pika mound patches will be scientific issues that need to be deeply investigated.

4.2. Response Mechanisms of Plateau Pika Activities to Soil Erosion and Meadow Degradation

Plateau rodent activity significantly drives soil degradation and meadow degradation in the source zone of the Yellow River [53,54]. Degraded meadow areas induced by plateau pika activities can provide a large number of material sources for soil erosion, while repeated and continuous soil erosion can, in turn, exacerbate the degree of meadow degradation [55]. During the field investigation, we found that in areas with high activity of plateau pika, their digging behavior not only destroyed the vegetation community and soil structure of native meadows [56], but also caused the meadow vegetation to rot and wither due to the accumulation and burial of loose pika mounds after turning out the shallow soil, and eventually formed pika mound patch lands that were significantly different from the native meadows [11,57]. The degraded bald patches are also subject to solar radiation, toxic grass invasion, and climate change, which constantly cause an increase in the number and area of patches, which leads to an expansion in alpine meadow degradation. It has been shown that meadow degradation and soil erosion are intrinsically related [58,59], in particular, the mound patches accumulated by the activities of plateau pika are prone to eroding under the action of wind, water, and freeze–thaw due to their loose structure, fine particles, and particular shape. In our study, degraded patches of land with different numbers of pika mounds responded positively to the soil water erosion conditions. The total soil loss of the degraded patch lands with one, two, three, four, and five pika mounds were significantly different (p < 0.05). Their soil erosion rates were significantly higher than the revegetated pika mounds (p < 0.01). Furthermore, soil erosion is the process by which fine soil particles are separated and transported and includes the loss of soil nutrients [60]. The succession of pika mound patches, due to the lack of nutrients in the vegetation, leads to the expansion of pika mound patches in all directions and even causes patch-to-patch connectivity, which eventually forms a bidirectional vicious cycle between soil erosion and meadow degradation, continuously increasing the level of meadow degradation [12]. Soil erosion and meadow degradation are extraordinarily complex and closely interrelated processes, although they are closely related to the meteorological and hydrological conditions, geographic environment, and the adaptation of patch micro-geomorphic development and vegetation replacement. However, due to the random nature of plateau pika activities and the particular soil structure of the pika mound, it is also more sensitive in influencing meadow ecological processes and soil erosion, so to a certain extent, the frequent activities of plateau pika are the primary initiation of meadow degradation and soil degradation in the source zone of the Yellow River.

4.3. Properties of Alpine Pika Mound Patches and Their Effects on Ecological Processes in Alpine Meadows

Alpine meadows in the source zone of the Yellow River are an important protective barrier for the entire Yellow River Basin and the environmental eco-systems of the Tibetan Plateau [61]. As a basic configuration of the vegetation landscape pattern in this region, the different quantitative distribution patterns of pika mound patches will change and reshape the landscape pattern of the meadow area, impacting the stability and sustainability of the meadow ecological processes. In this study, we simulated fresh uncrusted pika mounds of the same size, weight, and shape but with different number classes, aiming to clarify the effects of different pika mound numbers on soil water erosion and its hydrodynamic characteristics. Our results show that plateau pika activities intensify meadow degradation and increase soil erodibility by reducing vegetation cover and runoff resistance. However, the timing of mound formation, number of mounds, and the shape and size of mounds vary among different periods of plateau pika behaviors, so this study also quantified and compared the mounds with the revegetated mounds in the following years. We found significant differences (p < 0.05) in the landscape index, soil erosion and sand production, and hydrodynamic characteristics of the revegetated pika mounds compared to other pika mound patches. This reason could be related to the spatial scale instead of the temporal scale (i.e., at the early stage of plateau pika mound patch succession, the soil particles of the pika mound patches were dispersed and loose, and the landscape was severely fragmented, so that the edge length index, area index, and volume index were more extensive). The destructive mechanical effects of plateau pika on meadows involve not only meadow degradation but also soil degradation (i.e., a massive reversal of the soil properties of pika mounds before and after [62]), leading to more intense soil and nutrient loss under the action of external erosion, which has a direct negative impact on meadow ecological processes. During the field investigation, it was also found that the succession of plateau pika mounds in different years was a spatially dynamic process, with the radius, height, area, and volume of the following year’s mound soil decreasing to different degrees with time succession, but the soil erosion resistance increased significantly. The main reason for this is that although the mounding behavior of plateau pika destroys meadow vegetation and soil, the microtopographic environment formed by the pika mounds can renew the subsurface organic matter composition and create more ecological niches for the subsequent invasion and colonization of other dominant species. To a certain extent, it has contributed to the improvement in the properties of pika mound patches and their benign cycle with the ecological processes of the meadow [63]. In summary, the presence of pika mound patches has both advantages and disadvantages for meadow ecological processes, but in the context of local climatic and environmental characteristics and the current status of meadow degradation, it is still necessary to artificially control the population density and mound numbers of plateau pika in the outbreak area. In addition, we concluded that aside from the number of pika mounds, the slope and vegetation cover also are essential factors influencing the soil water erosion processes of the pika mounds. We suggest that a combination of different vegetation cover and slope patterns can be integrated into the restoration of meadow vegetation to enhance the soil and water conservation function of degraded bald patches, which is fundamentally important in promoting sustainable management and the utilization of meadow resources.

5. Conclusions

We aimed to clarify the effects of different pika mound numbers on the soil erosion and hydrodynamic characteristics by conducting simulated rainfall experiments. The results showed that the amount of soil loss increased and then decreased with rainfall time and that 15–20 min of rainfall was a sensitive period for soil loss in different pika mound patch lands. The degradation of meadows due to the activities of plateau pika is an essential factor influencing soil erosion, and the soil erosion rate is positively correlated with both the pika mound number and slope. The mean flow velocity can better describe the soil erosion process, and the magnitude of its value increases with the number of pika mounds and slope. The Reynolds number ranged from 57.85 to 153.63 (Re < 500), and it was preliminarily determined that all slope runoff was laminar. The Froude number increased with the pika mound numbers and slope, and they were both significant factors (p < 0.05). The resistance coefficient instead decreased with the pika mound numbers and slope, and the inhibitory effect of vegetation on soil erosion was likely to be limited when the number of mounds reached a certain level. Based on the results of the grey correlation method and Pearson correlation analysis, changes in the number of pika mounds caused variability in the slope surface flow hydrodynamic characteristics and soil erosion conditions by changing the landscape scale effect, where the number of mound patches (NP), edge length index (TE), area index (AREA), and volume index (V) were the key influencing factors on the soil erosion rate and each hydrodynamic parameter.

We can conclude that outbreaks of plateau pika activity exacerbate meadow degradation, and that the development of pika mound patches leads to severe landscape fragmentation through altered landscape-scale effects, and ultimately causes decreases in the vegetation cover and runoff resistance and increases in the soil erodibility. To effectively restore the meadow ecosystem, we suggest that reasonable control of plateau pika activities should be combined with a reduction in the density of pika mounds to realize the dual purpose of soil erosion control.