Short-Term Artificial Revegetation with Herbaceous Species Can Prevent Soil Degradation in a Black Soil Erosion Gully of Northeast China

Abstract

Understanding the effects of short-term artificial revegetation on preventing soil degradation in erosion gullies of black soil areas is essential to choosing the most suitable species of vegetation for controlling the development of erosion gullies. A field experiment with short-term artificial revegetation with herbaceous species (Medicago sativa L., Glycyrrhiza pallidiflora Maxim., Elytrigia repens (L.) Desv. ex Nevski, Rheum palmatum L., Asparagus officinalis L., Trifolium repens L., Bromus inermis Leyss., Elymus dahuricus Turcz.) and a runoff scouring test were conducted in a typical erosion gully in a black soil area. Soil erosion, physicochemical characteristics, and shoot/root characteristics were measured to evaluate the effects of short-term artificial revegetation. Short-term artificial revegetation significantly decreased (p < 0.05) sediment yield by 91.1% ± 7.2% compared with that of bare soil. Soil total nitrogen (TN), total potassium (TP), available phosphorus (AP), cation exchange capacity (CEC), water-stable aggregates > 0.25 mm (WR_0.25), and aggregate mean weight diameter (MWD) and mean geometric diameter (GWD) were significantly correlated with vegetated treatments, indicating they were factors sensitive to short-term artificial revegetation. Except for total potassium (TK), the other soil characteristics decreased in vegetated treatments. In addition to increasing TK, vegetated treatments also increased soil available nitrogen (AN)/TN ratios in the short term. The overall effects of different herbaceous species on soil and water conservation, soil quality, and vegetation growth were evaluated, and Trifolium repens L. is the most suitable for preventing soil degradation in an erosion gully. The results of this study will provide a reference for the restoration and protection of the ecological environment in black soil areas with gully erosion.

1. Introduction

Soil erosion is a global phenomenon that is becoming increasingly important, especially in the context of climate change. Gully erosion occurs across a wide range of environments as a sign of land degradation. Agricultural fields are damaged by gully erosion, sediments are lost, and surface soil quality is compromised [1,2]. In Northeast China, the mainly black and fertile soils are essential for grain production, and therefore play a crucial role in national food security. However, as of 2020, 666.7 thousand gullies were reported in the region (Songliao River Water Resources Commission of the Ministry of Water Resources, http://www.slwr.gov.cn/, accessed on 1 May 2024). As erosion gullies expand annually, they encroach on arable land by 7.39 km² y⁻¹ [3]. The annual loss in grain yields due to gully erosion is approximately 36.2 × 10⁸ kg, which represents 10% of the country’s total grain supply (National Earth System Science Data Center, http://northeast.geodata.cn, accessed on 1 May 2024).

Combating the gully erosion is very necessary to prevent land degradation in black soil. Erosion gullies develop by headcuts and expanding gully slopes. Trees and shrubs are commonly used types of vegetation to prevent gully wall expansion and effectively control gully erosion [4]. Although the mechanisms by which vegetation conserves water and soil are well understood [5,6], planting trees or shrubs to control gully erosion has some drawbacks. First, the trees planted may not be suitable for the environment. It is generally true that native forest trees retain soil and water better than shrubs and grasslands, but when planted, the effects can be very different [7]. Second, the effects of trees and shrubs are slow to develop because of the long growing periods. Third, in contrast to nonintercepted drops, those intercepted by tall trees without an understory can be larger and have greater kinetic energy, which can lead to soil crusting, surface runoff generation, and gully formation [8,9]. Last, shrubland is hydraulically more efficient than grassland, leading to higher runoff and erosion rates [10].

Herbaceous revegetation can effectively reduce soil and water loss [11,12], and it conserves more soil and water compared with trees and shrubs on gully slopes [7,13]. On gully slopes, herbaceous canopies, roots, and litter all play a role in controlling soil erosion [14]. In addition to intercepting and diverting rainwater, herbaceous vegetation canopies provide physical protection to soils by reducing the impact of raindrops and reducing “splash” effects [15]. As with the canopy, litter can serve as a protective barrier, while roots can hold soil in place, trap sediment, and add organic substances to soil, thus improving soil structure [16]. The infiltrability and structural stability of soil are improved by plant roots and reduce gully erosion [4]. Roots play a major role in soil resistance to concentrated flow erosion by virtue of the characteristics of their fibrils (fibrils < 1.0 mm in diameter), which are distributed densely at a depth of 0–30 cm [17]. It may be helpful to use herbaceous revegetation studies to understand the relationships between herbaceous species and soil characteristics, and to select the species that are most suitable for restoration [18,19]. Several soil quality characteristics are commonly evaluated under different vegetation treatments, including total phosphorus, bulk density, aggregate stability, saturated hydraulic conductivity, and macroporosity [20,21]. It is crucial to fully understand how to restore degraded soils, determining the responses of soil nutrients and structure to revegetation.

Compared with trees and shrubs, herbaceous vegetation has the advantages of low cost, efficiency, and multiple roles. According to Wang et al. [22], grassland has important effects on erosion control. In addition, the cost of controlling gully erosion by herbaceous vegetation is expected to be only a third of that by trees or shrubs. Last, planted forage grass has a short growing period, and the effects are apparent in one year. Yan et al. [23] also found that herbaceous vegetation on gully slopes decreased runoff and soil loss by 19–30% and 78–97%, respectively. However, few managers use herbaceous vegetation to measure whether it can control the active degrading soil quality in gullies. Whether short-term artificial herbaceous revegetation is suitable for controlling erosion gullies and preventing soil degradation on gully slopes in black soil regions remains uncertain. In addition, little is known about which herbaceous species are suitable for preventing soil degradation in gully slopes.

To examine the effects of different species of short-term herbaceous vegetation on the soil quality in gully slopes and then select the species best suited for controlling soil degradation in an erosion gully of a black soil region, we conducted a field artificial revegetation experiment and a runoff scouring experiment with eight species of herbs. Study objectives included determining (1) the role of herbaceous vegetation in preventing soil and water loss on gully slopes; (2) the effects of short-term artificial herbaceous vegetation restoration on soil physicochemical properties; and (3) the suitable herbaceous species to combat soil degradation in an erosion gully. The hypotheses are as follows: (1) herbaceous revegetation can reduce soil erosion on gully slopes; (2) short-term artificial herbaceous revegetation will improve the soil quality.

2. Materials and Methods

2.1. Study Area

The erosion gully was in Zhaoguojun Village of Baiquan County in Heilongjiang Province (47°23′45.23″–47°23′54.18″ N, 126°16′21.03″–126°16′25.52″ E; Figure 1), China. The length of the erosion gully was 789 m, and the area of the gully was 1.1 ha. The mean depth of the erosion gully was 5 m. The longitudinal gradient of the gully was 61.3‰. The average annual temperature in the region is 1.5 °C, and the average annual rainfall is 530 mm. The soil is a silt loam soil and classified as Mollisol (also called black soil), with a soil depth of 30–50 cm. Soils and their parent materials are generally acidic to alkaline, with varying textures ranging from loam to clay loam, with most sola consisting mainly of clay loam. The land use types are forests and sloped croplands (3° to 5°).

2.2. Experimental Setup

In May 2018, eight species of perennial herbaceous plants were planted on the gully slope (Table S1), and bare soil was used for the control (BARE). Rheum palmatum L. (RH) is a perennial herb in the Caryophyllales, Polygonaceae, and Rheum L. Asparagus officinalis L. (AS) is a perennial herb in the Asparagales, Liliaceae, and Asparagus L. Elymus dahuricus Turcz. (ET) is a perennial herb in the Gramineae, and Elymus Linn. Medicago sativa L. (ME) is a perennial herb in the Rosaceae, Leguminosae, and Medicago L. Glycyrrhiza pallidiflora Maxim. (GL) is a perennial herb in the Rosaceae, Leguminosae, and Glycyrrhiza L. Elytrigia repens (L.) Desv.ex Nevski. (EN) is a perennial herb in the Gramineae, and Elytrigia Desv. Trifolium repens L. (TR) is a perennial herb in the Fabales Bromhead, Leguminosae, and Trifolium L. Bromus inermis Leyss. (BR) is a perennial herb in the Gramineae, and Bromus L. Fabaceae can fix nitrogen (N), a potentially important contribution in degraded soils. The Fabaceae and Gramineae plants were achieved by strip sowing with a 30 cm row spacing and 2–3 cm depth. AS was also sown using strip sowing with a row spacing of 30 cm and a sowing depth of 4–6 cm. RH was transplanted with a 30 cm row spacing with 1–2 plants per hole, growing to 15 cm, and a plant spacing of 50 cm. The germination rate and clarity were obtained (GB 6141 seed quality grading of Leguminosae and GB 6142 seed quality grading of Gramineae). Six replicate photos of each plot were taken at a height of 1.5 m, and PCOVER software (Version 3.00) was used to obtain the vegetation coverage (VC, %).

2.3. Runoff Scouring Experiment

Sixteen months after sowing, we conducted a field runoff scouring experiment, in September 2019. Three months before the runoff scouring experiment, each species of herbaceous plant was surrounded in three replicate runoff plots (2 m × 1 m). To isolate treatments, a steel sheet was embedded to a soil depth of 0.1 m around each runoff plot. Plastic pipes, valves, a buffer tank, and a water supply tank were used to supply water to the runoff plot [23]. An attachment was designed to connect the buffer tank to the water supply tank at its bottom. Until the runoff plot was filled with water, water from the water supply tank spilled into the buffer tank, ensuring uniform water flow that was stable and nonpressurized. In order to ensure adequate water supplementation during the experiment, two trucks supplied filtered water. The flow discharge was controlled using an adjusting valve and measured three times before and after each experiment. First, the soil of a runoff plot was saturated with water with a watering can to prevent the water from infiltrating into it. According to the maximum rainfall intensity (75 mm h⁻¹) in the area (China Meteorological Data Service Centre, https://data.cma.cn/, accessed on 1 May 2024), the flow discharge was set at 0.24 L s⁻¹, which was controlled by a switch. Runoff velocity was measured with a staining method. Each runoff scouring experiment was conducted three times. After the start and at the end of the experiment, three 5 s runoff samples were collected. Sediment yield was collected for 5 s every 15 s. As a measure of sediment concentration, samples were dried at 105 °C until a constant weight was reached (SC, g cm⁻³).

Surface runoff (SR, mm), sediment yield (SY, g m⁻²), and soil erosion rate (ER, g m⁻² s⁻¹) during the observation time (t, s) were calculated using Equations (1)–(4), respectively, as follows:

SR = V/LW × 10³

(1)

SY = M/LW

(2)

M = V × SC × 1000

(3)

ER = SY/t

(4)

where V is the runoff volume (m³); L and W are the length (m) and width (m) of the plot, respectively; and M is the amount of soil loss (g).

The mean flow velocity (u, m s⁻¹) was calculated as the product between surface flow velocity (u_s, m s⁻¹) and the correction factor (α = 0.67). We measured the u_s every 2 min at a marked 1 m distance with five replications using the KMnO₄ coloration technique. An accurate 1.0 mm steel rule was used to measure flow width (w, m) at three runoff observation sections. The flow hydraulic parameter unit flow discharge (q, m² s⁻¹), flow depth (h, m), flow shear stress (τ, Pa), stream power (ω, W m⁻²), Reynolds number (Re), and Froude number (Fr) were calculated using Equations (5)–(10), respectively, as follows:

q = Q/w

(5)

h = q/u

(6)

τ = ρgRJ

(7)

ω = uτ

(8)

Re = uR/υ₀

(9)

$$\mathrm{Fr}=\mathrm{u}/\sqrt{\mathrm{g}\mathrm{h}}$$

(10)

where Q is flow discharge (m³ s⁻¹); ρ is the water density (kg m⁻³); g is the constant of gravity (m s⁻²); R is the hydraulic radius (m), which is equal to the mean flow depth; J is the hydraulic gradient, which is equal to the sine of the soil surface slope angle; and υ₀ is the kinematical viscosity (m² s⁻¹).

2.4. Soil Sampling and Laboratory Analysis

Before Frost’s Descent in 2018, a 1 m long sample section was randomly selected within the plant row of each plot. Three replications were carried out. We recorded the number of plants in each sample area. In the following year, the number of surviving plants was investigated after the plants in the sample area had resumed growth to calculate the resume growth rate of the plant. The resume growth rate = N (number of surviving plants before winter)/N1 (number of surviving plants after winter) × 100%.

Soil samples were collected outside the runoff plot with shovels before runoff scouring experiments at depths of 0–5 and 5–10 cm. Three samples were collected from the top to bottom of the gully slope and then composited as one soil sample. The sampling process was repeated three times per vegetated treatment and the bare soil. Drying was allowed to take place naturally by storing soil samples in well-ventilated conditions. Dried soils were sieved through a 200 μm mesh. Sieved soils were analyzed for physical and chemical properties.

The aggregate size distribution was measured by a slow-wetting method, with 50 g of air-dried soil that passed through a 10 mm sieve placed on the top of three nested sieves (2.0, 0.25, and 0.053 mm mesh) and submerged in tap water for 10 min. For 2 min, the wet sieve apparatus (DIK-2001, Daiki Rika Kogyo Co., Ltd., Saitama, Japan) was oscillated 3 cm up and down at a rate of 30 cycles per minute. Using the resistant soil materials on top of each sieve and the remaining unstable aggregates (<0.25 mm), all materials were oven-dried at 50 °C for 48 h. The wet aggregate stability (WAS) of the soil aggregates that remained above the sieves was calculated by weighing them after oven-drying:

WAS = (weight of aggregates on sieves/50 g) × 100%

(11)

The bulk density (BD) of the undisturbed soil was determined using a cutting ring (volume = 100 cm³) and oven-dried at 105 °C to a constant weight. By analyzing soil particle size distribution with a Mastersizer 2000 laser sizer (Malvern Instruments Ltd., Malvern, UK), soils were divided into sand, silt, and clay fractions. Soil BD, field capacity (FC), water content (SWC), and porosity (P) were measured separately by steel ring (volume = 100 cm³) and gravimetric methods. Soil aeration porosity was calculated as the difference between P and FC.

Soil organic carbon (SOC) and total nitrogen (TN) contents were determined by a Euro single elemental analyzer (Euro Vector, Milan, Italy). We measured soil total potassium (TK) using flame photometry, and soil nitrogen (AN) using Kang Hui dishes [24]. Soil available phosphorus (AP) was measured using a molybdenum–antimony resistance colorimetric method [25]. The soil cation exchange capacity (CEC) was determined by the CH₃COONH₄ exchange method, while the soil pH level was measured using a pH meter (S20P-K; Mettler-Toledo, Greifensee, Switzerland) in suspension with a 1:2.5 soil-to-water ratio.

After runoff scouring experiments, plants were clipped above the soil surface, sorted by species, and oven-dried for 48 h at 65 °C to a constant mass, then weighed to calculate shoot dry weight (SDW, kg ha⁻¹). Topsoil root samples were collected by digging a randomly selected soil block of 20 cm (width) × 20 cm (length) × 30 cm (depth). After immersing a soil block in water for 12 h, intact and clean roots were collected by washing three times in water. Specific root length (SRL, m g⁻¹), root/shoot ratio (RSR), root length density (RLD, cm cm⁻³), root surface area density (RSD, cm² cm⁻³), and root volume (RV, cm³) were analyzed by a scanner (Expression 1640XL, Epson, Nagano-ken, Japan) and the WinRHIZO 2004a root analysis system (Regent Instrument Inc., Quebec, QC, Canada). RLD and RSD are the ratios of root length and surface area, respectively, and soil volume. Root biomass was determined by oven-drying at 65 °C for 48 h to a constant mass.

2.5. Statistical Analyses

The soil properties of eight species were compared with analysis of variance (ANOVA; t-test) and least significant difference (LSD), both at significance level p < 0.05. The coefficients of correlation between soil clay content, root characteristics, and aggregate characteristics were determined using Pearson correlation analysis with p < 0.05. Analysis of the data and data processing were performed using SPSS 22 (SPSS Inc., Chicago, IL, USA) and Origin 2021 (Origin Lab Corp., Northampton, MA, USA). ArcGIS 10.6 (Esri Inc., Redlands, CA, USA) was used to generate the study area map. Principal component analysis was performed with SPSSAU (https://spssau.com/, accessed on 1 May 2024) to obtain scores to evaluate the effects of short-term artificial revegetation.

3. Results

3.1. Adaptability of Artificial Revegetation

Differences in the timing of growth stages demonstrated the adaptability of artificial revegetation on the gully slope (Tables S1 and S2). Resume growth is an important indictor in evaluating vegetation adaptability. BR and RH were the first species to resume growth on 3 May. Most of the species returned to green during the first 10 d period of May. The last species to resume growth was GL on June 10. The resume growth ratio varied from 60% to 92% and was greater than 90% for ME, GL, and TR.

There were large differences in shoot and root characteristics among treatments with different herbaceous species. GL had the largest shoot dry weight (34,817 ± 12,106 kg ha⁻¹, p < 0.05), whereas AS had the smallest (425 ± 150 kg ha⁻¹, p > 0.05, Figure 2A). EN had the largest specific root length (1.094 ± 0.240 m g⁻¹, p > 0.05), whereas RH had the smallest (0.004 ± 0.004 m g⁻¹, p > 0.05, Figure 2B). RH had the largest root/shoot ratio (2.93 ± 0.76, p < 0.05), whereas EN had the smallest (0.36 ± 0.06, p > 0.05, Figure 2C). As shown in Figure 2D, BR had the largest root length density (0.94 ± 0.10 cm cm⁻³, p < 0.05), whereas GL had the smallest (0.03 ± 0.01 cm cm⁻³, p > 0.05). BR had the largest root surface area density (0.06 ± 0.00 cm² cm⁻³, p > 0.05), whereas GL had the smallest (0.01 ± 0.00 cm² cm⁻³, p > 0.05, Figure 2E). As shown in Figure 2F, RH had the largest root volume (135.2 ± 68.1 cm³, p < 0.05), whereas ET had the smallest (8.1 ± 0.2 cm³, p > 0.05).

3.2. Factors Sensitive to Short-Term Artificial Revegetation

According to the correlation matrix shown in Figure 3, the factors SY, ER, TN, TK, AP, CEC, water-stable aggregates (>0.25 mm) (WR_0.25), aggregate mean weight diameter (MWD), and aggregate geometric mean diameter (GWD) were significantly correlated with vegetated treatments (p < 0.05). Thus, those factors were sensitive to short-term artificial revegetation. The factors are analyzed in detail in the following paragraphs.

3.3. Effects of Artificial Revegetation on Reducing Soil and Water Loss

An important function of herbaceous vegetation in controlling gully wall expansion is to reduce surface runoff and sediment loss. Compared with bare soil without vegetation coverage (BARE), vegetated treatments reduced average sediment yield by 91.1% ± 7.2% during runoff scouring (Figure 4). Among the herbaceous species, sediment yield was most reduced with TR (p > 0.05). By contrast, the highest sediment yield was with AS. Vegetation coverage of RH, AS, and ET was lower than that of ME, GL, EN, TR, and BR. Surface runoff was not significantly different between vegetated treatments and BARE (p > 0.05). Among the herbaceous species (Table S3), flow hydraulic parameters varied with species. AS had the largest q and RH had the smallest q. AS had the largest Re and RH had the smallest Re. BR had the largest Fr and GL had the smallest Fr.

3.4. Responses of Soil Physicochemical Characteristics to Artificial Revegetation

Soil physical characteristics varied with herbaceous species (Figure 5 and

Table 1. Soil physical characteristics with different herbaceous species.

Herbaceous Species	Soil Bulk Density (g cm−3)	Soil Water Content (V/V, %)	Field Capacity (V/V, %)	Soil Porosity (%)	Soil Aeration Porosity (%)	WR0.25 (%)	MWD (mm)	GWD (mm)	Soil Texture
									Clay (%)	Silt (%)	Sand (%)
RH	1.34 ± 0.05	33.17 ± 1.40	38.03 ± 1.80	49.32 ± 1.80	11.30 ± 1.33	79.39 ± 1.78	7.38 ± 1.01	0.79 ± 0.16	0.2 ± 0.0	8.0 ± 0.1	91.8 ± 0.1
AS	1.34 ± 0.01	32.16 ± 0.56	37.15 ± 0.36	49.25 ± 0.30	12.10 ± 0.66	82.65 ± 4.41	7.87 ± 2.15	1.27 ± 0.61	0.2 ± 0.0	10.0 ± 0.1	89.8 ± 0.1
ET	1.63 ± 0.01	18.88 ± 4.88	21.40 ± 5.95	38.39 ± 0.41	16.99 ± 5.53	91.03 ± 2.47	5.97 ± 0.08	1.46 ± 0.19	0.2 ± 0.0	2.0 ± 0.0	97.8 ± 0.0
ME	1.41 ± 0.09	36.50 ± 0.59	38.25 ± 0.27	46.78 ± 3.37	8.54 ± 3.09	90.21 ± 1.93	6.75 ± 2.91	2.5 ± 1.97	2.2 ± 0.1	5.8 ± 0.2	92.0 ± 0.2
GL	1.37 ± 0.02	29.93 ± 0.23	33.47 ± 0.30	48.40 ± 0.92	14.94 ± 1.80	88.71 ± 1.76	7.12 ± 1.35	1.7 ± 0.32	2.2 ± 0.2	12.0 ± 0.2	85.8 ± 0.0
EN	1.40 ± 0.03	31.87 ± 0.64	35.87 ± 0.45	47.28 ± 1.14	11.42 ± 1.59	82.38 ± 1.76	7.97 ± 1.08	1.16 ± 0.11	2.2 ± 0.0	14.1 ± 0.2	83.7 ± 0.2
TR	1.30 ± 0.08	33.69 ± 0.90	36.86 ± 0.98	51.08 ± 3.03	14.22 ± 2.06	87.11 ± 0.78	9.75 ± 4.46	3.53 ± 3.81	2.2 ± 0.0	14.0 ± 0.0	83.8 ± 0.0
BR	1.57 ± 0.02	29.38 ± 0.54	31.81 ± 0.46	40.59 ± 0.60	9.28 ± 1.06	87.81 ± 1.72	12.34 ± 2.81	4.09 ± 3.59	6.2 ± 0.0	14.0 ± 0.0	79.8 ± 0.0
BARE	1.49 ± 0.06	36.68 ± 2.43	36.58 ± 1.51	43.88 ± 2.32	7.31 ± 2.16	92.21 ± 2.7	12.2 ± 4.09	7.89 ± 6.75	2.2 ± 0.0	10.0 ± 0.0	87.8 ± 0.0

Note: soil water content and field capacity using volumetric water content (V/V). RH: Rheum palmatum L., AS: Asparagus officinalis L., ET: Elymus dahuricus Turcz. ME: Medicago sativa L., GL: Glycyrrhiza pallidiflora Maxim., EN: Elytrigia repens (L.) Desv.ex Nevski, TR: Trifolium repens L., BR: Bromus inermis Leyss. BARE: no vegetation coverage.

). Soil aggregate fractions varied with species of herbaceous vegetation (Figure 5). Four fractions of soil aggregates were evaluated: macroaggregates (>2.0 mm), small macroaggregates (0.25–2.0 mm), microaggregates (0.053–0.25 mm), and silt plus clay (<0.053 mm). In both 0–5 cm (Figure 5A) and 5–10 cm (Figure 5B) soil layers, the silt plus clay fraction in vegetated plots was larger than that in BARE. To analyze the relations, t-tests were used to compare the differences in soil aggregate fractions between BARE and vegetated treatments. In the 0–5 cm layer, the silt plus clay fraction in BARE (1.63% ± 1.86%) was lower than that in vegetated treatments (7.51% ± 3.34%, p = 0.020). In the 5–10 cm layer, the silt plus clay fraction in BARE (4.43% ± 3.82%) was also lower than that in vegetated treatments (5.76% ± 2.13%, p = 0.471). In the 0–5 cm layer, WR_0.25 were a larger fraction in BARE (92.11% ± 4.32%) than in vegetated treatments (84.78% ± 4.88%, p = 0.049). In the 5–10 cm layer, WR_0.25 were also a larger fraction in BARE (92.31% ± 2.25%) than in vegetated treatments (84.78% ± 4.88%, p = 0.093). Thus, vegetated treatments increased the silt plus clay fraction and decreased the macroaggregate fraction of soil aggregates. However, the differences between BARE and vegetated treatments in the 5–10 cm layer were not significant (p > 0.05). In the 0–5 cm soil layer, the MWD (20.68 ± 9.21 mm) and GWD (14.44 ± 13.81 mm) of BARE were greater than those in vegetated treatments (MWD: 8.67 ± 3.97 mm; GWD: 1.82 ± 1.85 mm). By contrast, in the 5–10 cm soil layer, the MWD (3.72 ± 1.59 mm) and GWD (1.34 ± 0.33 mm) of BARE were smaller than those in vegetated treatments (MWD: 7.61 ± 1.66 mm; GWD: 2.30 ± 1.62 mm).

Soil chemical properties (

Table 2. Soil chemical characteristics with different herbaceous species.

Soil Layer	Species	pH (mol L−1)	CEC (cmol kg−1)	SOC (g kg−1)	TN (g kg−1)	AN (mg kg−1)	AN/TN	AP (mg kg−1)	TK (g kg−1)
0–10 cm	RH	5.95 ± 0.24	18.72 ± 1.31	16.10 ± 4.20	1.37 ± 0.30	46.22 ± 3.08	0.04 ± 0.01	31.91 ± 2.93	6.38 ± 2.15
	AS	6.17 ± 0.08	18.87 ± 0.77	8.68 ± 1.31	0.92 ± 0.07	43.29 ± 3.04	0.05 ± 0.00	19.12 ± 8.91	8.44 ± 1.33
	ET	6.17 ± 0.10	19.34 ± 0.90	1.03 ± 0.45	0.23 ± 0.07	37.68 ± 6.05	0.19 ± 0.08	18.13 ± 8.66	3.45 ± 0.44
	ME	5.81 ± 0.18	19.37 ± 1.80	6.74 ± 5.66	0.74 ± 0.53	41.83 ± 7.89	0.11 ± 0.13	31.04 ± 3.48	6.11 ± 1.13
	GL	6.03 ± 0.09	20.50 ± 1.53	6.17 ± 1.61	0.54 ± 0.17	38.81 ± 1.50	0.08 ± 0.03	22.22 ± 5.37	6.34 ± 3.62
	EN	6.03 ± 0.04	19.35 ± 1.05	17.80 ± 7.70	1.31 ± 0.54	47.33 ± 10.79	0.04 ± 0.02	19.82 ± 7.22	8.90 ± 0.68
	TR	5.92 ± 0.13	20.51 ± 1.11	14.28 ± 6.03	1.23 ± 0.44	43.18 ± 0.29	0.04 ± 0.02	28.71 ± 0.40	8.60 ± 0.82
	BR	6.00 ± 0.08	19.78 ± 0.79	12.35 ± 7.19	1.06 ± 0.44	42.95 ± 3.00	0.05 ± 0.02	20.61 ± 7.05	8.33 ± 0.81
	BARE	6.07 ± 0.05	22.94 ± 1.43	13.46 ± 0.97	1.53 ± 0.35	46.00 ± 1.24	0.03 ± 0.01	36.08 ± 0.84	4.43 ± 0.08
	Mean value of vegetated treatments	6.01 ± 0.12	19.55 ± 0.67	10.39 ± 5.71	0.92 ± 0.40	42.66 ± 3.28	0.07 ± 0.05	23.94 ± 5.67	7.07 ± 1.86

Note: RH: Rheum palmatum L., AS: Asparagus officinalis L., ET: Elymus dahuricus Turcz. ME: Medicago sativa L., GL: Glycyrrhiza pallidiflora Maxim., EN: Elytrigia repens (L.) Desv.ex Nevski, TR: Trifolium repens L., BR: Bromus inermis Leyss. BARE: no vegetation coverage.

) are expressed as the difference between vegetated treatments and BARE at the 0–10 cm depth in Figure S1. The mean value of the CEC under vegetated treatments (19.55 ± 0.67 cmol kg⁻¹) was lower by approximately 15% than that under the BARE condition (22.94 ± 1.43 cmol kg⁻¹, Figure S1d). The mean values of total N (0.92 ± 0.40 g kg⁻¹, Figure S1e) and AP (23.94 ± 5.67 mg kg⁻¹, Figure S1b) were lower under vegetated treatments by approximately 39% than those under the BARE condition (TN: 1.53± 0.35 g kg⁻¹; AP: 36.08 ± 0.84 mg kg⁻¹). By contrast, the mean value of TK (7.07 ± 1.86 g kg⁻¹, Figure S1a) under vegetated treatments increased by more than 59% compared with that under the BARE condition (4.43 ± 0.08 g kg⁻¹). The mean value of the AN/TN ratio under vegetated treatments (0.07 ± 0.05, Figure S1f) was greater by approximately 134% than that under the BARE condition (0.03 ± 0.01). The not-significant related factors (p > 0.05) in Figure 3 are AN (Figure S1c) and SOC (Figure S1g), which were lower under vegetated treatments (42.66 ± 3.28 mg kg⁻¹ and 10.39 ± 5.71 g kg⁻¹) than those under the BARE condition (46.00 ± 1.24 mg kg⁻¹ and 13.46 ± 0.97 g kg⁻¹).

3.5. Evaluation of Short-Term Artificial Revegetation

Principal component analysis was conducted to evaluate the short-term effects of artificial revegetation on an erosion gully. The evaluation was based on effects on soil and water conservation, soil quality, and vegetation growth (

Table 3. Comprehensive evaluation scores for effects of artificial revegetation by different herbaceous species in an erosion gully on soil and water conservation, soil quality, and vegetation growth.

Species	Evaluation Aspects
	Soil and Water Conservation	Soil Quality	Vegetation Growth
RH	0.08	0.15	−1.26
AS	3.16	−0.61	−0.06
ET	−0.01	−3.42	−0.26
ME	−0.11	−0.02	−0.37
GL	−0.44	−0.13	−1.50
EN	−0.97	0.87	0.86
TR	−0.98	1.79	0.68
BR	−0.73	1.37	1.92

Note: RH: Rheum palmatum L., AS: Asparagus officinalis L., ET: Elymus dahuricus Turcz. ME: Medicago sativa L., GL: Glycyrrhiza pallidiflora Maxim., EN: Elytrigia repens (L.) Desv.ex Nevski, TR: Trifolium repens L., BR: Bromus inermis Leyss.

). The evaluation of soil and water conservation was based on SY, SR, τ, ω, Re, and Fr (Table S3). The weighting of each parameter was 18.19%, 14.40%, 17.32%, 16.25%, 17.75%, and 16.09%, and the evaluation scores for the eight herbaceous species were ranked AS > RH > ET > ME > GL > BR > EN > TR. In this evaluation, the higher the ranking was, the worse the soil and water conservation effect. The evaluation of soil quality was based on SOC, TN, C/N ratio, sand, silt, clay, pH, BD, FC, SP, TK, AP, AN, CEC, WR_0.25, MWD, and GWD. The weighting of each parameter was 5.24%, 5.50%, 4.98%, 5.91%, 6.11%, 4.77%, 4.89%, 6.29%, 5.62%, 6.29%, 5.70%, 5.59%, 5.99%, 4.70%, 5.91%, 5.49%, and 4.82%, and the evaluation scores were ranked TR > BR > EN > RH > ME > GL > AS > ET. The evaluation of vegetation growth was based on VC, SDW, SRL, RSR, RLD, RSD, and RV. The weighting of each parameter was 12.67%, 13.50%, 13.79%, 15.06%, 14.39, 15.23%, and 15.37%, and the evaluation scores were ranked BR > EN > TR > AS > ET > ME > RH > GL.

The selection of suitable herbaceous species for artificial revegetation is very important in preventing soil degradation in erosion gullies. Based on the objectives, the suitable herbaceous species varied a bit. However, the TR, BR, and EN consistently had the top three rankings by score in all three aspects evaluated. This means that the three herbaceous species (TR, BR, and EN) were suitable for preventing soil degradation in erosion gullies. Based on the characteristics and growing period of the herbaceous vegetation planted (Tables S1 and S2), TR was better than BR and EN. Therefore, TR was the most suitable species for use in artificial revegetation to prevent soil degradation in an erosion gully.

4. Discussion

4.1. The Effects of Artificial Revegetation with Herbaceous Species on Soil Erosion

Reducing soil erosion is critical for controlling gully development [4]. Artificial revegetation can quickly alter soil surface properties and reduce soil losses during surface runoff [23]. In Figure 3, vegetation treatments had a strong relationship with sediment yield (SY) and erosion rate (ER). MWD and GWD were significant factors influencing the SY and ER. Increased aggregate stability will reduce soil erosion by minimizing rainfall-induced aggregate disturbance and increasing soil pores [26,27]. However, the lower MWD and GWD with revegetated treatments (Table 1) did not result in increased sediment output (Figure 3). This meant that vegetation cover played an essential role in lowering sediment yield.

Figure 4 shows how vegetation treatments reduced soil erosion. Soil erosion was directly linked to vegetation features [7]. Previous research has shown that vegetation coverage, and shoot and root traits all have an impact on soil erosion. However, in our investigation, vegetation coverage was not significantly related to sediment output (p > 0.05). In prior research, the role of vegetation coverage was to minimize rainfall kinetic energy and prevent soil particles from dispersing [11]. The influence of herbaceous vegetation cover was not investigated because our study focused on runoff scouring tests. However, the roots of herbaceous vegetation play an important role in soil stabilization [28]. AS produced more sediment than other vegetated treatments due to its poor vegetation coverage (Figure 4).

4.2. Changes in Soil Physicochemical Characteristics

Short-term (16 months) artificial revegetation of an erosion gully slope with herbaceous vegetation did not improve soil quality. It significantly decreased soil nutrient concentrations (except TK), WR_0.25, and aggregate water stability compared with bare soil.

Although sediment yield decreased with revegetated treatments, the WR_0.25, MWD, and GWD suggested short-term herbaceous species revegetation destroyed soil structure and, instead of conserving nutrients, released them for plant uptake. Abiotic and biotic processes (e.g., SOM and microorganisms) interact to form soil aggregates as a result of aggregation and fragmentation processes [29]. The decreased aggregate stability can be attributed to vegetation root penetration. By affecting the size of aggregates and the connectivity or size by producing [30] or clogging [31] pores, root penetration primarily affects soil structure. Growing roots can cause soil particles and aggregates to become pressed by growing roots during the initial stages of vegetation growth [32,33]. It is unavoidable that the growth of roots leads to a reduction in the porosity of rhizosphere soil when the penetration resistance of soil is greater than the pressure. Also, thickened and lengthened roots can disperse soil particles or aggregates [34] and crack macroaggregates [35]. Although short-term artificial revegetation did not increase aggerate stability in this study (Figure 3), aggerate stability appeared to tend toward future increases, with increased large macroaggregates (Figure 5).

In previous studies, soil nutrients increased after vegetation restoration [36,37,38]. However, in our study, SOC and AN were not sensitive to short-term artificial revegetation. The soil nutrient-preserving capability (represented by the CEC) decreased with the vegetated treatments (Figure 3). With the exception of TK, soil nutrients were generally lower under vegetation treatment than in bare soil. One reason to explain the lack of sensitivity could be that because the herbaceous litter input was relatively small, the C input to the soil was also relatively small during the experiment. Secondly, a shallow root system, weak soil consolidation, and the need for plants to absorb nutrients are also factors to consider. The input of soil nutrients was less than the nutrients consumed by vegetation growth in the short term. The third reason was the breakdown of soil aggregate. Newly imported nutrients can benefit from soil aggregates because they provide physical protection and reduce their accessibility by microbes [39,40,41]. However, with the breakdown of soil aggregate by root penetration, the soil nutrients are mineralized and lost [42]. The decrease in WR_0.25 (Figure 5) mainly caused by the breakdown of soil aggregates could contribute to the loss of SOC and soil nutrients. The last reason could be the “priming effect”. The input of vegetation litter with a high C:N ratio can lead to priming effects and accelerate the decomposition of native SOC [43]. Compared with farmland soils, the addition of exogenous organic matter (vegetation litter) to gully slope soil with a relatively low SOC content may lead to greater stimulation of microbial activity and consequently drive increases in the decomposition of native SOC [44]. Although the TN under vegetated treatments (0.09% ± 0.04%) was lower than that under bare soil (0.15% ± 0.04%), the AN was not significantly different between bare soil and vegetated treatments. The results indicated that short-term artificial revegetation increases the AN/TN ratio. Available N consists of “temporary and slow-acting reservoirs” of SON, which are the main sources of easily mineralized N in soil [45]. As herbaceous vegetation has a relatively short life cycle and turnover, the soil microorganisms use readily decomposable litter from herbaceous vegetation to increase their AN/TN ratios. The increased AN/TN ratio can lead to a relatively rapid rate of soil N cycling [46]. The increase in TK indicated that soil K in an erosion gully could be restored under artificial revegetation in a short time, consistent with the results of Liu and Wang [47]. One reason for the rapid recovery could be that K is not a limiting element in the black soil area. Another reason for the increase could be the high contents of K returned in plant litter leading to soil K enrichment.

4.3. The Evaluation of Herbaceous Species

The goal of our study was to find suitable herbaceous species to prevent soil degradation in a black soil erosion gully based on three aspects. To our surprise, the top-ranked three species (BR, EN, and TR, in Section 3.5) were the same in all aspects evaluated. Vegetation coverage is a critical factor affecting soil and water loss [11]. Yan et al. [23] found that increasing vegetation coverage will reduce soil and water loss in the same place. In that regard, the TR was best in controlling gully development. Improving soil quality is another role of artificial revegetation [48]. TR revegetation increased the SOC contents, reduced soil nutrient loss, and functioned best among the herbaceous species, followed by BR and EN. The vegetation growth status is also very important in controlling gully development [23]. Root characteristics are important vegetation growth indexes that decide the role of herbaceous species in changing soil characteristics (Figure S2). A root density increase leads to the formation of stable macroaggregates because roots and associated fungi form a mesh that entangles fine soil particles. In addition, roots produce exudates that can serve as adhesives and may stimulate the activity of microbes, both of which contribute to soil aggregate stability [49,50]. Based on the above reasons, the BR was best in preventing soil degradation among the herbaceous species, followed by TR and EN. Although BR, EN, and TR were the top-ranked three species in all aspects evaluated, the TR was the best of them based on the germination rate, clarity, and resume growth ratio.

Artificial revegetation is typically accomplished with trees and shrubs in degraded areas [51]. However, comparing the results with herbaceous species with those with trees and shrubs, herbaceous species played a more important role in reducing soil and water loss, soil nutrient loss, and soil carbon loss [13,52]. Especially in some serious eroding environments like erosion gully slopes, trees and shrubs have been difficult to grow, with few nutrients and poor soil structure [53]. Herbaceous species were demonstrated to be a viable alternative in our study.

5. Conclusions

The effects of short-term artificial revegetation with herbaceous species in preventing soil degradation in an erosion gully of a black soil area were evaluated. A field runoff scouring experiment verified that artificial revegetation conserved soil and water, conserving soil by decreasing sediment yields. There were significant differences in soil TN, TK, and AP between vegetated treatments and bare soil. With a decrease in CEC, soil TK and AN/TN increased under short-term vegetated treatments. Although WR_0.25 decreased under vegetated treatments, the silt plus clay fraction (<0.053 mm) was greater than that with bare soil. On the basis of a comprehensive evaluation of the effects on soil and water conservation, soil quality, and vegetation growth, TR is recommended as the most suitable herbaceous species to prevent soil degradation on an erosion gully slope. Although short-term revegetation did not improve soil quality in gullies, continuous vegetation growth will strengthen the potential to regulate gully formation and soil degradation. Leguminosae herbaceous species such as TR, which have a high vegetation coverage, are excellent alternatives for gully restoration. Artificial revegetation with herbaceous species has been shown to be an effective method for reducing soil erosion in gullies. Given the cost and time involved with herbaceous species, we believe that artificial revegetation with herbaceous species or a combination of trees, shrubs, and herbaceous species is a sustainable strategy to minimize soil degradation in a black soil erosion gully in Northeast China.