Geometry Morphology and Distribution Characteristics of Permanent Gullies in the Greater and Lesser Khingan Mountains Forest Region of China

Abstract

The Greater and Lesser Khingan Mountains (GKM and LKM), together form one of the main resources of the terrestrial natural ecosystem in northeast Asia and play a crucial role in climate regulation and soil and water conservation due to their distinctive geographical features and abundant vegetation cover. Nonetheless, the morphology and distribution of gullies in the two study areas remain unclear. This study focused on an investigation area of approximately 100 km² within the forest areas of the GKM and LKM, where field measurements were conducted to record and analyze the morphological characteristics of the gullies. The study also explored the impact of slope and the aspects of gully development and established a gully volume estimation model in the study area. The findings indicate the following. Firstly, that the proportions of gullies with a length of 200–1000 m, a width of 2–6 m and a depth of 1–2 is 59.4%, 51.3% and 45.9%, respectively in the GKM, and 42.5%, 75.7% and 56%, respectively in the LKM. The measured gully density in the GKM was 0.3 gullies per km², with an average length, width, and depth of 524.4 m, 2.4 m, and 1.0 m, respectively. In contrast, the measured gully density in the LKM was 0.45 gullies per km², with an average length, width, and depth of 560.1 m, 3.9 m, and 1.8 m, respectively. Secondly, as the slope increased, the density of gullies and the degree of surface fragmentation gradually decreased. In the measured area of the GKM, gullies developed faster on the semi-sunny slope. However, in the measured area of the LKM, gullies were more evenly distributed across different slopes. A significant power function relationship between the volume and area (V-A) of gullies in the measured areas of the GKM (V = 0.37 A^1.11, R² = 0.94) and LKM (V = 0.32 A^1.17, R² = 0.94) was observed. These findings have important implications for soil conversation in forested areas of the black soil region in Northeast China.

1. Introduction

Gully erosion is a severe form of land degradation that poses significant threats to land, water resources, and the ecological environment [1,2]. Moreover, gully erosion can exacerbate flood disasters, worsen water quality, and disrupt ecosystems [3]. Gully erosion has become an increasingly prominent issue in recent years, exacerbated by global climate change and human activities, particularly in ecologically fragile areas such as the black soil region in Northeast China [4,5]. As a response to this issue, a growing number of studies have been dedicated to exploring the morphological characteristics, distribution patterns, and control methods of gully erosion, aiming to provide a scientific basis for decision making. Gullies are not limited to marly badlands or mountainous areas, but can also be found in soils affected by soil crusting, such as the loess found in Europe, in the Chinese Loess Plateau, and in North America, as well as the sandy soils in the Sahelian zone and northeast Thailand [6,7]. They can also occur in soils that are susceptible to piping and tunnelling, such as dispersive soils [8]. Meanwhile, Zheng et al. [9] have provided a comprehensive review of the research progress on gully erosion over the past 60 years, covering topics that include comparisons of gully erosion concepts, the critical models for gully erosion occurrence, the influencing factors of gully erosion process, the prediction models of gully erosion, research methods and technologies, and proposed key areas for future research. Additionally, recent studies have focused on investigating the relationship between gully erosion and ecological factors, such as land use and vegetation cover, with the aim of improving the understanding of the mechanisms and laws of gully erosion [10,11,12,13].

The Northeast of China is subject to multiple erosion forces, such as hydraulic, wind, and freeze–thaw. These alternate or synchronize in time and overlap or interweave in space, leading to the coexistence of multiple erosion types that interact with each other [14]. According to the 2022 China Water Resources Bulletin, the area’s soil erosion accounts for 26.79% of the total land area, with gully erosion alone causing land damage exceeding 5000 km². The characteristics of gully erosion in this region are significantly different from those in other regions, with shorter formation time and faster development speed. Current research on gullies primarily focuses on spatial differentiation characteristics, influencing factors, and field measurement and remote sensing interpretation to study the distribution of gullies [15,16,17]. Wang et al. [18] used unmanned aerial vehicle (UAV)-derived orthoimages and digital terrain models (DTMs) with centimeter-level resolution obtained from 2015 to 2020 to periodically monitor the geomorphic, morphological, and volume changes of a stabilized gully in a small agricultural watershed (6 ha) in the southern black soil region of northeast China. The study conducted annual and seasonal assessments of erosion rates in the region and demonstrated the strong adaptability and various advantages of multitemporal UAV data for short to medium-term (≤5 years) erosion rate assessments compared with submeter resolution satellite images. Li et al. [19]. used high-resolution WorldView 2 stereo images (0.5 m) and ENVI5.3 software to analyze the morphology characteristics of gullies in the mountainous and hilly areas of Northeast China’s black soil region and developed a volume estimation model. Li et al. [13] investigated the evolution of gullies in Guangrong village, Hailun city, Heilongjiang province. The authors found that 93% of the incised gullies originated from pre-existing channels in 1968, and that their development process was mainly characterized by channel merging and gully wall widening. Despite the advanced capabilities of current LiDAR technology to penetrate cloud cover and areas with dense vegetation, it still encounters limitations with regard to pristine forested regions and expansive watersheds [20]. The interpretation results often underestimate the actual outcomes. The Chinese Ministry of Water Resources used 2.5 m resolution remote sensing images to conduct the first national water conservancy census in 2013, finding that more than 80% of the 295,663 gullies in the black soil region of Northeast China had been developed into farmland [21]. However, identifying gullies in the forested areas was challenging due to the dense vegetation cover. The thick canopy and undergrowth made it difficult to visually spot and delineate the gully features accurately. Moreover, in some areas, gullies were partially obscured by fallen leaves and debris, further complicating the field measurements. Therefore, the field investigation is still a more effective method by which to obtain the three-dimensional information of each gully than are remote sensing interpretation and LiDAR in a larger region.

The Greater and Lesser Khingan Mountains, which protect over 1/10 of China’s cultivated land and the largest grassland, are critical sources and water conservation areas for the Nenjiang and Heilongjiang river systems and their tributaries [22]. Thus, they have a special status in the overall national ecological protection strategy. Despite this, little research has been undertaken on gullies in the forested areas, and the morphology and development status of gullies remain unclear. This is extremely detrimental to the search for a comprehensive and in-depth understanding of the current status of gully erosion in forest areas, to the making of informed decisions on gully management, and to the implementation of effective prevention measures.

In this study, field measurements were conducted to determine the morphological parameters of every gully in an area of approximately 100 km² in both the Greater and Lesser Khingan Mountains. The aim was to clarify the development and distribution characteristics of gullies, gully volume estimation and the relationship between gully morphology and topographic distribution patterns in the forest areas of the Greater and Lesser Khingan Mountains. The results of the study deepen our understanding of the current situation of gully erosion in the Northeastern black soil region of China and provide a scientific basis for decision making on gully erosion control measures.

2. Materials and Methods

2.1. Study Area

The study conducted field measurements in two forested areas: Dashimen forest in the southern part of Zhalantun city, situated in the Greater Khingan Mountains with an average slope of 11.8° and Jianshe forest in the eastern part of Beian city, situated in the Lesser Khingan Mountains with an average slope of 3.6° (Figure 1). Zhalantun City is located on the eastern foot of the Greater Khingan Mountains and to the west of the Songnen plain in China. The region is characterized by mountainous terrain in the northwest and hilly terrain in the southeast. It mainly falls into three major types: medium-to-low mountainous areas, hilly, and river valleys. The predominant rock types in this area are schist, followed by granite and quartzite, which are widely distributed throughout the region. The winter in this region is long and bitterly cold, while the summer is short with concentrated rainfall. The spring and autumn experience rapid temperature changes. The average annual temperature is 2.7 °C, and the average annual precipitation is 500 mm. Within the territory of Zhalantun city, the soil horizon is characterized by black soil, belonging to the black soil zone of the Songnen plain. There are primarily six types of soils, namely brown coniferous forest soil, dark brown soil, black soil, meadow soil, marsh soil, and paddy soil. The forest in this area is sparsely distributed, and the dominant tree species include Quercus serrata and black birch.

On the other hand, for Beian city (

Table 1. Comparison of environmental factors.

Parameters	The Great Khingan Mountains	The Lesser Khingan Mountains
Vegetation cover	67%	91%
Gravel content	27%	52%
Average shape	11.8°	3.6°
Average temperature	2.7 °C	0.8 °C
Average rainfall	500 mm	566 mm
Soil texture	Clay	Loam

), located on the southwest foothills of the Lesser Khingan Mountains, the terrain of the entire area presents an east-high-west-low and north-high-south-low pattern. This complex topography results in considerable climate variability. Most of the city is characterized by lacustrine deposits, and the geological conditions for mineralization are relatively poor. The average annual temperature is approximately 0.8 °C, while annual precipitation varies from 358.8–773.8 mm. During winter, early spring, and late autumn, it is primarily influenced by the Siberian cold high-pressure system, which is characterized by prevailing northerly winds, low precipitation, and dry air, and results in cold temperatures. In the territory of Beian city, there are primarily two soil types: black soil and dark brown soil. The natural forests in this area are characterized by poplar and birch, while the planted forests are mainly composed of larch and camphor.

2.2. Methodology for Field Survey of Gully

The field survey was conducted in April and May 2022, in the LKM and GKM, respectively. The survey area was determined based on topographic features and land use. Prior to the field survey, high-definition satellite images of the survey area were collected using Omap (version 9.7.1 Beijing Yuanshenghua Network Technology Co., Ltd. Beijing, China, www.ovital.com, accessed on 15 April 2022) to preliminarily identify the distribution of gullies within the area. The survey area was divided into sub-areas, and specific survey routes were developed. Each team, comprising two individuals, surveyed the sub-areas. Upon locating a gully, the team numbered it and measured the position of the gully head, tail, and cross-section using GPS. A laser rangefinder (NF-273L red light) was used to measure the width above and below the gully, as well as the depth in and far from the gully. The development status and land use type were also recorded, and photographs of the cross-sections were taken [9]. To ensure that the measured sections represented the morphological characteristics of gullies, at least 5 sections were set for gullies <100 m in length, and 8–15 sections were generally set for gullies >100 m in length. Each surveyed gully was marked on the Omap to avoid duplication or omission. After the field survey was completed, the information of each gully in the surveyed area was summarized, and the geographic location information of the gullies was imported into ArcMap 10.2. The line file number of each gully corresponded to the number of the gully in the survey and was manually corrected using high-definition satellite images to minimize error. Figure 2 shows the real status of gullies in the two study areas.

The gully data obtained from field measurements were used to extract the length of each gully using ArcMap10.2. Microsoft Excel 2021 was utilized to compute various parameters, including gully area, gully volume, ground lacerative degree, and gully density. These parameters were calculated using the following methods.

The number density of gullies is the number of gullies per unit area, calculated as shown in Equation (1):

$$f=\frac{N}{A}$$

(1)

where f represents the density of gullies, expressed in units of per km²; N represents the total number of gullies; A represents the area of the region, in km².

The gully area is calculated as shown in Equation (2):

$$SA=\frac{{\displaystyle \sum _i^n{W}_i}}{n}\cdot L$$

(2)

where SA represents the gully area in km², W_i denotes the width of the i-th measuring point on the gully, L represents the length of the gully in m, and n is the number of measurements taken.

The formula for calculating the trench volume is shown in Equation (3):

$$V_G={\displaystyle \sum _{i=1}^n\left[\frac{(W_{iu}+W_i{}_d)}{2}\cdot D_i\cdot L_i\right]}$$

(3)

where V_G represents gully volume in m³, W_iu is the average width of the gully at the i-th measurement point, W_id is the average width of the gully at the bottom of the i-th segment, D_i is the average depth of the i-th segment, and L_i is the length of the slope measured for the i-th segment.

The gully density is an indicator of the total length of gullies per unit area, which is calculated according to Equation (4):

$$GD=\frac{L_g}{A}$$

(4)

where GD represents gully density in km/km², and L_g is the total length of gullies within the area, in km.

The ground lacerative degree represents the total area of gullies per unit area and is calculated as shown in Equation (5):

$$GLD=\frac{S}{A}$$

(5)

where GLD is the ground lacerative degree in km²/km⁻² and S is the total area of gullies in the region, in km².

The root mean square error is the square root of the average of the squared differences between predicted values and true values, divided by the number of observations, n, and is calculated according to Equation (6):

$$RMSE=\sqrt{\frac{{\displaystyle {\sum }_{i=1}^n{(V_{Gi}-V_M{}_i)}^2}}{n}}$$

(6)

where V_Gi and V_Mi are the measured and modelled gully volumes (m³), respectively.

Downloading 12.5 m resolution DEM within the study area as a data source via Google Earth, various tools in ArcMap10.2′s ArcToolbox, including 3D Analyst and Spatial Analyst, were used to intersect the gullies polyline shape file with slope layer from DEM, the average ditch width in the study area is much less than the DEM resolution, and the topographic slope is considered to be approximately equal to the gully slope. The slope gradients of the Greater and Lesser Khingan Mountains were reclassified into the following categories: 0–2°, 2–4°, 4–6°, 6–8°, 8–10°, 10–12°, 12–14°, 14–16°, 16–18°, 18–20°, 20–22°, 22–24°, 24–26°, 26–28°, 28–30°, and 30–32°. Gully parameters for different slopes and aspects were obtained from this information. By overlaying the gully parameters of the same slope and aspect, all of the gully parameter values for that slope and aspect were obtained. The slope aspect classification is as follows: N represents the north-facing slope (shaded slope), S represents the south-facing slope (sunny slope), NW and NE represent the northwest and northeast slopes, respectively (semi-shaded slopes), and SW and SE represent the southwest and southeast slopes, respectively (semi-sunny slopes).

Correlation analysis of the parameters of length, width, depth, area, and volume in the study areas of GKM and LKM was conducted using IBM SPSS Statistics 27.0.1, and the correlation statistical graphs were plotted using Origin 2023.

3. Results

3.1. General Characteristics of Gully Morphology

After field measurements, a total of 87 gullies were identified in the study area of the GKM and LKM (Figure 3). As shown in

Table 2. Morphology parameters of the gullies in the study area of the Great and Lesser Khingan Mountains.

Parameters	The Great Khingan Mountains	The Lesser Khingan Mountains
Study area (km2)	114.87	124.45
Average length (m)	524.42	560.13
Average width (m)	2.45	3.93
Average depth (m)	1.09	1.79
Average area (km2)	1243.93	2445.94
Average volume (m3)	1153.18	3365.55
f (gully km−2)	0.32	0.40
GD (km/km−2)	0.12	0.12
GLD (km2/km−2)	0.05	0.03

, the GKM study area had 37 gullies with a density of 0.32 gullies per km², and the gully lengths ranged from 26.33 m to 3303.2 m, while the widths and depths ranged from 0.40 m to 7.65 m and 0.4 m to 2.48 m, respectively. The average length, width, depth, area, and volume of the gullies in this area were 524.42 m, 2.45 m, 1.09 m, 1243.93 m², and 1153.18 m³, respectively. The GD in this area was 0.12 km/km⁻², and GLD was 0.05%. In the LKM study area, there were 50 gullies, resulting in a gully density of 0.40 gullies per km². The range of gully lengths, widths, and depths were 14.70 m to 2182.23 m, 0.9 m to 9.0 m, and 0.5 m to 8.0 m, respectively. The average gully length, width, and depth were 560.13 m, 3.93 m, and 1.79 m, which were 1.07 times, 1.60 times, and 1.64 times greater than those in the GKM. The individual gully area and volume in this area could reach 2455.94 m² and 3365.55 m³, or 2.0 times and 2.9 times greater than those in the Greater Khingan Mountains forest area, respectively. Although the GD was similar to that of the GKM, the GLD was only 0.03 km/km⁻².

3.2. Gully Morphology Distribution Characteristics

According to Figure 4, over 75% of the gullies in both study areas are longer than 100 m, with the highest percentage of gullies in the LKM study area being in the 200–500 m and 1000–2500 m length categories, at 30% and 24%, respectively, followed by the 100–200 m length category, at 20%. Gullies with lengths less than 100 m account for only 14%. However, in the GKM study area, gullies are concentrated in the 200–1000 m length range, accounting for 62.2% of the total, while gullies in the <50 m, 50–100 m, and 1000–2500 m length categories are almost equally distributed, accounting for 10.8%. The 100–200 m length category has the lowest percentage, at only 5.4%.

The width of gullies in the LKM is mostly distributed between 2–6 m, accounting for 62% of the total, followed by the 4–8 m category at 36%, while gullies with a width of 8–10 m are the least common, accounting for only 2%. In contrast, in the GKM, gullies are mostly distributed between 0–4 m, accounting for 86.5% of the total, followed by the 4–6 m category at 8.1%, while gullies with a width of 6–8 m are the least common, accounting for only 5.4%.

The frequency distribution of gully depths in the study areas of the GKM and LKM follows an increasing-then-decreasing trend with increasing width categories. In the LKM, gully depths are mostly distributed within the range of 1.5–2 m, accounting for 38% of the total, followed by the range of 0.5–1.5 m, which represents 36% of the total. The percentage of gullies deeper than 2.5 m is the lowest, at only 6%. In the GKM, gully depths are mostly distributed within the range of 0.5–1.5 m, accounting for 73% of the total, followed by the range of 1.5–2 m, which represents 16.2% of the total. The percentage of gullies in the remaining width categories is only 10.8%.

The gully area distribution in the LKM is relatively even among the same level. The average proportion of the three level ranges, namely <0.05 hm², 0.10–0.2 hm², and >0.3 hm², is about 25.3%. The proportion of the 0.05–0.1 hm² level is 14% of the total, while gullies with an area of 0.2–0.3 hm² only account for 10%. On the other hand, in the GKM, the gully area is mainly concentrated in the level of <0.05 hm², accounting for 43.2% of the total.

In the LKM, the single gully volume is mainly distributed in the range of 500–2000 m³, accounting for 36% of the total, followed by 100–500 m³, accounting for about 20%. The gullies in the three volume ranges of 2000–5000 m³, 5000–10,000 m³, and >10,000 m³ have an average proportion of about 12.7%. The proportion of gullies with a volume less than 100 m³ is the smallest, at only 6%. In the GKM, gullies with volumes of 100–500 m³ and 500–2000 m³ account for the largest proportion, reaching 40.5% and 32.4% respectively, followed by gullies in the <100 m³ and 2000–5000 m³ ranges, accounting for 10.8%. The proportion of gullies in the 5000–10,000 m³ range is the smallest, at only 5.5%.

3.3. Topographical Distribution

The variations of gully density and surface fragmentation with slope in the forest area of GKM and LKM are shown in Figure 5. In the study area of the GKM, GD and GLD are highest at slopes between 0–1°. As a slope increases from 0–6°, GD and GLD sharply decrease from 0.89 km/km⁻² to 0.12 km/km⁻² and from 0.0017 km²/km⁻² to 0.0003 km²/km⁻², respectively. Subsequently, both of these show a slow fluctuating decrease with the increase of slope. The fitted results indicate that GD and GLD exhibit a significant exponential decrease with slope (p < 0.05).

Figure 6 illustrates the changes in GD and GLD along eight slope orientations: north (N), northeast (NE), east (E), southeast (SE), south (S), southwest (SW), west (W), and northwest (NW) in the GKM and LKM. In the study area of the GKM, the gully density (0.33 km/km⁻²) and surface fragmentation (0.0007 km²/km⁻²) were highest on the east-facing slope, followed by the southeast-facing slope. The northwest-facing slope had the lowest gully density (0.06 km/km⁻²) and surface fragmentation (0.0002 km²/km⁻²). In the study area of the LKM, the southeast-facing slope had the highest gully density (0.29 km/km⁻²), followed by the south-facing slope, while the southwest-facing slope had the lowest gully density (0.17 km/km⁻²). In contrast with the GKM study area, surface fragmentation was highest on the north-facing and northwest-facing slopes (0.0012 km²/km⁻²) in the LKM study area, followed by the southeast-facing slope, and lowest on the east-facing slope (0.0007 km²/km⁻²).

3.4. Gully Volume Estimation

As shown in Figure 7a, in the measured area of the GKM, a gully’s volume is strongly positively correlated with the gully area and length, and significantly positively correlated with the gully width and depth. Among these, the correlation with the gully area is the strongest (correlation coefficient of 0.93), followed by gully length, width, and depth. Therefore, the gully area can be used as the preferred parameter for estimating the volume of gullies in the GKM. Regression analysis shows that the gully volume is in an extremely significant power function relationship with the gully area (V = 0.37 A^1.11, R² = 0.94), and that this relationship can be used to estimate the volume of gullies in the GKM (Figure 8a).

The correlation between gully volume and other gully morphological parameters in the measured area of the LKM is almost identical to that of the GKM, where the gully area is the parameter with the highest correlation with volume, followed by gully length, width, and depth (Figure 7b). As shown in Figure 8b, the gully volume in the LKM is in an extremely significant power function relationship with the gully area (V = 0.32A^1.17, R² = 0.94).

4. Discussion

4.1. The Uniqueness of Gully Morphology in the Greater and Lesser Khingan Mountains

The main factors that affect gully erosion are terrain, rainfall, composition of the ground material, soil thickness, land use, and human activities [23,24]. The average annual rainfall in Beian city is higher than that in Zhalantun city (Table 1). The soil in the LKM is relatively thick, with a loose texture and low gravel content, making the soil highly erodible. Compared with the LKM, the soil in both the small and large ranges of the GKM is relatively thin and contains a high percentage of gravel. The average slope of the LKM (3°) is smaller than that of the GKM (12°), resulting in a larger catchment area for the LKM. It can be observed from research that the area of forest destruction is greater in the LKM, and the vegetation coverage is lower than that of the GKM. Due to the loose texture and low gravel content of the soil in the LKM, the gully downcutting is more severe. Additionally, lateral erosion during water flow scouring causes bank collapse and gully widening. Therefore, the average length, width, and depth of the LKM are slightly greater than those of the GKM.

Compared with other regions (

Table 3. The gully parameters vary between different study areas [19,27,28,29,30,31].

Study Area	Average Length (m)	Average Width (m)	Average Depth (m)
The Greater Khingan mountains	524.5	2.5	1.1
The Lesser Khingan mountains	560.1	3.9	1.8
Rolling hilly region in northeast China [27]	522.32	15.05	2.77
Mountainous and hilly region of northeast China [19]	81.9	5.3	1.05
Loess plateau [28]	58.4	9.0	6.2
Yuanmou dry–hot Valley in Southwestern China [29]	62.92	18.69	-
The upper Blue Nile basin, Ethiopia [30]	138.25	5.5	2.4
The Lake Manyara basin, northern Tanzania [31]	283.54	65.6	1.21

), the average length of gullies in this study area is significantly higher than those of the Loess plateau, Yuanmou dry–hot valley, Ethiopia, and Tanzania. However, the average width of gullies is smaller. Previous studies have indicated that the slopes of the Loess plateau and Yuanmou dry–hot valley are steeper than those of the Northeast China black soil region, while their watershed areas are much smaller. In contrast, the Northeast China black soil region has an average annual temperature of 0 °C and is prone to freeze–thaw erosion, which promotes gully development. In the Loess plateau, the gullies tend to be wider and deeper due to the collapsibility of the loess [25]. The Yuanmou dry–hot valley, characterized by high annual temperatures, steep slopes, and loose soil, has wider gullies. In Tanzania, the region is prone to high erosive rainfall, which contributes to soil erosion. Additionally, the topographic features and soil properties are also driving factors in the formation of gullies. The soil structure in the area is relatively weak, primarily composed of sandy soil, but the subsoil contains a higher clay content, making the soil more susceptible to erosion [26]. The erosion degree of gullies is also closely related to climatic conditions, such as rainfall. The study area is located in the Northeast China black soil region, where the annual rainfall is relatively low, providing a suitable environment for gully development. Apart from climatic and geological conditions, human activities also have an impact on the formation and development of gullies. Excessive land use and unreasonable water resource development may exacerbate soil erosion, thereby promoting gully erosion. The study area of this article is located in the black soil region of China, and its soil structure is not significantly different from those of the rolling hilly or mountainous hilly regions. Therefore, it is inferred that the differences in the morphology of gullies in two areas may be attributed to the influence of other environmental factors, such as differences in climatic conditions, vegetation cover, human activities, and land use. Climatic conditions, such as precipitation patterns and temperature variations, can directly affect the erosive forces acting on the landscape. Vegetation cover acts as a natural protective barrier against erosion, and its presence or absence can impact the vulnerability of the soil to erosion. Additionally, human activities and land use practices, such as deforestation, agriculture, and construction, can alter the surface characteristics and exacerbate erosion processes.

4.2. Topographic Distribution Pattern

The results show that the GD and GLD degree in the study area decreased with increasing slope, while different results were found in other related studies. In the northeast Manchurian sandy land, the gully density and GLD increased with slope until reaching a peak, after which they decreased with increasing slope. This suggests that the relationship between GD, GLD, and slope may be influenced by various factors, such as topography, land use, and human activities. The average slopes in the GKM and LKM were 11.8° and 3.6°, respectively, while in the rolling hilly region in Northeast China, it was 2.44°. In the GKM forest area, areas with high slope had extremely high vegetation cover and almost no human activity. Vegetation cover may be a factor that leads to the different relationship between slope and GD and GLD in this area because vegetation can reduce the flow velocity of water and erosion, thus reducing the formation and development of erosion gullies. On the other hand, in the Rolling hilly region inNortheast China, due to the relatively flat terrain and low vegetation cover, gullies were promoted by both natural and human factors, leading to a different relationship between slope and GD and GLD compared with other areas. In addition, although the GLD and GD exhibit an exponential distribution in the GKM, while a linear distribution is observed in the LKM, the relationship between GD and GLD in the GKM is nearly linear within areas of an average slope of 0–12°. This pattern suggests that in the GKM, the development of gullies is primarily influenced by water flow and human activities, particularly in areas with lower slopes. Water flow plays a significant role in erosion and channel deepening, while human activities alter surface hydrological conditions and promote the formation of gullies.

The research indicates that GD and GLD on the semi-sunny slopes of the GKM are higher than those on the shady slopes. Li et al. [19] have also found that erosion on sunny or semi-sunny slopes is stronger than on shady or semi-shady slopes. In the study area, the sunny and semi-sunny slopes receive more solar radiation during the day than the shady and semi-shady slopes, and the temperature difference between day and night is also greater. This leads to differential effects of freeze–thaw cycles on the development of gullies. In addition, the sunny slopes also accelerate snow melting, leading to concentrated water flow and to the surface soil texture becoming loose, thereby decreasing erosion resistance, which is one of the reasons for the higher density of erosion gullies on the sunny slopes. Furthermore, local wind direction and rainfall also affect the development of erosion gullies. In Northeast China, warm and humid air currents mostly come from the Pacific, and in summer, there are prevailing south winds that have a greater impact on the erosion of the windward slopes by raindrops [32].

The LKM, however, has a gentler slope and a more evenly distributed solar radiation due to its lower overall slope. Consequently, the gullies that form in this area display a relatively uniform pattern. This implies that the slope could be the primary factor that influences the development of gullies in the LKM.

4.3. Estimation Model for Gully Volume

A gully’s volume (V) is a key parameter that reflects the degree of gully erosion [33]. Obtaining parameters such as gully length (L) and gully area (A) is simpler and more accurate than estimating gully volume. Therefore, studying the relationship between V–A and V–L is of great significance for further research on gully erosion. It has been found that the V–A relationship in both the GKM and the LKM follows a power function V = a·A^b, where b can be considered as the growth rate of gully depth per unit area of gully cross-section [34,35]. The b value in the GKM (1.11) is smaller than that in the LKM (1.17), indicating a greater risk of gully erosion in the latter. This phenomenon can be attributed to the smaller average slope angle and larger watershed area in the LKM, which facilitates gully formation, and the low content of soil gravel, which makes the gully easy to undercut. Additionally, the large deforestation area and frequent human activities in the LKM also increase the risk of gully erosion. Further studies have found that the b value in the study area is larger than the rolling hilly region in Northeast China but smaller than the mountainous and hilly region there (

Table 4. The estimation model for gully volume between different study areas [27,28,29,30,31].

Study Area	Volume Estimation Model
The Greater Khingan Mountains	V = 0.37 A1.11, R2 = 0.94 RMSE = 31.48
The Lesser Khingan Mountains	V = 0.32 A1.17, R2 = 0.94 RMSE = 26.84
Rolling hilly region in Northeast China [27]	V = 2.96 A0.97, R2 = 0.85
Mountainous and hilly region of Northeast China [28]	V = 0.18 A1.25, R2 = 0.90
Loess Plateau [29]	V = 1.71 A1.14, R2 = 0.85
Yuanmou dry–hot valley in Southwestern China [30]	V = 3.24 L1.27, R2 = 0.77
The upper Blue Nile basin, Ethiopia [31]	V = 1.05 A1.15, R2 = 0.95

), indicating a slower rate of gully incision per unit area of gully cross-section in the rolling hilly region. The gully development in the rolling hilly region may be mainly characterized by advancing gully heads and widening gully banks, while gully incision is the dominant process in the mountainous and hilly region. The b value in the study area is similar to that in the Loess plateau of China and to Ethiopia.

4.4. Significance and Limitation of This Study

In summary, this study utilized manual field measurements to gain insights into the actual conditions of gullies developed in forested areas of Northeast China. The findings provide valuable references for local gully control and soil management. Additionally, we built a V–A gully volume model, so that gully volume could be calculated using the gully area obtained by a combination of field measurement and remote sensing interpretation. Moreover, the main factors influencing gully erosion in forested areas have been identified as terrain, soil, climate, vegetation, and human activities, providing critical parameters/factors for the development of gully erosion models in future studies. However, there is a limitation in this study. Although field investigations are indeed effective in problem-solving, they still involve some degree of error and can be time consuming and labor intensive. In the future, a combination of manual and high precision measurement technology can be employed to enhance measurement accuracy. The freeze–thaw cycle has been proven to have a great influence on gully development and should be analyzed in future studies.

5. Conclusions

The aim of this study was to investigate the morphological characteristics and influencing factors of, and to develop an estimation model for, erosion volume in the forested areas of the Greater and Lesser Khingan Mountains (GKM and LKM). Based on field surveys, 37 gullies with an average length of 524.42 m, width of 2.45 m, depth of 1.09 m, area of 1243.93 m², and volume of 1153.18 m³ were found in a 114.87 km² area of the GKM, while 50 gullies with an average length of 560.13 m, width of 3.93 m, depth of 1.79 m, area of 2445.94 m², and volume of 3365.55 m³ were discovered in a 124.45 km² area of the LKM. Gullies in the LKM exhibit larger dimensions than those in the GKM. The GD of the two study areas is similar, but the GLD in the GKM is slightly higher than in the LKM. Overall, GD and GLD decrease with increasing slope. Gullies develop more quickly on the semi-sunny slopes of the GKM, while gullies in the LKM are distributed more evenly on different slope aspects. The volume estimation models for the GKM and LKM are V = 0.37 A^1.11 (R² = 0.94) and V = 0.32 A^1.17 (R² = 0.94), respectively.