Potential Reduction of Spatiotemporal Patterns of Water and Wind Erosion with Conservation Tillage in Northeast China

Abstract

Conservational tillage (NT) is widely recognized globally for its efficacy in mitigating soil loss due to wind and water erosion. However, a systematic large-scale estimate of NT’s impact on soil loss reduction in Northeast, China’s primary granary, remains absent. This study aimed to investigate the spatial and temporal variability of soil erosion under NT compared to conventional tillage (CT) in the black soil region and to analyze the underlying mechanisms driving these erosions. The Revised Universal Soil Loss Equation (RUSLE) and the Revised Wind Erosion Equation (RWEQ) models were employed, incorporating previously published plot/watershed data to estimate the potential reduction of water and wind erosion by NT in this region. Results indicated that under CT practices, water- and wind-induced soil losses were widely distributed in the arable land of Northeast China, with intensities of 2603 t km⁻² a⁻¹ and 34 t km⁻² a⁻¹, respectively. Furthermore, the erosive processes of water and wind erosion were significantly reduced by 56.4% and 91.8%, respectively, under NT practices compared to CT. The highest efficiency in soil conservation using NT was observed in the mountainous regions such as the Changbai Mountains and Greater Khingan Mountains, where water erosion was primarily driven by cropland slopes and wind erosion was driven by the wind speed. Conversely, the largest areas of severe erosion were observed in the Songnen Plain, primarily due to the significant proportion of arable land in this region. In the plain regions, water-induced soil loss was primarily influenced by precipitation, with light and higher levels of erosion occurring more frequently on long gentle slopes (0–3°) than on higher slope areas (3–5°). In the temporal dimension, soil loss induced by water and wind erosion ceased during the winter under both tillage systems due to snow cover and water freezing in the soil combined with the extremely cold climate. Substantial reductions were observed under NT from spring to autumn compared to CT. Ultimately, the temporal and spatial variations of soil loss under CT and NT practices were established from 2010 to 2018 and then projected onto a cropland map of Northeast China. Based on this analysis, NT is recommended as most suitable practice in the southern regions of Northeast China for maintaining soil health and crop yield production, while its suitability decreases in the northern and eastern regions.

1. Introduction

Soil erosion, a serious global issue, results in annual soil displacements of approximately 3590 and 58.3 billion tons by water and wind, respectively [1,2]. This erosion process leads to soil fertility loss, decreased productivity, increased flooding, and air pollution, posing significant threats to socio-ecological sustainability [3,4]. China’s vast drylands, the world’s largest, are significantly influenced by water and wind erosion, affecting 37.27% and 34.39% of these regions, respectively [5,6]. Specifically, in the Northeast Plain of China, the nation’s primary granary and the world’s fourth-largest region of black soil (Mollisols), soil degradation due to wind and water erosion presents a significant threat to soil health, national food security, and agricultural sustainability. The multi-year average soil water erosion modulus and amount were estimated at 1.52 t ha⁻¹ and 1.88 × 10⁸ t, respectively, from 2000 to 2020 across Northeast China [7]. Similarly, the estimated values of wind erosion were 1.13 t ha⁻¹ and 1.4 × 10⁸ t during the same period [7]. These erosional processes have led to a significant depletion of fertile black soil thickness, with a mean erosion rate of 2.22 mm/a, which would result in complete erosion in approximately 113 years [8]. Consequently, compared with forest, grassland, wetland, and urban land, soil erosion were observed to be more severe in the cropland in the black soil region in the Northeast China. Therefore, it is urgent need to develop sustainable soil management strategies to mitigate soil erosion while maintaining black soil thickness in this region.

However, soil erosion in the arable lands of Northeast China has been exacerbated by traditional tillage practices, including autumn plowing, deep ploughing, rotary tillage, and ridge tillage [1,9]. Annually, several tillage practices are carried out on the arable land before planting. Soil clods are often crushed into smaller pieces, and soil roughness is reduced due to mechanical modifications with rotary tillers or moldboard plows, resulting in a loosened soil structure [9]. The crop straw and residue are frequently collected and used for fodder or fuel or sometimes buried in the field under traditional agriculture. Soil organic carbon is therefore significantly decreased with the duration of traditional tillage [10]. Consequently, sheet erosion is likely to increase under traditional tillage due to the loosened and bare soil surface [11]. Conservation tillage, i.e., no tillage with straw mulching (NT), is widely utilized to mitigate soil loss from wind and water erosion on a global scale [1]. Several studies have demonstrated the positive effects of NT practices on Northeast China’s soil erosions. For instance, Zhang et al. [12] found a reduction in runoff by 71% to 98% and sediment loss by 89% to 99% under conservation tillage compared to conventional tillage (CT) in Hailun City. Wang et al. [13] documented an 82.6–85.5% decrease in runoff and an 87.9–91.5% reduction in soil erosion under NT compared to CT at the Lishu station. Jia et al. [14] and Chen et al. [15] demonstrated a reduction in soil wind erosion under NT in Changchun and Hailun. However, current studies in Northeast China primarily focus on plot monitoring or small watersheds, necessitating caution when advocating for widespread adoption. These studies do not provide a comprehensive understanding of soil erosion reduction across different regions and scales. Additionally, the suitability range of NT practices in the black soil region of Northeast China remains underexplored. Therefore, a systematic and large-scale estimate of the reduction in soil loss attributed to NT in Northeast China is essential. This research aims to fill these gaps by providing a comprehensive assessment of NT’s effectiveness in mitigating soil erosion across various spatial and temporal scales.

The models of soil erosion predicting are invaluable tools for assessing the potential impact of NT and CT on regional scale. Various models have been developed. For instance, the Water Erosion Prediction Project (WEPP) is designed to simulate continuous and spatial changes in land use and soil properties across different slope and watershed scales [16,17]. The Limburg Soil Erosion Model (LISEM) operates on individual erodible-rainfall events to simulate soil loss at a watershed scale [18]. Nevertheless, obtaining the model’s input indicators requires intensive field monitoring, posing a significant challenge. The European Soil Erosion Model (EUROSEM) is a highly accurate soil water erosion model capable of simulating sediment, runoff, and storm hydrographs in European contexts [19]. However, its applicability in Northeast China requires parameter adjustments to account for local soil and climate properties. In contrast, the Reversed Universal Soil Loss Equation (RUSLE) from America is the most widely used worldwide for assessing water erosion due to its robustness, ease of application, and adaptability to various geographic and climatic conditions [20]. In Northeast China, Zhang et al. [21] refined the RUSLE model using numerical data from 31 runoff test areas, 9 years of monitoring, and information from 20 weather stations. This refinement achieved a high accuracy of 0.9 in matching prediction data with measured data, demonstrating its reliability and effectiveness in the region. For wind erosion, the enhanced version of the Revised Wind Erosion Equation (RWEQ), derived from the Wind Erosion Equation (WEQ), is the most frequently used model in mainland China compared to others such as the Wind Erosion Prediction System (WEPS), Texas Erosion Analysis Model (TEAM), Wind Erosion Stochastic Simulator (WESS), Wind Erosion on European Light Soils (WEELS), Wind Erosion Assessment Model (WEAM), and the Integrated Wind Erosion Modelling System (IWEMS) [22]. In Northeast China, the RWEQ model has been shown to effectively predict wind erosion rates, making it suitable for assessing the wind erosion potential in the black soil zone due to its ability to account for the complex interactions among soil, vegetation, and climatic factors influencing wind erosion [2,3]. Therefore, the RUSLE and RWEQ models were selected to estimate the soil loss induced by water and wind erosion in the present study.

In the RUSLE/RWEQ modelling approach, both natural and human influences are integrated through various quantitative factors, including climate variables (such as rainfall, wind speed, and snowfall), landform characteristics, soil roughness, vegetation cover, and tillage practices. Many of these factors are based on findings from previous studies, as detailed in the following methods section. Specifically, the P factor in the RUSLE model, which reflects soil loss reduction due to human management practices, is calculated as the ratio of soil loss observed in control experiment plots without protective measures to that observed in plots with conservation management, as described by Zhang et al. [21]. To simulate the spatial variation of the P factor under CT and NT practices in Northeast China, a dataset was compiled using annual soil loss amounts obtained from field observations documented in the peer-reviewed literature for the present study. Additionally, a new soil freezing factor (F) was developed in the present study, defined as the absence of soil loss induced by wind and water erosion under frozen conditions. This factor is specifically tailored for Northeast China, characterized by extremely low soil and air temperatures throughout the winter months each year. The equation for the FE factor is shown below. Finally, the spatial and temporal variability of soil erosion under CT and NT from 2010 to 2018 was projected onto a cropland map of Northeast China, and then the underlying mechanisms driving these erosions were analyzed in this study.

2. Materials and Methods

2.1. Study Area

The study investigates the black soil region of Northeast China (115°31′–135°05′ E, 38°43′–53°33′ N), covering an expansive land area of 1.25 × 10⁶ km², comprising Heilongjiang, Jilin, and Liaoning provinces and the eastern segment of Inner Mongolia (Figure 1). Subsequently, this area is subdivided into five sub-regions based on distinct climatic and ecological characteristics: the Greater Kinggan Mountain, the Songnen Plain, the Liao River Plain, the Changbai Mountains, and the Sanjiang Plain regions, as depicted in Figure 1. Specifically, only arable lands that adopt conventional or conservation tillage practices and cultivate major crops such as soybean, maize, and wheat are estimated in the preset study, while paddy soils are excluded due to their lower soil erosion. Furthermore, the distribution of dryland in Northeast China is predominantly concentrated in the Songnen Plain, followed by the Greater Khingan Mountains regions, with statistical areas of 114.55 × 10³ km² and 79.26 × 10³ km², respectively, in 2018. The Changbai Mountains contain approximately 48.61 × 10³ km² of dry land, ranking third in terms of area. Ultimately, the dryland areas in the Liao River Plain and Sanjiang Plain are comparatively lower, with statistical areas of 37.87 × 10³ km² and 32.96 × 10³ km², respectively (Figure 1).

2.2. Data Resource and Processing

Data on the daily rainfall, air temperature, wind speed, and snow cover of China mainland were downloaded from the China Meteorological Data Service Center (http://data.cma.cn/ accessed on 14 April 2024). Data on the soil sand, clay, silt, organic carbon contents, and cropland data of the Northeast China were obtained from the Resource and Environment Science and Data center (https:/www. resdc.cn/Default.aspx, accessed on 14 April 2024). Soil slope and surface 30 m resolution digital elevation mode (DEM) data were downloaded from the Geospatial Data Cloud (https://www. gsclod accessed on 14 April 2024). The monthly normalized difference vegetation index (NDVI) products came from the National Science Data Center (http://www.csdata.org/ accessed on 14 April 2024). Moreover, a local erosion dataset of different tillage practices was created by extracting the soil loss amount from water and wind erosion under CT and NT practices from 58 peer-reviewed literature [12,13,14,15,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76] sources in the Northeast China. These papers were collected from various sources including Web of Science, Google Scholar, Springer Link, Scopus, and China Knowledge Resource Integrated Database (CNKI) during the period of 2010 to 2018. Subsequently, all the datasets were normalized to a standard resolution of 1 km² using the “GeoPandas” package within the Python software (version 3.7) for further analysis.

2.3. RUSLE Model

The water-induced soil erosion amount (E, expressed in Mg ha⁻¹ yr⁻¹) under conventional tillage (CT) and conservation tillage (NT), separately, are estimated using the Revised Universal Soil Loss Equation (RUSLE; Equation (1)) by multiplying seven factors ad follows [77]:

$$E=R·K·L·S·C·F·P$$

(1)

Here, R is calculated as follows [24]:

$$Ri=a\sum _1^{n}D{j}^b$$

(2)

Ri represents the rainfall-runoff erosivity force during the half month of i, measured in (MJ mm)/(ha⁻¹ h⁻¹); Dj denotes the daily rainfall (mm) on day j within a specific half month period, with days receiving over 12 mm of rainfall noted, and otherwise the value is noted as 0; n represents the count of days with rainfall exceeding 12 mm in each half-month period; parameters a and b are directly derived from Zhang et al. [78], who computed these values for all of mainland China using nonlinear regression analysis, minimizing the sum of squared residuals using the Levenberg–Marquardt method. The 15th day of each month is chosen to divide one month into two half-months, resulting in 24 half-months per year.

$$\begin{array}{c}K=\left(0.2+0.3\exp \left(-0.0256Sa\left(1-{\displaystyle \frac{Si}{100}}\right)\right)\right)·{\left({\displaystyle \frac{Si}{Ci-Si}}\right)}^{0.5}·\\ (1-{\displaystyle \frac{0.25SOC}{SOC+\exp \left(3.72-2.95SOC\right)}})·(1-{\displaystyle \frac{0.75Si}{Si+\mathrm{e}\mathrm{x}\mathrm{p}(-5.51+22.95Si)}})\end{array}$$

(3)

The K factor represents the soil erodibility factor for water erosion; the Sa, Si, and Ci denote the soil sand, silt, and clay content, respectively, expressed as percentages. SOC indicates the soil organic carbon content, which is also expressed as a percentage.

$$L={(l/22.13)}^{\alpha }$$

(4)

$$\alpha =\beta /(\beta +1)$$

(5)

$$\mathsf{\beta }=(sin\theta /0.0896)/[3.0{\left(sin\theta \right)}^{0.8}+0.56]$$

(6)

$$\mathrm{S}=\left\{\begin{array}{c}10.8 sin\theta +0.03 \theta <9\%\\ 16.8 sin\theta -0.50 9\%<\theta <14\%\\ 21.9 sin\theta -0.96 \theta >14\%\end{array}\right.$$

(7)

L and S factor denote the slope length and steepness factors, respectively, with l representing the slope length measured in meters and calculated from DEM using the following equation: DEM/ sin(θ*π/180). α and β represent the model parameters for slope length and steepness, while θ signifies the slope angle or gradient, expressed as a percentage.

$$\mathrm{C}=\mathrm{e}\mathrm{x}\mathrm{p}[-\mathsf{\alpha }·{\displaystyle \frac{NDVI}{\beta -NDVI}}]$$

(8)

The C factor indicates the effect of crop covering on soil water erosion, represented by a dimensionless value ranging from 0 to 1. The values of α and β, set at 2 and 1, respectively, are derived from Li et al. [79]. The Normalized Difference Vegetation Index (NDVI) represents the fractional vegetation cover calculated from remote sensing data acquired by MODIS.

$$\mathrm{F}=\left\{\begin{array}{c}1 P(T>0℃)\\ 0 P(T<0℃)\end{array}\right.$$

(9)

The soil freezing factor (F) has been introduced in the present study to account for the water-induced soil erosion halted by soil water freezing in the black soil region of Northeast China, particularly during winter rainy periods. The variable P(T < 0) takes on the values of 0 or 1, indicating whether the air temperature is above or below 0 degrees Celsius, resulting in a corresponding value of 0 or 1 for F. A value of 0 for F signifies no soil erosion, while a value of 1 indicates the absence of freezing effects.

$$\mathrm{P}={\displaystyle \frac{A_{NT}}{A_{CT}}}$$

(10)

The P factor signifies the protective effect of conservation tillage practices against soil water erosion in the field. Here, A_NT represents the soil erosion amount in NT practices, while A_CT represents the soil erosion amount in CT practices.

2.4. RWEQ Model

Soil wind erosion under CT and NT systems was computed utilizing the Revised Wind Erosion Equation (RWEQ) model [80], which integrates empirical and process-based components, considering seven influencing factors. The model’s formulas are provided below:

$${\mathrm{E}}^{\prime }={\displaystyle \frac{2z}{{Qs}^2}}{Q}_{max} e^{-{\left({\displaystyle \frac{z}{Qs}}\right)}^2}$$

(11)

$$\mathrm{Q}\mathrm{S}=150.71{(WF·K^{\prime }·SCF·RN·F^{\prime }·C^{\prime }·P^{\prime })}^{-0.3711}$$

(12)

$$Q_{max}=109(WF·K^{\prime }·SCF·RN·F^{\prime }·C^{\prime }·P^{\prime })$$

(13)

where E′ represents the wind-induced soil loss amount (E′, measured in units of t ha⁻¹ a⁻¹); QS denotes the critical field length for determining wind erodibility in specific areas, with a unit of meters (m); Q_max signifies the maximum transport capacity of soil sediment by wind erosion in a field (kg m⁻¹); and Z represents the downwind maximum distance at which the maximum wind erosion occurs, with a fixed value of 50 m in the present study.

$$\mathrm{W}\mathrm{F}=W_f· {\displaystyle \frac{\rho }{g}} ·SW·SD$$

(14)

$$W_f={\displaystyle \frac{{\sum }_1^n{v}_2{(v_2-v_1)}^2}{n}} · n_d$$

(15)

$$\mathrm{S}\mathrm{W}={\displaystyle \frac{{ET}_p-(R+I)·\frac{R_d}{N_d}}{{ET}_p}}$$

(16)

$$\mathrm{S}\mathrm{D}=1-\mathrm{P}(H_{snow}>25.4\mathrm{m}\mathrm{m})$$

(17)

WF is the weather factor component computed from the wind forces (W_f), soil wetness index (SW), and the snow cover depth (SD). Here, ρρ and g represent the air density and the gravitational acceleration, respectively, with fixed values of 1.29 kg m⁻³ and 9.81 m s⁻², respectively; n is the number of wind-erosion events in each month; N_d is the total number of these events; v₁ is the threshold wind speed set at the standard of 5 m s⁻¹ in this study; and v₂ represents the observed wind speed at a height of 10 m (m s⁻¹). ET_p represents the potential evapotranspiration, while R and I represent the daily rainfall amount (mm) and irrigation (mm), with Rd and Nd denoting the number and total number of rainfall and/or irrigation days within each month, respectively. In this study, the I was set to 0 mm given that Northeast China is a region reliant on rainfall, without the presence of irrigation practices. Furthermore, P(H_snow > 25.4 mm) represents the probability of the snow depth exceeding 25.4 mm.

$${\mathrm{K}}^{\prime }={\displaystyle \frac{29.09+0.31Sa+0.17Si+0.33\left(\frac{Sa}{Ci}\right)-2.59SOM-0.95CaC{O}_3}{100}}$$

(18)

$$\mathrm{S}\mathrm{C}\mathrm{F}={\displaystyle \frac{1}{1+0.0066{Ci}^2+0.021{SOM}^2}}$$

(19)

K′ and SCF represent the soil erodibility factor for wind erosion and the soil crust factor, respectively, where Sa, Si, and Ci denote the soil sand, silt, and clay content, respectively, as mentioned previously. SOM is calculated from SOC, while CaCO₃ is not considered here.

$$\mathrm{R}\mathrm{N}=e^{(1.86{RN}_r-2.41{RN}_r^{0.94}-0.127C_{rr})}$$

(20)

$${RN}_r=0.2 {\displaystyle \frac{{RH}^2}{RS}}$$

(21)

$$C_{rr}=17.46 {RR}^{0.738}$$

(22)

Here, RN represents the soil roughness factor (dimensionless), with RNr indicating ridge or oriented roughness, Crr representing chain random roughness, and RR denoting aggregate roughness. RH and RS represent ridge height (cm) and ridge spacing (cm), respectively, as derived from the DEM in Northeast China.

Furthermore, the F′ and P′ factors for wind erosion as affected by soil freezing and conservation tillage protection are calculated following similar formulations as the water erosion equations as mentioned previously in Equations (9) and (10).

$${\mathrm{C}}^{\prime }=e^{-0.043·SC}$$

(23)

$$\mathrm{S}\mathrm{C}={\displaystyle \frac{NDVI-{NDVI}_{min}}{{NDVI}_{max}-{NDVI}_{min}}}$$

(24)

Similarly, the C′ factor denotes the influence of crop covering on wind erosion, using the NDVI data mentioned previously in Equation (8).

2.5. Subsection

The Soil Erosion Classification Standard (SL190–2007) from the People’s Republic of China was employed to categorize the simulated water and wind erosion into several levels for evaluating the black soil erosion risks under CT and NT practices. These classifications include tolerable, slight, moderate, severe, very severe, and destructive erosion, defined by thresholds of 0–200, 200–2500, 2500–5000, 5000–8000, 8000–15,000, and (>15,000) t km⁻² a⁻¹, respectively.

2.6. Model Accuracy

To assess the accuracy of the RUSLE and RWEQ models in this study, linear regression was performed between the simulation results and the measurement data obtained from the publications included in the tillage dataset as mentioned above. Additionally, the root mean square error (RMSE), mean absolute error (MAE), and R-square between the simulation and measurement were calculated using the equations from Wang et al. [81].

3. Results

3.1. Model Factors

The results derived from the RUSLE and RWEQ models are illustrated in Figure 2 and Figure 3. The R factor demonstrated a significant range from 167.2 to 7896.4 MJ mm hm⁻² h⁻¹ a⁻¹ across the entirety of the black soil region, particularly pronounced in the southern parts of Northeast China (Figure 2A). The K factor was low in the Songnen plain but high in the Liao River plain, corresponding well to the spatial pattern of soil clay content [82]. Furthermore, soil L and S factors are generally greater in mountainous than flat plain regions. The C factor notably decreased from 1 to 0.1 during May to June within arable land during the growth of crops and then stabilized until October when crop harvest (Figure 2c and Figure S1). The F factor remained at 0 throughout the entire winter and transitioned to 1 as soil ice melted in March each year (Figure S2). Ultimately, the P factor of NT increased from 0.10 to 0.21 spatially, but temporal evaluation was hindered by data scarcity (Figure 2F). For comparison, the P factor of CT was set to 1 in the RUSLE model.

In the black soil region, the distribution of wind forces (W_f) in the RWEQ model were highly scattered (Figure 3A). Moreover, soil SW, surface SD, and F′ factors varied across different areas, weather conditions, and/or seasons (Figure 3B,C,G, and Figure S1). Soil K′ factors were higher in the north and east, while the SCF factor was greater in the west and south areas, with soil RN factor being higher in the central regions (Figure 3D–F). Similarly, the P′ factor for wind erosion was set as 1 for CT and varied from 0.99 to 0.085 due to increased straw covering and reduced soil disturbance in NT over the months (Figure 3H and Figure S4).

3.2. Spatial Patterns of Soil Erosion as Affected by NT Practice

Soil water and wind erosion were widely distributed in the arable land of Northeast China (Figure 4). The regional average estimated 2603 t km⁻² a⁻¹ of water erosion and 34 t km⁻² a⁻¹ of wind erosion across the entire black soil region under CT practices in 2018 (Figure 4A,D;

Table 1. Average soil loss of different conservation regions under CT and NT practices in the black soil regions in Northeast China.

Regions	Water Erosion (t km−2 a−1)			Wind Erosion (t km−2 a−1)
	CT	NT	Reduction	CT	NT	Reduction
I	2196	999	1206	70	6	64
II	2279	829	1457	24	2	22
III	2327	821	1508	40	3	37
IV	4370	2470	1895	7	1	6
V	2454	959	1489	13	1	12
Total	2603	1134	1511	34	2.8	28.2

). These values were reduced by 56.4% and 91.8% with NT practices, respectively (Figure 4B,E; Table 1). Consequently, this reduction was more pronounced in the Changbai Mountains for water erosion and in the southwest of the Greater Khingan Mountains for wind erosion, where the intensity of water and wind erosion was most evidently observed corresponding (Figure 4C,F; Table 1). In the Changbai Mountain, soil areas of water erosion classified as very severe (8.03 × 10³ km²) and severe (7.62 × 10³ km²) under CT practices thereby reduced to only 2.31 × 10³ km² and 5.61 × 10³ km² under NT practices, respectively (Figure 5A). Meanwhile, the areas of wind erosion in the Greater Khingan Mountains shifted from 1.59 × 10³ km² (class moderate) and 4.31 × 10³ km² (class slight) under CT to 0 km² under NT (Figure 5B).

Considering area, the largest water and wind erosion areas were observed in the Songnen plain, which encompasses the majority of arable land in Northeast China (Figure 4 and Figure 5). Here, soil erosion caused by water was significantly higher than that caused by wind, with the highest water-induced erosion areas predominantly distributed in the northeast part of the plain (Figure 4). Similarly, NT practice significantly reduced high-level erosion. Under NT, areas previously subjected to soil water erosion classified as moderate, severe, and very severe under CT decreased to tolerable and slight levels, with the areas of very severe reduced to 0 km² (Figure 5A). The average change in soil amount caused by water erosion therefore declined from 2327 t km⁻² a⁻¹ in CT to only 821 t km⁻² a⁻¹ under NT in this plain (Table 1).

In the Liao River and Sanjiang plains, both the intensity and areas of water erosion were estimated to be low, and there was no obvious wind erosion observed (Figure 4 and Figure 5; Table 1). However, the potential reduction of water erosion caused by NT practice was still significant in specific areas within these regions (Figure 4C). Furthermore, detailed soil erosion amounts and areas under CT and NT across the other arable land of Northeast China are presented in Figure 5 and Table 1.

3.3. Temporal Dynamics of Soil Erosion as Affected by NT Practice

The average annual and monthly soil loss amount was computed from 2010 to 2018 in this study (Figure 6 and Figure 7). The annual average of soil loss amount due to water erosion remained relatively stable over these nine years, while the values for wind erosion fluctuated sharply, decreasing, increasing, and then decreasing again under CT treatments (Figure 6). Consequently, the monthly average soil loss amount due to water erosion under CT and NT was near zero t km⁻² m⁻¹ at the beginning (January, February, March) and the end (November, December) of each year in Northeast China (Figure 7A). It initially increased from 39.4 ± 21.2 t km⁻² m⁻¹ to 165.7 ± 44.0 t km⁻² m⁻¹ from April to July and then steadily decreased to 0 by November under CT practices. From April to October, soil water erosion was proportionally reduced by NT compared to CT. The greatest reduction was observed from June to August, with lower reductions in May, September, and October. For wind erosion, soil loss was observed annually under CT from March to May, with a predominance in the latter two months. Furthermore, this wind erosion was reduced by 91.5% under NT practices in each month (Figure 7B).

3.4. Influence of Various Factors on Soil Erosion as Affected by NT Practice

In Northeast China, water erosion was mainly influenced by precipitation, while wind erosion was driven by wind speed (Figure 8). Under CT and NT practices, precipitation contributed 24.9% and 30.9% to soil water erosion, respectively, while wind speed contributed 26.5% and 26.9% to soil wind erosion, respectively. Consequently, field slope, soil clay content, and local aridity index significantly influenced soil loss under water erosion in both CT and NT practices (Figure 8A). Wind erosion, on the other hand, was notably influenced by arid conditions, soil silt content, field length, and local precipitation under both tillage practices in the same regions (Figure 8B).

4. Discussion

4.1. Model Validation

Our models performed well in simulating the variation of soil wind and water erosion under CT and NT practices in the dry lands of Northeast China, with credible model evaluation indices of R² (0.65, 0.56, and 0.66), MAE (891, 409, and 151), and RMSE (1277, 916, and 268) for the RUSLE-CT, RUSLE-NT, and RWEQ-NT models, respectively (

Table 2. Performance of the RUSLE and RWEQ models as evaluated by the correlation between the simulated and measured data (collected from the publications).

Erosion	Tillage	Number	The Relationship between the Simulated and Observed Data	MAE	RMSE	R2
Water	CT	44	y = 0.2377× −3.6909	891	1277	0.65
	NT	39	y = 0.0772× +53.263	409	916	0.56
Wind	CT	5	y = 1.1519× +110.45	262	425	0.15
	NT	9	y = 0.7060× −10.943	151	268	0.66

). However, the correlation between simulated and measured soil loss under wind erosion in the RWEQ-CT model is poor (R²: 0.15, MAE: 262, RMSE: 425; Table 2), primarily due to the limited number of verification data and their concentrated distribution, with only five observations mainly in eastern Inner Mongolia.

Furthermore, our simulated water erosion values under CT across the cropland of Northeast China, at light and higher levels, covered an area of 182,265.52 km² in 2015. This value closely aligns with the 181,130.18 km² reported in the National Soil Erosion Investigation of China (NSEI-C) conducted by the Chinese Academy of Sciences (CAS) in 2015, using an integrated evaluation method [83]. However, our estimation of wind erosion under CT for the same year and levels amounted to only 74,223.85 km², which is significantly lower than the 99,210.43 km² reported by the NSEI-C. This disparity may be attributed to the fact that our research focused solely on cropland, while their study included a broader range of land uses. Moreover, the spatial patterns of wind and water erosion distribution under CT practice closely resembled those observed in several previous studies [7,83,84]. We failed to compare the NT results with those of other models due to the lack of corresponding predictions in the current regions.

4.2. Spatial Patterns of Soil Erosion as Affected by NT Practice

NT practices in the mountainous regions (Changbai Mountains, Greater Khingan Mountains) demonstrated the highest efficiency in conserving soil water and wind erosion (Figure 4). This can be primarily attributed to the fact that these areas are among the most severely eroded regions in Northeast China. Furthermore, precipitation and cropland slope contributed the most at approximately 30.9% and 26.5%, respectively, to soil water erosion under NT practices on the spatial scale (Figure 7). Precipitation primarily drove soil runoff in croplands [78]. However, in the Changbai Mountains regions, the R factor of the rainfall and runoff erosivity force was not the highest (Figure 2A), indicating that the steeper slopes of croplands in these areas were the main reason for more severe soil water erosion (Figure 2C). The intensity of water erosion significantly increased with soil slopes from 5° to 18° under CT in the mountainous regions (Figure 9), attributed to increased runoff with steeper slopes during rainfall events [85]. Additionally, the loosened soil structure, breakdown of soil aggregates, and removal of residue coverage by mechanical till disturbance exacerbated soil loss induced by water runoff in the mountains areas [60]. In contrast, under NT practices, soil mulched by straw residue and soil structure stability improved due to continuous non-disturbance from no-tillage practices [86]. Consequently, surface sealing formation declined and water infiltration increased, resulting in a significant reduction of surface runoff and providing a mechanical barrier to soil displacement [87]. Subsequently, this cover under NT also reduced wind-induced soil erosion in the Greater Khingan Mountains regions, where the most severe wind erosion occurred due to high wind speeds and dry climate conditions [88]. Notably, the success of NT in soil erosion conservation in mountainous regions does not imply its failure in plain regions, where the substantial reduction gap observed in erosion reduction was due to the highest initial erosion values occurring under CT practices.

In the plain areas such as the Songnen, Liao River, and Sanjiang plains, the regions experiencing wind erosion at light and higher levels (defined as noticeable soil loss) were estimated to be near zero under both CT and NT practices, owing to the absence of strong wind events in these areas (Figure 5). This finding is consistent with Wang et al. [81]. Furthermore, water erosion at light and higher levels was widespread and significantly reduced by NT compared to CT practices (Figure 4), particularly in the Songnen plain, which has the largest area of arable lands in Northeast China. Furthermore, severe (light and higher) water erosion was observed more frequently in small arable land slopes of 0~3° than in higher slope areas of 3~5° in the plain regions (Figure 9), consistent with the findings of Wang et al. [30]. This suggests a pronounced pattern of precipitation driving soil loss under the topographical features of long gentle slopes [85].

4.3. Temporal Dynamics of Soil Erosion as Affected by NT Practice

Annually, soil water and wind erosion under CT and NT fluctuated with local climate variations (rainfall and wind speed, Figure 6), as well-documented [60,84,89]. Interestingly, on a monthly basis, soil wind and water erosion cease in January, February, November, and December each year in the arable land of Northeast China (Figure 7). This cessation is attributed to the snow covering and soil water freezing under the extremely cold climate in this area. The seasonal average surface air temperature was lower than −15 °C during the winter from 2010 to 2018 in this study (data not shown), resulting in soil freezing to depths of over 200 cm per year, as observed by Xu et al. [90] in Heilongjiang province. This result highlights the significant benefit of naturally lower temperatures in reducing black soil loss through erosion during the long winter period each year.

Consequently, this frozen state of the soil water continues until early March each year (Figure S2), and then the ice begins to melt from the south to the north of Northeast China within this month (Figure 6). Thus, slight wind erosion occurs in March, particularly in the early warmer areas of southern Northeast, but there is no water erosion due to lack of rainfall during this period [91]. Furthermore, black soil experiences concentrated wind erosion in April and May each year under CT practices, primarily due to high levels of soil disturbance and the removal of residue cover. Conversely, these highest wind erosion rates were significantly reduced to near zero by NT, owing to mulching and the preservation of undisturbed soil structure [14]. This finding suggests that the greatest benefits of conservation tillage on mitigate soil loss by wind erosion occur in April and May, preceding the planting season each year. This provides valuable information for directing focus toward wind erosion conservation efforts during these two months.

Subsequently, soil water erosion under CT sharply increases from May to June, peaking in July, which aligns with the maize growth season. From August to September, black soil loss due to water erosion diminishes. This reduction is attributed to the dense canopy of maize, which mitigates the direct impact of raindrops on the soil surface [92] and extends the time taken for water to reach the soil, thereby reducing surface runoff [87,93]. During the entire process, soil water erosion was proportionally reduced by NT practices (Figure 6). However, we were unable to provide detailed insights into the performance of NT on water erosion within monthly dimensions due to the lack of supervisory data. This underscores the necessity for further measurements and research in the future.

4.4. The Implementation of NT Practice in Northeast China

Overall, NT practice shows enormous potential in reducing soil erosion and preserving the ecological benefits of black soil. However, the current adoption of conservation tillage in black soil regions of China remains limited [79,94], possibly due to the yield decrease associated with NT [95]. Therefore, striking a balance between economic gains and ecological benefits in this region will be a worthwhile issue to address. Based on the results of this study, specific recommendations for NT practices are as follows: (i) NT with straw mulching is most suitable in the southern regions of the Greater Khingan Mountains and Songnen Plain, the entire Liao River Plain, and the Changbai Mountains, where it demonstrates the highest efficiency for conserving soil erosion (water and wind), and it has been shown to achieve higher yields according to our previous study [96]. (ii) Caution is needed when implementing NT in the northern regions of the Songnen Plain and the Sanjiang Plain. In these areas, the reduction of soil erosion was minimal (Figure 4), and yields of NT were reduced [96]. (iii) Continuous monitoring of soil erosion and crop yields under NT practices is essential. The data collected should be used to refine and adapt NT techniques to local conditions, ensuring that both ecological and economic benefits are maximized.

5. Conclusions

In Northeast China’s arable lands, conventional tillage (CT) leads to intense and widespread soil water and wind erosion. No tillage (NT) with straw mulching significantly reduces these erosive processes. Soil erosion ceases in winter under both systems, with NT showing a marked reduction in soil loss from spring to autumn compared to CT. Spatial distribution maps revealed that NT is most effective in conserving soil in mountainous regions. Based on existing data, NT is best suitable in the southern regions of Northeast China for maintaining soil health and crop yields, while its effectiveness diminishes in the northern and eastern regions.