Abstract
To explore the effect and applicability of contour reverse slope terraces in water and soil conservation of karstic red clay soil area in northern Guangxi, rainfall simulation experiments were conducted on bare red clay soil slopes laid with contour reverse slope terraces to find the erosion pattern in response to different rainfall intensities (15, 30, and 50 mm/h) and bedrock porosities (1%, 3%, and 5%). The results showed that contour reverse slope terraces can change the spatial distribution of surface runoff, divert surface runoff to the underground, and increase underground sediment yielding. It is found that contour reverse slope terraces can effectively reduce surface soil erosion though it still remains dominant erosion for red clay terraced farmland, with surface runoff reduced by 71%, 52%, and 46% and surface sediment loss reduced by 70%, 65%, and 63%, respectively, when the rainfall intensity is set at 15 mm/h, 30 mm/h, and 50 mm/h. It is concluded that contour reverse slope terraces can help reduce water and soil erosion of red clay terraced farmland in northern Guangxi.
1. Introduction
As the second largest carbon pool after the oceans and one of the most important natural resources, soil plays a crucial role in providing humans with vital goods and services to sustain life and regulate the flow of matter and energy between the lithosphere, biosphere, hydrosphere, and atmosphere [1]. In recent years, soil erosion and degradation have become a key environmental concern worldwide due to overexploitation and lack of effective management [2, 3].
The karst region distributed in Yunnan, Guizhou, and Guangxi in southwestern China is the typical karst region in the world [4, 5]. Pure carbonate rocks are widely distributed in this region, and the favorable climatic conditions and long-term karstification have given rise to various types of precipitous landscapes which make it hard to conserve soil with the continuous heavy rainfall of the subtropical monsoon climate [6]. Karstification has not only formed surface landforms such as peak clusters and stone buds but also underground landforms such as underground rivers and sinkholes, which creates a unique surface-underground dual spatial structure in the karst region [7], providing an underground “invisible” channel [8] for soil and water loss in this region that is difficult to observe and calculate. Rainwater forms runoff and carries sediment away from underground channels along pore fissures and sinkholes, causing nonzonal drought [9]. Moreover, there are very few acid-insoluble substances in carbonate rocks in karst regions, and it takes more than 40,000 years to form 1 cm thick layer of soil [10], which is 10~80 times longer than the soil formation in nonkarst regions [11]. Long-term soil erosion has led to vegetation degradation and bedrock exposure, causing karst rocky desertification, which is difficult to recover in a short period of time [2, 12]. Field investigation showed that rock exposure, soil thinning, and vegetation degradation appeared on some hilly slopes of karst areas. Therefore, it is of great scientific significance to further understand the erosion pattern of sloping farmland and provide effective control measures.
Due to the complexity and concealment of underground landforms in karst areas, the limitations of research methods and tools, and the lack of long-term positioning and monitoring tests, indoor simulation tests based on fieldwork are widely adopted. Chun-heng et al. [13] investigated the effects of soil macropore structure on soil and water loss processes in karst lime soil area through artificial rainfall experiments. Ziyang et al. [14] studied the surface and underground runoff yielding mechanism in the karst area of Chongqing city by simulating bedding/inverse layer slope characteristics and underground porosity. Chun-hua et al. [15] simulated different relative positions of rock in the soil layer and studied the soil erosion characteristics of different soil and rock mosaic slopes in karst loess area during successive rainfalls. Liu et al. [16] simulated natural rainfall and groundwater flow to study the effect of raindrop impact on sediment loss under free drainage and groundwater seepage conditions. There are also several simulation experiments of surface and underground erosion processes in response to various factors including rainfall intensity, bedrock porosity, slope, bedrock exposure rate, and rainfall duration [17–21].
Various soil and water conservation measures are commonly used currently. Tai-kui et al. [22] showed that the combination of slope-into-terrace and grass planting could effectively reduce soil and water loss on sloping land, and both were reduced by more than 30%. The simulation results of Xiong-fei et al. [23] indicated that side ditches worked well in reducing surface sediment loss, and the sediment reduction rate was positively correlated with rainfall intensity. Xing-yi et al. [24] pointed out that applying long-term straw mulching on black soil farmland could significantly reduce runoff and soil loss. Field experiments by Liu Gang et al. [25] revealed that the erosion control efficiency of upslope drainage technique in loess stairstep cut slopes was better than vegetation planting and deploying reverse slope and could achieve the balance between erosion control efficiency and cost. Several studies on contour reverse slope terrace have shown that it can not only effectively reduce surface runoff, sediment, and nutrient loss [26–28] but also improve the physiochemical properties and enzyme activity of soil [29, 30]. While most of the studies on contour reverse slope terrace have focused on red soil areas in Yunnan province, loess plateau areas, and northern soil and rocky mountain areas [31], there are few reports on its application and experiments in karst areas. Due to the special surface-underground dual spatial structure of karst areas, the underground soil and water loss process is hidden and complicated, and the existing soil and water conservation techniques in nonkarst areas are not necessarily suitable [2].
Therefore, rainfall simulation experiments were carried out in the present study by using artificial rainfall equipment and a self-made angle-adjustable wooden soil trough to explore the response of surface and underground soil erosion to different rainfall intensities and different bedrock in red clay bare slope farmland with contour reverse slope terrace under the condition of fixed slope gradient and rainfall duration. The soil and water erosion pattern of contour reverse slope terrace was compared with that of the ordinary bare slope to find out the soil and water conservation efficiency and applicability of contour reverse slope terraces in the karst red clay region and to provide a reference basis for soil erosion control and ecological restoration in karst regions.
2. Materials and Methods
2.1. Overview of the Study Area
The study area, located in Guilin, northern Guangxi (110°1848E, 25°1643N), has a subtropical monsoon climate, with an annual average temperature of 19.8°C and annual average precipitation of 1978.4 mm. Rainfall is concentrated in months from April to August, accounting for more than 60% of the annual precipitation. The rainy season, especially from April to July, is often accompanied by high temperatures. The arable land covers an area of about 3021.5 square kilometers, and the main crops are rice, sugar cane, citrus oranges, etc. Pure carbonate rocks are widely distributed in this region, with large-scale exposure on the surface of hillsides in the field.
2.2. Materials
The soil used in this experiment was red clay, which is derived from carbonate rocks with minimal acid nonsoluble substances weathered by the subtropical monsoon climate for years [32]. Red clay is mainly distributed in Guangxi, Yunnan, Guizhou, and other regions in southern China and typically found in Guangxi [33]. The soil sample was collected from within 20 cm of the surface layer of the sloping farmland in Gongcheng County, a karst area in northern Guangxi, with the organic matter content of 49.0 g/kg, the total nitrogen content of 13.7 g/kg, the total phosphorus content of 0.75 g/kg, the total potassium content of 2.56 g/kg, and pH value of 7.13. After the soil dried, large soil clumps were ground finer to a diameter of less than 5 mm, and impurities like gravel, plant roots, and garbage were removed.
The experiment equipment consisted of stainless steel bracket, a wooden soil trough, and artificial rainfall equipment. The wooden soil trough was fixed on the stainless steel bracket and could be steplessly rotated at the angle between 0 and 45° with the bracket. The wooden soil trough was divided into three identical rectangular parts along the longitudinal direction by two partitions. The internal dimensions of each part were 1.8 m long, 0.25 m wide, and 0.3 m deep, with a bottom area of 0.45 m2. Round holes were evenly punched at the bottom of each part in different numbers to simulate the pores and fissures on the surface of the bedrock, and bedrock porosity was taken to measure the extent of karstification of the bedrock. The wooden soil trough was equipped with a pipe for collecting surface slurry and a tank for collecting underground slurry, as shown in Figure 1.
The artificial rainfall equipment contained a water pump, switch, flowmeter, and rainfall sprinkler. The effective rainfall area was , which could cover the soil trough, and the rainfall intensity could change between 15 and 100 mm/h. The uniformity coefficient of the rainfall sprinkler was verified up to 85%, and the rainfall intensity was corrected by a rain gauge.
2.3. Experimental Design
The average gradient of the erosion survey unit in Guangxi is 18.49°, and the average gradient of arable land is 11.06° [34], and the proportion of arable land with gradient below 15° reaches 90.22%, so 15° was taken as the test gradient for its typicality.
Rainfall intensity and bedrock porosity are the two variables in this study. In terms of rainfall intensity, the results of Ying et al. [35] showed that the threshold of erosive rainfall amount () and maximum 30-minute rainfall intensity () for red soil in southern China are 6.1 mm and 5.2 mm/h, respectively. Wen-yuan et al. [36] indicated that of karst yellow soil ranges from 9.6 to 10.2 mm/h, considering the characteristics of subtropical monsoon climate with simultaneous rain and high temperature, abundant rainfall, and frequent rainstorm, along with the frequent occurrence of extreme precipitation due to global warming since the 20th century [37, 38]. Previous studies on karst area in Guizhou Province [19, 21] indicated that due to the spores and fissures in the bedrock, there is a critical rainfall intensity that will cause surface runoff in karst area, i.e., surface runoff only occurs when the rainfall intensity is higher than a specific value. Otherwise, there will be only underground runoff but no surface runoff, and it is also pointed out that the critical rainfall intensity in Guizhou calcareous soil area is between 30 and 50 mm/h. In summary, 15 mm/h, 30 mm/h, and 50 mm/h were taken as the test rainfall intensities. The adopted rainfall intensities fit the definition of heavy rainfall by the China Meteorological Administration (cumulative precipitation of 50 mm within 24 hours).
Bedrock porosity is the percentage of the hole area to the bottom area of the trough. It was set at the three levels of 1% (micro porosity), 3% (low porosity), and 5% (medium porosity) in this study, with the respective hole areas being 45 cm2, 135 cm2, and 225 cm2 and the respective hole numbers and diameters being 160 holes of 6 mm diameter, 478 holes of 6 mm diameter, and 447 holes of 8 mm diameter. The schematic of the wooden trough and the holes is shown in Figure 2.
A wooden soil trough was used to simulate the sloping farmland in the karst area. The steps were as follows: (1)Adjust the trough to the horizontal, fill it with the red clay prepared as mentioned above to a height of 15 cm, smooth the surface with a trowel, and compact the soil near the corners of the trough manually and slightly to reduce the boundary effect(2)Spray water evenly onto the soil surface with an atomization nozzle to wet the soil without damaging the surface. Stop spraying water when there is minor ponding on the surface and runoff in the underground collection tank, which indicates that the soil is saturated(3)Cover the trough with tarpaulin and leave it for at least 24 hours to ensure that the natural settlement has been completed, and then adjust the trough to 15° for subsequent rainfall simulation experiments
Based on the climatic characteristics of northern Guangxi, the duration of a single rainfall ranges from 15 to 30 minutes, so in this study, the duration of a single artificial rainfall was set to 20 minutes, starting from the time when the raindrops hit the soil surface. A total of 9 rounds of tests on the bare slope were conducted to find the erosion patterns of the bare slope in response to three different rainfall intensities (15, 30, and 50 mm/h) and three bedrock porosities (1%, 3%, and 5%). After each round, the rill produced by runoff scouring was filled again with red clay, and the same steps were repeated.
After the bare slope rainfall simulation tests were completed, two contour reverse slope terraces were laid on each bare slope, as shown in Figure 3, and the simulation rainfall experiment of the contour reverse slope terraces was carried out following the same steps. The shape of the single contour reverse slope terrace is shown in Figure 3(a), and the position of the two terraces is shown in Figure 3(b).
(a)
(b)
(c)
2.4. Specimen Collection and Data Processing
At the start of the rainfall, sealable plastic containers were placed at the outlets of the collecting pipe and collecting tank to collect slurry from the surface and underground. The containers were replaced every 2 minutes to measure the rate of runoff and sediment yielding in different periods. After the artificial rainfall, the slurry was transferred to glass containers, dried in an oven at 105°C to constant weight, and weighed on an electronic balance with an accuracy of 0.01 g.
In this study, multiple indicators were used to describe the experimental phenomenon and results. Surface cumulative runoff modulus () and underground cumulative runoff modulus () were adopted to measure the volume of surface and underground runoff yielded per square meter of the slope during a single rainfall event in ml/m2. Surface runoff proportion () and underground runoff proportion () were adopted to measure the proportion of surface and underground runoff to the total runoff, which were calculated as follows:
Like and , surface cumulative sediment modulus () and underground cumulative sediment modulus () were applied to measure the weight of sediment lost from the surface and underground per square meter of the slope during a single rainfall event in g/m2. Surface sediment proportion () and underground sediment proportion () were applied to measure the proportion of surface and underground sediment to the total sediment, which were calculated as follows:
3. Results and Discussion
3.1. Soil Erosion of the Bare Slope
3.1.1. Runoff Yielding of the Bare Slope
Rainfall is the main driving force of soil erosion. Raindrops converge into runoff on the surface. When runoff flows downhill, it carries soil particles away and forms rills. The pores and fissures produced by karstification on the bedrock and the caves and underground rivers below the bedrock constitute the channels for underground soil and water leakage. As shown in Table 1, both surface cumulative runoff modulus () and underground cumulative runoff modulus () were significantly influenced by rainfall intensity and positively correlated with rainfall intensity. While underground cumulative runoff modulus () increased with the rise of bedrock porosity, the surface cumulative runoff modulus () decreased as bedrock porosity increased.
Surface runoff proportion () followed the same pattern as surface cumulative runoff modulus (), positively correlated with rainfall intensity, and negatively correlated with bedrock porosity, and underground runoff proportion () showed the opposite.
More patterns can be obtained by analyzing the whole process of runoff rate. Runoff rate is the runoff volume per square meter of the slope per unit time in ml/(m2·min), divided into surface runoff rate and underground runoff rate according to the spatial location of runoff. Figure 4 shows the bare slope surface and underground runoff yielding process in light of different rainfall intensities and bedrock porosities.
(a) 1% bedrock porosity
(b) 3% bedrock porosity
(c) 5% bedrock porosity
Figure 5 indicates that regardless of rainfall intensities, both surface and underground runoff rates showed an overall trend of going upward and then stabilizing. When bedrock porosity increased from 1% to 3% at a low rainfall intensity (15 mm/h), both surface cumulative runoff modulus () and surface runoff proportion () decreased sharply, and when bedrock porosity increased to 5%, surface cumulative runoff modulus () increased a little, but surface runoff proportion () decreased again (Table 1), and at 3% and 5% bedrock porosities, the surface runoff yielding delayed significantly, which indicated that the 15 mm/h rainfall intensity is close to the critical rainfall intensity for causing surface runoff on red clay sloping farmland with 3% and 5% bedrock porosities [19].
(a) 1% bedrock porosity
(b) 3% bedrock porosity
(c) 5% bedrock porosity
The results of partial correlation analysis in Table 2 showed that rainfall intensity was highly and significantly positively correlated with surface cumulative runoff modulus () () and surface runoff proportion () () with a partial correlation coefficient of 0.973 and 0.769, respectively, and significantly negatively correlated with underground runoff proportion () () with a partial correlation coefficient of -0.769. Rainfall intensity was significantly positively correlated with underground cumulative runoff modulus () () with a partial correlation coefficient of 0.716.
Bedrock porosity was highly significantly negatively correlated with surface cumulative runoff modulus () () with a partial correlation coefficient of -0.92 and highly significantly positively correlated with underground cumulative runoff modulus () () with a partial correlation coefficient of 0.889. Bedrock porosity was significantly negatively correlated with surface runoff proportion () () with a partial correlation coefficient of -0.726 and was significantly positively correlated with underground runoff proportion () () with a partial correlation coefficient of 0.726.
3.1.2. Sediment Yielding of the Bare Slope
As shown in Table 3, due to bedrock pores and fissures, the soil would leak underground while being lost along the surface, but the weight and proportion of underground loss were much smaller than those of surface erosion. Both surface and underground cumulative sediment modulus increased with the increase of rainfall intensity, but surface sediment yielding far outweighed underground sediment yielding in the absolute value of weight and weight growth ratio, so surface sediment proportion () showed a trend of increasing with rainfall intensity, indicating that surface erosion is the dominant pattern of soil erosion on red clay bare slope, regardless of rainfall intensity.
Like the runoff rate in Figure 4, sediment rate describes the weight of sediment per square meter of the slope per unit time in g/(m2·min), which is divided into surface sediment rate and underground sediment rate according to the spatial location of sediment yielding. The process of surface sediment yielding on the bare slope in light of different rainfall intensities and bedrock porosities is shown in Figure 4.
Compared with the surface runoff rate in Figure 5, the surface sediment rate showed greater volatility. Sediment was partly formed by dissociative soil particles on the slope surface and partly produced by rill erosion. The rill kept developing throughout the rainfall, so the surface sediment rate tended to increase with time (Figure 6).
During the development of the rill, the following situations might arise to temporarily reduce the surface sediment rate: (i)Rill development encounters obstacles, such as soil clumps or stone particles that are too large to be carried by the runoff, and thus slows down or the rill is diverted(ii)Soil particles coming down with the runoff block the rill for a period
If the rill is diverted successfully or the soil and stone particles blocking the rill are finally carried away under the continuous impact of runoff and raindrops, the rill would develop smoothly again and increase surface sediment rate. The above situation may occur repeatedly in the cycle throughout the rainfall.
In addition, in the early stage of the rainfall, the surface sediment rate increased due to the development of the rill. In the middle stage, the surface sediment rate decreased to a certain extent because the raindrop hit the soil surface and compacted the soil to form a relatively strong crust, and the wash durability of the soil increased to a small extent. At the later stage, when the crust was gradually destroyed by the combined force of rill development and raindrops, the wash durability of soil weakened and the surface sediment rate increased again.
Therefore, under the combined effect of the above factors, the surface sediment rate in Figure 4 fluctuated a lot throughout the whole process.
The process of underground sediment yielding on the bare slope with different rainfall intensities and bedrock porosities is shown in Figure 7. The curves of underground sediment rate fluctuated in a small range. The seepage force of water in the soil is the driving force for the movement of soil particles. Due to seepage force, underground soil particles encountered internal scour, and fine soil particles moved through the pores and fissures with water flow. It was possible for them to fill the pores and fissures in the bedrock and the void in the soil so that only a very small number of extremely fine soil particles could pass through the pores and be carried away with the water, which explains the decrease in underground sediment rate as shown in the curve. Under the continuous action of water flow, the soil particles stuck in the pores and voids would be washed down, so that the underground leakage channel would be open again, resulting in the increase of underground sediment rate.
(a) 1% bedrock porosity
(b) 3% bedrock porosity
(c) 5% bedrock porosity
The results of the partial correlation analysis are shown in Table 4. Rainfall intensity was highly significantly positively correlated with surface cumulative sediment modulus () () with a partial correlation coefficient of 0.919. Rainfall intensity was significantly positively correlated with underground cumulative sediment modulus () () and surface sediment proportion () () with a partial correlation coefficient of 0.771 and 0.792, respectively, and significantly negatively correlated with underground sediment proportion () () with a partial correlation coefficient of -0.792.
Bedrock porosity was significantly negatively correlated with surface cumulative sediment modulus () () and surface sediment proportion () (), with a partial correlation coefficient of -0.831 and -0.786, respectively. Bedrock porosity was significantly positively correlated with underground cumulative sediment modulus () () and underground sediment proportion () (), with a partial correlation coefficient of 0.807 and 0.786, respectively.
3.2. Soil Erosion of the Bare Slope with Contour Reverse Slope Terraces
3.2.1. The Regulating Effect of Contour Reverse Slope Terraces on Runoff
The surface and underground cumulative runoff modulus and the percentages on the bare slope and contour reverse slope are shown in Table 5.
Compared with the bare slope, contour reverse slope terraces would significantly decrease surface runoff proportion () and significantly increase underground runoff proportion (), which tells that contour reverse slope terraces would change the spatial distribution of runoff in the surface and underground and promote the transfer of surface runoff to the underground.
The pattern of surface runoff of the contour reverse slope terraces was similar to that of the bare slope. Surface cumulative runoff modulus () and surface runoff proportion () increased with rainfall intensity and decreased with bedrock porosity. Underground cumulative runoff modulus () basically increased with bedrock porosity, but its relationship with rainfall intensity was not obvious.
The effect of contour reverse slope terraces in regulating the spatial distribution of runoff is manifested as a decrease in surface runoff and an increase in underground runoff, as shown in Table 6, where the regulating ratio is calculated based on the bare slope data.
Table 6 shows that the contour reverse slope terrace was effective in reducing surface runoff at different rainfall intensities, and the effect was particularly strong at low rainfall intensity (15 mm/h). At different rainfall intensities, the average regulating ratio of surface runoff was, respectively, 71% (15 mm/h), 52% (30 mm/h), and 47% (50 mm/h), negatively correlated with rainfall intensity. Contour reverse slope terraces would increase underground runoff with average regulating ratios of 34% (15 mm/h), 94% (30 mm/h), and 45% (50 mm/h), with no obvious correlation with rainfall intensity.
The regulation effect in the runoff of the contour reverse slope terraces was not only manifested in the cumulative runoff modulus but also in the process of runoff yielding. Figure 8 shows the surface (Figures 8(a)–8(c)) and underground (Figures 8(d)–8(f)) runoff yielding process of the bare slope and contour reverse slope terraces.
(a) 15 mm/h rainfall intensity
(b) 30 mm/h rainfall intensity
(c) 50 mm/h rainfall intensity
(d) 15 mm/h rainfall intensity
(e) 30 mm/h rainfall intensity
(f) 50 mm/h rainfall intensity
As shown in Figures 8(a)–8(c), at low and medium rainfall intensities of 15 and 30 mm/h, the contour reverse slope terrace weakened the peak of surface runoff rate and stabilized surface runoff yielding. During intense rainfall of 50 mm/h, the contour reverse slope terrace could also maintain the stability of surface runoff rate curves for about 14 minutes. After 14 minutes, the surface runoff rate increased because the reverse slope was filled to overflowing with water.
For underground runoff in Figures 8(d)–8(f), the contour reverse slope terrace increased the slope of the curves at different rainfall intensities, thus significantly increasing underground runoff rate.
The results of partial correlation analysis are shown in Table 7. For the contour reverse slope terrace, rainfall intensity was highly significantly positively correlated with surface cumulative runoff modulus () () and surface runoff proportion () () with a partial correlation coefficient of 0.974 and 0.871, respectively, and was highly significantly negatively correlated with underground runoff proportion () () with a partial correlation coefficient of -0.871. Rainfall intensity was significantly positively correlated with underground cumulative runoff modulus () () and regulating ratio () with a partial correlation coefficient of 0.717 and 0.73, respectively. There was no significant correlation between rainfall intensity and modulating ratio.
Bedrock porosity was highly significantly negatively correlated with surface cumulative runoff modulus () () with a partial correlation coefficient of -0.868. Bedrock porosity was significantly negatively correlated with surface runoff proportion () () with a partial correlation coefficient of -0.785 and was significantly positively correlated with underground runoff proportion () () with a partial correlation coefficient of 0.785. Rainfall intensity had no significant correlation with underground cumulative runoff modulus (), regulating ratio, and regulating ratio.
3.2.2. The Regulating Effect of Contour Reverse Slope Terraces on Sediment
Runoff is the driving force of sediment loss, so the contour reverse slope terrace is bound to change the spatial distribution of sediment while regulating the distribution of runoff. The surface and underground cumulative sediment modulus and the percentages on the bare slope and contour reverse slope terrace are shown in Table 8.
Except for the cases of 15 mm/h rainfall intensity, 3% and 5% bedrock porosities where surface cumulative runoff and sediment modulus are minimal due to approaching critical rainfall intensity, contour reverse slope terraces did not change the dominance of surface erosion in soil erosion. The contour reverse slope terrace promoted the transfer of surface runoff to the underground while promoting the increase of underground sediment yielding. Compared with the bare slope, the underground sediment proportion () on the contour reverse slope terrace increased to different degrees with different rainfall intensities and bedrock porosities.
The effect of contour reverse slope terraces in regulating sediment is shown in Table 9, where the regulating ratio is calculated based on the bare slope data.
As shown in Table 9, the contour reverse slope terrace was able to significantly reduce surface sediment yielding at different rainfall intensities and simultaneously increase underground sediment yielding to some extent. At different rainfall intensities, the average regulating ratio of surface sediment reduction was, respectively, 70% (15 mm/h), 65% (30 mm/h), and 54% (50 mm/h), negatively correlated with rainfall intensity. The average regulating ratio of underground sediment was 39% (15 mm/h), 45% (30 mm/h), and 60% (50 mm/h), showing a positive correlation with rainfall intensity.
The surface and underground sediment yielding process of the bare slope and contour reverse slope terrace is shown in Figure 9, where surface sediment rates are shown in Figures 9(a)–9(c) and underground sediment rates are shown in Figures 9(d)–9(f).
(a) 15 mm/h rainfall intensity
(b) 30 mm/h rainfall intensity
(c) 50 mm/h rainfall intensity
(d) 15 mm/h rainfall intensity
(e) 30 mm/h rainfall intensity
(f) 50 mm/h rainfall intensity
Figure 9 shows that the contour reverse slope terrace significantly decreased the fluctuation of the surface sediment rate. At 50 mm/h rainfall intensity and 5% bedrock porosity, the regulation effect of the contour reverse slope terrace was especially pronounced because of the large amount of surface erosion under the impact of intense rainfall. At other rainfall intensities and bedrock porosities, the contour reverse slope terrace could play a quite good role in reducing surface sediment.
Contour reverse slope terraces also increased underground sediment yielding. In the overall underground sediment rate, there was a slight ladder-type increase with time at the low and medium rainfall intensities of 15 mm/h and 30 mm/h and a continuous increase at intense rainfall of 50 mm/h.
The results of partial correlation analysis in Table 10 showed that for the contour reverse slope terrace, rainfall intensity was significantly positively correlated with surface cumulative sediment modulus () (), underground cumulative sediment modulus () (), and surface sediment proportion () () with a partial correlation coefficient of 0.776, 0.749, and 0.765, respectively, and was significantly negatively correlated with underground sediment proportion () () with a partial correlation coefficient of -0.765. Rainfall intensity was highly significantly positively correlated with regulating ratio () with a partial correlation coefficient of 0.843. There was no significant correlation between rainfall intensity and regulating ratio.
Bedrock porosity was significantly negatively correlated with surface cumulative sediment modulus () (), surface sediment proportion () (), and regulating ratio () with a partial correlation coefficient of -0.717, -0.793, and -0.815 respectively, and was significantly positively correlated with underground cumulative sediment modulus () () and underground sediment proportion () () with a partial correlation coefficient of 0.801 and 0.793, respectively. Bedrock porosity was highly significantly positively correlated with regulating ratio () with a partial correlation coefficient of 0.925.
3.3. Mechanisms of Contour Reverse Slope Terraces for Regulating Soil and Water Erosion on the Bare Slope
The principle on which contour reverse slope terraces regulate soil and water erosion lies in changing the spatial distribution of surface runoff, and the effect of regulating surface runoff varies with rainfall intensities, which is determined by the inherent characteristic of contour reverse slope terraces. As shown in Figure 3, two contour reverse slope terraces divide the bare slope into three parts: upper, middle, and lower parts. At the rainfall intensity of 15 mm/h, surface runoff from the upper and middle parts could be completely intercepted by and retained in the two reverse slopes and infiltrated into the soil, as shown in Figure 10. There was not much surface runoff in the lower part either since it was short and small, and few raindrops could reach it. As seen from the flat curves in Figure 8, the contour reverse slope terrace has relatively low surface runoff rates at low and medium rainfall intensities (15 mm/h and 30 mm/h).
(a)
(b)
At the intensity of 50 mm/h, the curves of 1% and 5% bedrock porosity in Figure 8 both show apparent and rapid increase after 14 minutes, because the reverse slope was full, and the overflowing water formed surface runoff. This also explains why the regulation effect of contour reverse slope terrace in reducing surface runoff is negatively correlated with rainfall intensity (Table 6). When the reverse slope overflowed, the new rainfall could not infiltrate in time, so the contour reverse slope terrace produced the best regulation effect at the rainfall intensity of 15 mm/h, and the regulation effect was significantly reduced at the rainfall intensity of 50 mm/h.
At the rainfall intensity of 50 mm/h, the regulation effect of the contour reverse slope terrace in reducing surface sediment would not be significantly affected. In Table 9, although the underground sediment regulating ratio is negatively correlated with rainfall intensity, the numerical difference is insignificant, indicating that the contour reverse slope terrace can exert a good effect at all three rainfall intensities for the reason that rainfall is the original driving force of sediment loss. At the rainfall intensities of 15 mm/h and 30 mm/h, the sediment carried in the surface runoff from the upper and middle parts was deposited in the two reverse slopes and could not go downhill unless the reverse slopes were filled up by water. In other words, if the reverse slope was not full, sediment loss would only occur at the lower part of the slope. Therefore, contour reverse slope terraces can produce a significant effect in reducing surface sediment at the rainfall intensities of 15 mm/h and 30 mm/h.
At the rainfall intensity of 50 mm/h, when the reverse slope was overflowing, the initial velocity of the runoff produced by the overflowing water could be regarded as zero. The runoff from the upper part would also be slowed down when passing through the reverse slopes, and the coarse sediment particles in the runoff would deposit in the reverse slopes and could not be transported further downhill. Both the runoff with zero initial velocity and the runoff decelerated by the reverse slope had much weaker scouring force to the soil surface than the runoff rushing downhill without the obstruction of the reverse slope. Moreover, whether the reverse slope was full or not, soil particles in it would not fly about under the impact of the rainfall. Therefore, contour reverse lope terraces have good regulation effect at different rainfall intensities.
Previous studies showed similar results. Shuai-bing et al. [39] pointed out that contour reverse slope terraces had good effect of reducing runoff and sediment regardless of rainfall intensity, with the best effect of reducing runoff at low and medium rainfall intensities and the best effect of reducing sediment at medium rainfall intensity. The research results of Ping et al. [40] indicated that the average regulation rate of surface runoff reached 65.3% and the average regulation rate of surface sediment reached 80.7% after the reverse slope was laid on red clay sloping farmland in the mountainous area.
The reason why the contour reverse slope terrace increases underground water and soil erosion lies in the increase of the time for surface runoff to infiltrate. When the reverse slope was not full of water, the surface runoff in the reverse slope had sufficient time to infiltrate and turn into the underground runoff. When the reverse slope was filled up by water, the surface runoff from the upper part decelerated after flowing through the reverse slope and had a longer time to move at the surface and infiltrate into the soil.
More underground runoff must result in more underground sediment. Therefore, regardless of rainfall intensities, contour reverse slope terraces would enhance underground runoff and sediment yielding to various degrees. At 50 mm/h rainfall intensity and 5% bedrock porosity, the increasing ratio of underground cumulative runoff modulus () was the lowest (7%, Table 6) because this was the first round of the test after laying contour reverse slope terraces when surface soil in and around the reverse slope was compacted, which resulted in relatively poor soil permeability, slow infiltration of surface water, and low increasing ratio of underground cumulative runoff modulus (). When the compacted soil was loosened under the impact of raindrops and runoff, the increasing ratios of underground cumulative runoff modulus () in all subsequent tests were higher than that in the first one.
4. Conclusion
(1)Under the conditions of short and heavy rainfall, soil erosion on red clay sloping farmland in the karst area is dominated by surface soil erosion. For bare red clay slope, rainfall intensity is highly significantly positively correlated with surface cumulative runoff modulus and surface cumulative sediment modulus and is significantly correlated with underground cumulative runoff modulus and underground cumulative sediment modulus. Bedrock porosity is highly significantly negatively correlated with surface cumulative runoff modulus and is significantly negatively correlated with surface cumulative sediment modulus. Bedrock porosity is highly significantly positively correlated with underground cumulative runoff modulus and is positively correlated with underground cumulative sediment modulus(2)Contour reverse slope terraces will change the spatial distribution of surface runoff and promote the transfer of surface runoff to the underground, while increasing the underground sediment yielding. Still, surface soil erosion remains the dominant erosion for red clay sloping farmland. At different rainfall intensities of 15 mm/h, 30 mm/h, and 50 mm/h, the average proportion of increased underground runoff is 34%, 94%, and 45%, and the average proportion of increased underground sediment is 39%, 45%, and 60%, respectively(3)At different rainfall intensities, contour reverse slope terraces can effectively reduce surface water and soil erosion. At the rainfall intensities of 15 mm/h, 30 mm/h, and 50 mm/h, the average proportion of surface runoff reduction is 71%, 52%, and 47%, and the average proportion of surface sediment reduction is 70%, 65%, and 54%, respectively. Laying contour reverse slope terraces on red clay sloping farmland in northern Guangxi can reduce water and soil erosion
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
No potential conflict of interest was reported by the authors.
Acknowledgments
This research was funded by the National Natural Science Foundation of China (Grant no. 41867039), the Foundation of Technical Innovation Center of Mine Geological Environmental Restoration Engineering in Southern Area (CXZX2020002), and the Guangxi Key Laboratory of Geotechnical Engineering (20-Y-XT-03).