Next Article in Journal
The Growth and Physiological Responses of Gleditsia sinensis Lam. Seedlings with Different Phosphorus Efficiencies to Low Phosphorus Stress
Previous Article in Journal
Different Responses of Terrestrial Carbon Fluxes to Environmental Changes in Cold Temperate Forest Ecosystems
Previous Article in Special Issue
Understorey Plant Functional Traits of Platycladus orientalis Depends on Crown Closure and Soil Properties in the Loess Plateau, China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 15
Issue 8
10.3390/f15081341
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Quantitative Study on the Effects of Vegetation and Soil on Runoff and Sediment in the Loess Plateau

by
Gaohui Duan
1,
Chunqian Leng
2,
Zeyu Zhang
3,
Cheng Zheng
1 and
Zhongming Wen
1,4,*
1
College of Grassland Agriculture, Northwest A&F University, Yangling 712100, China
2
College of Chemical Engineering, Shandong Institute of Petroleum and Chemical Technology, Dongying 257000, China
3
School of Economics and Management, Shanghai Maritime University, Shanghai 201306, China
4
Institute of Soil and Water Conservation, Chinese Academy of Sciences & Ministry of Water Resources, Yangling 712100, China
*
Author to whom correspondence should be addressed.
Forests 2024, 15(8), 1341; https://doi.org/10.3390/f15081341
Submission received: 5 June 2024 / Revised: 10 July 2024 / Accepted: 31 July 2024 / Published: 1 August 2024
(This article belongs to the Special Issue Forest Disturbance and Management)
Browse Figures
Review Reports Versions Notes

Abstract

:
Runoff and sediment (RAS) are important indicators of soil erosion in a watershed, playing a significant role in the migration of surface material and landform development. Previous studies have extensively documented the effects of trees, shrubs, herbs, and soil on runoff and sediment during erosive rainfall; however, the precise interactions among these factors and their influence on RAS yield within the vegetation hierarchy remain unclear. Using the random forest algorithm and the structural equation model, this research aimed to quantify the interaction of numerous variables within diverse vegetation hierarchies and how they affect RAS, as well as to identify critical indicators that influence RAS. The structural equation model results show that the grass properties have a direct effect on soil properties, and the grass properties and soil properties both affect the canopy properties directly; the soil properties and canopy properties are the main factors influencing runoff and sediment directly. In addition, the grass properties could affect RAS by influencing the soil properties indirectly, and the soil properties could also affect RAS indirectly by influencing the canopy properties. Height difference (HD) between two layers of vegetation had the highest weight of 1.043 among the canopy variables, showing that HD has a substantial effect on RAS. Among the soil properties, soil bulk density and maximum field capacity have a significant impact on RAS. We conclude that canopy properties have the greatest impact on RAS. In the future, more Caragana microphylla Lam and Robinia pseudoacacia Linn plants should be planted to prevent soil erosion. This study provides a scientific basis for vegetation planting management and soil erosion control on the Loess Plateau.

1. Introduction

Vegetation has been widely studied in areas affected by soil erosion due to its crucial role in controlling erosion and rehabilitating the environment [1]. Nevertheless, various types of vegetation exhibit diverse hierarchies that influence soil and water conservation to varying extents. A rational vegetation structure can mitigate splash erosion induced by erosive rainfall, particularly on the Loess Plateau, where erosive rainfall is prevalent during the period from July to September [2]. Therefore, understanding the impact of vegetation hierarchy and function on soil and water conservation under erosive rainfall conditions is critical for determining vegetation hierarchy and planting strategies in varied environments [3]. For example, establishing how to choose appropriate vegetation coverage in places with serious soil erosion might enhance the impact of ecological engineering [4]. Grasslands hold immense potential for optimizing the delicate balance between mitigating soil erosion and conserving surface water resources in semi-arid environments, thereby proving to be a particularly suitable ecosystem in these regions [5].
The influence of vegetation on soil and water conservation exhibits significant variations, stemming from latent factors inherent to distinct vegetation hierarchies, including canopy characteristics, grass properties (understory natural grass), and soil properties. Since the 1950s, extensive research has been conducted in this realm, focusing on the intricate relationships between soil, vegetation, species diversity, and soil erosion, with the overarching goal of mitigating soil erosion [6]. On the one hand, the canopy, shrub, and herbaceous layers play a crucial role in intercepting precipitation and dissipating its kinetic energy. Improvements in vegetation structure can modulate the obstructive and redistributive effects of vegetation on rainfall, thereby reducing the potential for rainfall-induced erosion on the soil surface and prolonging water infiltration periods, ultimately leading to a decrease in runoff [7]. For instance, the proliferation of herbaceous species beneath the canopy significantly influences sediment storage and interception processes [8]. Conversely, there exists a complex interplay between soil characteristics, plant biology, runoff, and sediment dynamics. Soil water and nutrients are fundamental to plant growth and development [9]. Moreover, soil bulk density and maximum water-holding capacity are the primary determinants of vegetation composition and runoff patterns [10]. These studies have illuminated the relevance and causality between influencing factors and runoff generation, as well as sediment yield [11]. However, it remains unclear how these diverse factors integrate to influence runoff and sediment yield and how they interact with each other within this complex process.
As research in runoff and sediment prevention and control has progressed, scientists have come to understand that investigating the factors influencing runoff and sediment necessitates not only establishing appropriate vegetation systems but also predicting runoff and sediment levels under various vegetation structures. Despite extensive research into the influence of canopy properties, grass properties, and soil properties on runoff and sediment production in semi-arid environments, the complex interactions among these factors and the factors of primary importance remain uncharted territory. This knowledge gap underscores the need for a thorough investigation of these influencing factors to delve into the connections between the canopy, grass, soil, and runoff-sediment production and to gain insights into the mechanisms governing the runoff-sediment production process. Consequently, the objectives of this study are to (a) select the most appropriate runoff and sediment indexes to gauge the extent of soil erosion in the watershed; (b) explore how interactions among canopy, grassland, soil, and diversity factors affect runoff and sediment latent variables; and (c) evaluate the effects of canopy, grass, soil, and diversity factors on runoff and sediment latent variables, and identify appropriate vegetation types and species to prevent soil erosion.

2. Materials and Methods

2.1. Study Area

The Xindiangou watershed in Suide Country, Shaanxi Province, China (37°27’–37°32’ N, 110°15’–110°20’ E; elevation 850–1287 m; area 1.44 km2), is one of the most typical areas of soil erosion in the Loess Plateau. As a National Soil and Water Conservation Demonstration Park with a history of 70 years, it has become a typical representative of comprehensive soil and water conservation management. Therefore, this watershed was selected as the research area in this study. This watershed is characterized by typical loess hilly and gully landforms, low vegetation coverage, and a sandy loam soil texture. The average annual precipitation is 486 mm, 70% of which falls between July and September. According to the WRB system, the soil type is Calcaric Regosols [12]. Due to the low rainfall and hot climate in the Loess Plateau, vegetation development is severely constrained by water availability [13,14]. The main tree species in the area are Platycladus orientalis (L.) Franco, Robinia pseudoacacia L., etc., while the majority of shrub species are Caragana microphylla Lam, Ziziphus jujuba var. spinosa (Bunge) Hu, Salix cheilophila Schneid, etc. More than 80 grass species dominate the vegetation in the study area, which mainly include Asteraceae Bercht. & J. Presl, Poaceae Barnhart [12].

2.2. Sample Plot Survey and Sample Collection

In this study, nine fixed plots of 100 m2 (including five different types of community structure, the details of which are in Table 1) were selected. Each plot contained five 1 m × 1 m grassland plots representing the general herbaceous information of the plot. The plots were set on the same slope (20°) with the same soil types (Figure 1 and Figures S1–S3). Prior to each erosive rainstorm, the amount, coverage, and proportion of dominant species and plant height of herbaceous, shrub, and tree species were surveyed. A DJI UAV was used to gather the total coverage, and it was then brought back to the lab for image processing using ArcGIS 10.6 to determine the coverage. After the investigation, five individuals of each species were randomly selected, and at least 10 fully healthy leaves were collected per individual in each quadrat. Meanwhile, soil samples were collected from seven randomly selected points in each plot (S-shape). After the litter was removed, the soil was collected with a 100 mm2 cylindrical metal sampler at depths of 0–10 cm,10–20 cm, and 20–30 cm. To increase the sample size, we additionally surveyed 34 temporary plots with similar terrain, vegetation, and soil to the fixed plots, and collected their vegetation and soil data.
Siphon rain gauges were erected in Xindian watershed to automatically record hourly rainfall data during the rainy season. This study selected rainfall events based on the following criteria: (1) the rainfall events were erosive, with a maximum 30 min rainfall intensity standard of 0.25 mm/min; (2) when it rained for more than 6 h, the rainfall event was broken into two events (Wischmeier 1959); and (3) the onset of rainfall events coincided with the onset time of runoff.
From 2018 to 2020, the runoff and sediment from fixed plots were collected and measured following seven erosive rainfalls events. The runoff pool with a scale was used to determine the clean water depth and muddy water depth. A measuring cylinder was used to determine the total amount of sediment in the runoff pool. After the supernatant was determined, all sediment samples from the runoff pool were brought back to the laboratory. These samples were dried in 105 °C ovens for 24 h and then weighed to determine the total amount of sediment [15]. After each rainfall event, the runoff pools were emptied in preparation for the next rainfall event. Finally, 63 groups of runoff and sediment data from 9 fixed plots were collected. The temporary sample plot’s runoff and sediment amount was calculated using the average runoff and sediment data of each vegetation structure sample plot from 2018 to 2020.

2.3. Data Analysis

The random forest algorithm (RF) was used to select essential dependent variables and exclude those with a low correlation [16,17]. RF is an ensemble algorithm that classifies by voting on multiple unbiased classifier decision trees (R 4.1.2 Boruta package). Then, to increase the model’s fit, a correlation test for variables was conducted using canonical correspondence analysis (CCA). Finally, the insignificant path in this model was removed, and a new partial least squares structural equation model (PLS-SEM) was established using the RF filtered variables. PLS-SEM was used to explain the correlation and causality between runoff and sediment variables and their influencing variables in this study. The CCA was performed using the R package Vegan, and PLS-SEM analysis was performed with SmartPLS 3. In general, a qualified PLS-SEM model should match the criteria below: (1) compound reliability values (CR) > 0.7; (2) average variance extraction values (AVE) > 0.5; (3) variance inflation factor values (VIFs) < 5.0; and (4) a standardized root mean square residual value (SRMR) < 0.10.

2.4. Candidate Variables

The PLS-SEM was built by combining four independent variables (canopy properties, grass properties, soil properties, and vegetation diversity) and one dependent variable (runoff and sediment properties). We used 6 indexes to indicate the characteristics of canopy properties: leaf dry weight (FLT), height difference (HD), top layer of vegetation height (TVH), canopy width of top layer of vegetation (CTV), vegetation coverage of the top floor (VCT), and number of canopies (CN). To describe grass properties in the study area, seven observational variables were chosen: leaf thickness (LT), leaf tissue density (LTD), leaf area (LA), specific leaf area (SLA), vegetation coverage (VC), average plant height (PH), and number of vegetation layers (VL). In addition, four independent observation variables, namely soil water content (SWC), soil bulk density (BD), maximum soil water capacity (BW), and soil porosity (SP), were selected as the observed variables to further explain the soil characteristics variable. Five alternative species diversity indexes, including the Shannon–Wiener index (SHA), Simpson index (SIM), Margalef index (MAR), Pielon index (PIE), and Gleason index (GLE), was selected to ensure the integrity of diversity information. Finally, 6 indexes were used to explain the runoff and sediment variation, including total runoff and sediment weight (TRS), total sediment weight (TS), clear water coefficient (CW), depth of clear water (DW), muddy water coefficient (MWC), and depth of muddy water (DMW). Indicator details are provided in the Supplementary Materials (Table S1).

2.5. Determination of Final Variables

The RF algorithm was used to eliminate the redundancy in runoff and sediment indexes. The weighted values of 6 runoff and sediment indexes were obtained through residual variables. We ranked all of the weighted values for each runoff and sediment index, and then chose the index with the most votes to represent the runoff and sediment status. Confidence levels and maximum runs of RF were set as 0.01 and 100 in the algorithm, respectively [12]. The algorithm was based on the Boruta package in R 4.1.2.

2.6. Establish the Preliminary Model

SEM deals with complex data relationships. It is mainly applied in the PLS-SEM and in covariance-based structural equation modeling (CB-SEM), etc. The PLS-SEM has better predictive accuracy when compared to CB-SEM, which focuses on parameter assessment and has a wider tolerance [12,18]. Therefore, the PLS-SEM was employed to quantify the relationship between variables. The preliminary model was established using three causality hypotheses: (a) grassland properties and diversity directly affect soil properties; (b) grassland and soil properties directly affect canopy properties; (c) canopy properties, grass properties, and soil properties all affect runoff and sediment (Figure 2).
To increase the model’s fit, a correlation test for variables was conducted using canonical correspondence analysis (CCA). Twenty-eight indexes were separated into five sets of variables, and correlation analysis was performed for each pair. Finally, the insignificant path in the model was removed, and a new PLS-SEM was established. The CCA was performed using the R package Vegan.

3. Results

In the plots, four different types of vegetation structure were identified: tree + shrub, shrub, mixed tree + shrub, and mixed shrub. Five tree species, six shrub species, and fifty-one herbaceous species were found across the plots. The ranges of the six dependent variables used to explain the latent variables of runoff and sediment in the sample plots were as follows: MWC: 0.32–2.57; CW: 0.31–2.55; TS: 7.46–12.6; TRS: 2.5–28.4; DW: 0.23–2.71; DMW: 0.34–2.74 (Figure 3).

3.1. Observed Variables of Runoff and Sediment

In the RF, the 22 independent observation variables of canopy properties, grass properties, soil properties, and vegetation diversity were used to classify the sic dependent variables. We obtained 22 groups of important values based on the regression analyses, as shown in Figure 4, by treating the contribution of dependent variables to independent variables as important values. The values of HD and FLT showed the highest significance, with all the important values of groups being accepted with high coefficients. Conversely, the TRS and TS groups were rejected by SIM and SHA; their important values were −0.39, −0.58, −0.19, and −0.55, respectively. The important values in MAR, SWC, and PIE were all found to be irrelevant, and all variables were rejected. The important values of canopy properties, grass properties, soil properties, and vegetation diversity were 16, 7, 2, and 8, respectively, among the various observed variables.
The six dependent variables are ranked according to 33 important values in Figure 4, and the results are reported in the following: TRS (7) > DW (6) DMW (6) > CW (5) MWC (5) >TS (4). TRS received the most votes, the majority of which came from canopy properties, while TRS received the only two votes from soil properties. The votes of DW and DMW were derived from two votes from canopy properties, grass properties, and diversity, respectively. All of the votes that TS received were due to canopy properties. CW and MWC each received two votes from canopy properties and vegetation diversity, and one vote from grass properties. Based on the votes of each dependent variable, we selected TRS, DW, and DMW as the observation variables of the runoff and sediment in PLS-SEM (Table S2).

3.2. Possible Path and Model

The data in Table 2 indicate that there was no statistically significant link between vegetation diversity and soil properties. In line with the interaction between canopy properties and other latent variables, the results for the interaction between vegetation diversity and grass properties, canopy properties and grass properties, canopy properties and runoff and sediment, and canopy properties and soil properties were highly statistically significant. Additionally, the CCA results of the relationships between soil properties and grassland properties, soil properties and runoff and sediment, and grassland properties and runoff and sediment all showed significant correlations, indicating that soil properties were strongly correlated with grassland properties, runoff and sediment, and grassland properties and runoff and sediment, respectively.
If the CCA results reveal no significant relationships between variables, we believe that this path does not exist in the final model. As a consequence, we established the preliminary path of the PLS-SEM based on the CCA results of different variable groups by eliminating a path from vegetation diversity to soil properties (Figure 5).

3.3. Model Fit

In the final model, the CR value is 0.790, indicating that the model has a good combination reliability. The AVE value of 0.660 indicates high aggregation validity [19]. The Fornell–Larcker criterion matrix was used to assess identification validity. The AVE square root of runoff and sediment was 0.813, which was higher than the correlation with the other variables. The VIF value being <5 suggests that there is no collinearity in all variables [20]. To avoid model misjudgment, the standardized root means square residual (SRMR) was used as a fitting measure. The SRMR value of 0.099 is acceptable in PLS-SEM [21,22].

3.4. Evaluation of Runoff and Sediment’ Effect Factor

The factor analysis of final model showed the weight of each observed variable in relation to the latent variable (Figure 6). Only TRS and DW were observed among the latent variables of runoff and sediment, with weights of 1.234 and −0.546, respectively. GLE and BD were the largest contributors among the latent variables of vegetation diversity and soil properties, with contribution values of 0.95 and 1.819, respectively. HD had the highest contribution value of 1.043 among the latent variables of canopy properties, while VCT and FLT had the opposite contribution values of 0.516 and −0.491, respectively.
According to the bootstrapping results, the T-statistics and p-values of all paths are acceptable, indicating that all effects of the final model are significant (Figure 6). The total effects of vegetation diversity on grass properties (T-statistics = 1.96, p-values < 0.05), grass properties on soil properties (T-statistics = 3.903, p-values < 0.001), soil properties on canopy properties (T-statistics = 3.492, p-values < 0.001), soil properties on runoff and sediment (T-statistics = 1.981, p-values < 0.05), and canopy properties on runoff and sediment (T-statistics = 9.757, p-values < 0.001) are 0.893, 0.598, −0.489, −0.213, and 0.743, respectively. In addition, the indirect effects of grass properties on runoff and sediment (T-statistics = 3.31, p-values < 0.01), grass properties on canopy properties (T-statistics = 2.861, p-values < 0.01), and soil properties on runoff and sediment (T-statistics = 3.159, p-values < 0.01) are −0.345, −0.292, and −0.363, respectively. According to bootstrapping, this is acceptable.

4. Discussion

4.1. Direct Effects of Vegetation Diversity on Grass Properties

According to Wang et al. (2016), vegetation, soil structure, and water management are all important for ecological restoration in the Loess Plateau, with grassland being the predominant vegetation type [23]. A major focus right now is on improving grass properties to enhance soil and water loss control [24]. The results of the SEM analysis demonstrate that vegetation diversity has a direct influence on grass properties. The herbaceous layer species diversity index is a partial indicator of vegetation competitiveness, with highly adapted species gradually outperforming less adapted ones [25].
The grass property results show that mutual competition between plants leads to variations in the VL index, characterizing the spatial structure of the vegetation hierarchy. Consequently, plots with higher vegetation diversity have more vegetation layers [26]. Furthermore, as indicated by the considerable fluctuations in PH, the intense competition within the community for light, water, heat, and other important resources reveals the substantial impact of vegetation diversity on PH [27]. Higher competition, in particular, drives plants to compete for light and necessary conditions by increasing leaf area, whereas higher species diversity contributes to increased nutrient availability in the soil, impacting LTD [28]. According to the SEM analysis, changes in vegetation diversity affect the latent variables of LTD. This is primarily due to the fact that plants fight for resources by growing larger leaf areas and crowns. Conversely, the harsh environment of the sparsely vegetated Loess Plateau impedes vegetation growth [29].

4.2. Direct Effects of Grass Properties on Soil Properties

Because of the scarcity of water resources, research into soil water physical properties has become a hot topic [30]. Water conservation is a vital function of the ecosystem, especially in the Loess Plateau, where water sources are limited and summer rainstorms are common. Understanding how to adjust the vegetation structure to improve soil physical properties is particularly important. Our results indicate that the grass properties also directly affected the soil properties [31]. Other researchers reported similar findings, hypothesizing that multiple soil functions (including water storage, water filtration, nutrient storage, and recycling) can be expressed by linking plant traits to a range of soil biochemical or physical structural properties that are frequently regulated by a small number of easily measurable plant traits [32].
Our findings demonstrate that the vegetation traits have a significant influence on BD and BW for the target variables. Since the study area is prone to heavy rain in the summer and the soil in the research area suffers most from soil erosion, the grassland properties will have a considerable impact on soil erosion [33]. This is supported by the findings of Zhao et al., who demonstrated how appropriate vegetation management practices in the Loess Plateau can improve soil conditions, hence promoting vegetation growth and habitat quality [34]. First, the bulk density of surface soil, which is determined by soil porosity, is the most affected in this study. Studies have shown that reasonable vegetation types can reduce the bulk density of surface soil and increase the porosity of soil, thereby improving the permeability function of soil and effectively increasing the content of soil nutrients and water [35]. Second, grassland properties have a significant effect on BW, and the increase in BW helps to increase the water storage capacity of the soil, which in turn delays and reduces the formation of runoff during erosive rainfall. Wang et al.’s study showed that the size of the maximum field capacity indexes can help us to understand the changes in soil moisture and drought [36]. These results suggest that different vegetation types have considerable effects on soil porosity, which may be ascribed to the changes in root status and nutrient absorption capabilities of different vegetation types, which ultimately affect the BW indicator [37].

4.3. Direct Effects of Soil Properties on Canopy Properties

In this study, the influence of soil characteristics on the canopy properties was quantified. The results show that soil characteristics constrained the development of canopy properties. In a related study, it was shown that water availability was the main factor limiting vegetation growth in the Loess Plateau region [38].
According to the model’s results, the soil properties had the greatest influence on HD, which is the difference in the canopy height between the top layer of vegetation and the canopy height of the next layer of vegetation (tree and shrub layers, tree and herb layers, shrub, and herb layers). Trees require a lot of water during their growth and development, but soil water is scarce in the arid Loess Plateau [39]. Herbaceous vegetation, on the contrary, has great competitiveness in water acquisition due to its small water demand, and thus is less constrained by water limitations [40]. Secondly, the model’s results show that the soil properties also had a great influence on FLT. The FLT index measures the average leaf thickness of vegetation, which reflects the response path and the mechanisms of the response of vegetation to environmental changes, and is of great significance to plants [41]. With an increase in FLT, plants can improve their water storage capacity and drought tolerance. Similar findings indicate that leaf thickness changes rapidly when soil water content falls below a particular threshold [42]. In summary, our results reveal the mechanism of how the characteristics of the soil and the characteristics of the canopy interact. On the one hand, the interaction between canopy properties and soil is reflected in the protection of soil by the canopy layer and the improvement of soil nutrient content (C, N, P and so on) by the stand through photosynthesis. On the other hand, the changes in soil water and soil nutrients will also affect the growth of canopy, which are complementary to each other and ultimately work together to enhance the effect of soil and water conservation and improve the watershed environment.

4.4. Direct Effects of Canopy and Soil Properties on Runoff and Sediment

In fragile and sensitive ecosystems, there is a pressing need to find effective ways to reduce soil erosion. We discovered the significant direct effects of canopy properties and soil properties on runoff and sediment. The effect of canopy properties on runoff and sediment was mainly caused by the changes in the HD, FLT and VCT indexes, and soil properties mainly affected the runoff and sediment through the BD and BW [43].
The results support the notion that canopy properties have a positive impact on runoff and sediment, indicating that the runoff and sediment indexes will increase with the increase in canopy properties indexes. Runoff and sediment were significantly impacted by the HD index among the canopy properties. However, previous studies have focused mostly on the protective effect of tree height on surface soil during rainfall [44]. This work offers fresh insights indicating that HD index can more accurately describe the protection of surface soil and reduce the kinetic energy of raindrops than the tree height index. Studies have shown that HD can play the best role when it is below 3 m, and a previous study also proved this. Their research showed that the diameter and kinetic energy of raindrops falling to the ground first decreased and then gradually increased when the vegetation height changed while the canopy cover remained constant [45]. Because the leaf thickness might indicate a plant’s ability to withstand raindrops, FLT also has a significant impact on the latent variables of runoff and sediment. Especially when the rainfall intensity is higher, vegetation has a better blocking effect when the leaf thickness is greater, preventing raindrops from spattering soil at a faster speed. Last but not least, runoff and sediment are also impacted by the VCT index [46]. The main reason for this may be that understory vegetation is crucial to water proofing and sediment reduction [47]. However, as canopy cover grows, understory vegetation struggles to obtain enough light, which leads to sparse herbaceous vegetation and increased runoff and sediment.
The results show that the soil properties have a direct impact on runoff and sediment, primarily through BD and BW indexes. The Loess Plateau is prone to soil erosion due to its low annual rainfall, primarily from summer rainstorms, limited vegetation species variety, low vegetation coverage, and poor soil water retention capacity [48]. Therefore, the latent variables of soil characteristics impacting runoff and sediment can thus be accurately represented by the BD and BW indexes. According to Marzen et al. (2017), soil’s susceptibility to erosion, particularly under intense rainfall, is significantly influenced by its texture and physical properties [49]. Furthermore, soil properties play a pivotal role in influencing the DW index, with high soil bulk density and field water holding capacity negatively impacting this index [50]. Through artificial rainfall experiments, some studies have also shown that the enhancement of soil physical properties had a positive impact on waterproofing and sediment reduction [51]. Finally, due to the short and strong rainfall periods in the study area, the underground root system of vegetation has little influence on runoff. In addition, the amount of litter on the surface of the soil is also lower, so this study did not investigate the litter in the sample area.
Although there are many factors affecting runoff and sediment, this work highlights how soil and canopy characteristics affect soil and water loss. Our findings provide a fresh perspective on how to reduce soil erosion by improving vegetation hierarchy and soil quality.

4.5. Indirect Effects of Grass Properties on Canopy Properties

In the PLS-SEM, we quantified the indirect effects of grass properties on canopy properties, which mainly affect soil properties. The results prove that there was no direct causal relationship between grass properties and canopy properties.
The composition, structure, and function of vegetation in an ecosystem are significantly influenced by plant interactions [52]. Generally speaking, two factors primarily limit the growth of vegetation in the canopy layer: one is a dense canopy of the upper layer of vegetation, which restricts the utilization of light resources by underlayer vegetation [53], and the other one is the fact that understory and overstory vegetation compete via their seedlings for limited living resources (water, nutrients, etc.). According to Haseltine and Prentice (1996), the upper vegetation has the advantage in light utilization, while the lower vegetation has the advantage in the competition for living resources [54]. However, the strong light conditions and limited water and nutrient resources will increase the advantages of herbaceous vegetation, and severely restrict the growth and development of canopy-layer vegetation in the Loess Plateau [55]. Among the latent variables of grass properties, the VL index, which measures the number of herbaceous vegetation layers, had the greatest influence on the canopy properties. In fact, the stratification phenomenon will be enhanced when herbaceous plants are provided with adequate living resources, so the competition for soil resources will be increased and the characteristics of canopy-stratified vegetation will be limited. Furthermore, the indexes such as LTD, PH, and VC have an impact on the nature of the stand, and they restrict the development of arbor-layer vegetation by enhancing their own functions to compete for soil resources. Finally, the SLA index having the least influence on the canopy-layer vegetation is mainly because the top layer of vegetation plays a dominant role in the acquisition of resources such as light, and the change in grass vegetation traits will not affect the leaf area and other traits of the canopy layer of vegetation [56].

4.6. Indirect Effects of Grass and Soil Properties on Runoff and Sediment

According to each variable’s specific influencing mechanisms, vegetation structure has an important influence on the latent variables of runoff and sediment and indirectly affects the runoff and sediment yield. The vegetation characteristics in the top layer of the sample plot may vary depending on the grass and soil properties [57]. According to the results of the structural equation, there was no direct relationship between grass properties and runoff and sediment yield, but both grass and soil properties affected the latent variables of canopy properties, which in turn indirectly affected runoff and sediment yield.
VL contributed the most to the observed variables related to grass properties, followed by PH, while SLA contributed the least. This conclusion could be attributed to the fact that the study area is resource-limited, while vegetation constantly enhances its own functions to gain advantages in order to compete for limited living resources [58], with PH and VL being most competitive traits. Among the observed variables related to soil properties, BD has a somewhat bigger contribution than BW. This could be because this study was conducted in the sub-humid and semi-arid temperate zones, where the soil’s water-holding capacity is one of the most important factors affecting vegetation growth. Furthermore, differences in BD are mostly caused by differences in vegetation structure and composition, and numerous studies have demonstrated that BD has a significant impact on vegetation structure [59]. As a result of their combined influence, canopy properties changed accordingly. Among the observed canopy properties, HD, VCT, and FLT contributed the most in descending order. The likely reason for this result is similar to the explanation for grass properties, which that plants are competing for living resources [60]. The main motive for the change in canopy properties, however, is to compete for light, followed by the competition for water and nutrition resources [61]. A series of interactions during the competition of living resources ultimately affect runoff and sediment yield in the resource-poor Loess Plateau. TRS contributes the most to runoff and sediment yield, which are indirectly influenced by canopy properties, followed by DW. TRS is the index that best summarizes this latent variable, since it indicates the total amount of runoff and sediment collected after a thunderstorm. DW reflects the depth of clean water in the collected samples, which can, to some extent, reflect the outcomes of latent variables of runoff and sediment. The results reveal that the scientific vegetation hierarchy has a good effect on soil erosion control [62].

5. Conclusions

A quantitative description of the relationship between variables in different vegetation hierarchies and their effects on runoff and sediment is of great significance to our understanding of the factors affecting soil and water loss under heavy rainfall in the Loess Plateau, which can provide new insights for vegetation management in the Loess Plateau. We used PLS-SEM to quantify the impact of the factors on runoff and sediment yield in this study. The findings were as follows: (a) the main factors affecting runoff and sediment were the canopy properties controlled by grassland and soil properties; (b) soil properties have limited influence on the latent variables of runoff and sediment yield, although some soil properties are strongly correlated with them; and (c) grassland properties indirectly affect runoff and sediment latent variables by affecting soil properties. In general, our findings show that the BD of soil properties and HD of canopy properties are the most important influencing factors, among others. The findings of the relationship between these factors and runoff and sediment latent variables are helpful for studying how to effectively control soil and water loss in the Loess Plateau, which may provide effective suggestions for dealing with soil and water loss in the Loess Plateau in the future. Vegetation measures are the most commonly uses means of preventing and controlling soil and water loss, and different vegetation types have different effects on soil erosion. In practice, this is combined with the scientific vegetation structure to control soil erosion. For example, more shrubs or small trees (such as Caragana microphylla Lam and Robinia pseudoacacia Linn) mixed with herbs could be used in the future to control the soil erosion.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/f15081341/s1, Table S1. Details of variables used in the research; Figure S1. Some fixed runoff plot photos; Figure S2. slope map of the study area; Figure S3. Slope aspect map of the study area; Table S2. The data details.

Author Contributions

Methodology, G.D.; Software, G.D. and C.L.; Investigation, G.D., Z.Z. and C.Z.; Writing—original draft, G.D.; Supervision, Z.W.; Funding acquisition, Z.W. All authors have read and agreed to the published version of the manuscript.

Funding

This work was financially supported by the National Natural Science Foundation of China (No. 41977077).

Data Availability Statement

Data are contained within the article.

Conflicts of Interest

The authors declare that they have no known financial interests or personal relationships that could have influenced the work reported in this paper.

References

  1. Chen, L.; Wei, W.; Fu, B.; Lü, Y. Soil and water conservation on the loess plateau in China: Review and perspective. Prog. Phys. Geogr. 2007, 31, 389–403. [Google Scholar] [CrossRef]
  2. Hou, G.; Bi, H.; Huo, Y.; Wei, X.; Zhu, Y.; Wang, X.; Liao, W. Determining the optimal vegetation coverage for controlling soil erosion in Cynodon dactylon grassland in North China. J. Clean. Prod. 2020, 244, 118771. [Google Scholar] [CrossRef]
  3. Chen, H.; Zhang, X.; Abla, M.; Lü, D.; Yan, R.; Ren, Q.; Ren, Z.; Yang, Y.; Zhao, W.; Lin, P.; et al. Effects of vegetation and rainfall types on surface runoff and soil erosion on steep slopes on the Loess Plateau, China. Catena 2018, 170, 141–149. [Google Scholar] [CrossRef]
  4. Stokes, A.; Sotir, R.; Chen, W.; Ghestem, M. Soil bio- and eco-engineering in China: Past experience and future priorities. Ecol. Eng. 2010, 36, 247–257. [Google Scholar] [CrossRef]
  5. Wu, G.L.; Liu, Y.F.; Cui, Z.; Liu, Y.; Shi, Z.H.; Yin, R.; Kardol, P. Trade-off between vegetation type, soil erosion control and surface water in global semi-arid regions: A meta-analysis. J. Appl. Ecol. 2020, 57, 875–885. [Google Scholar] [CrossRef]
  6. Minden, V.; Kleyer, M. Ecosystem multifunctionality of coastal marshes is determined by key plant traits. J. Veg. Sci. 2015, 26, 651–662. [Google Scholar] [CrossRef]
  7. Puigdefabregas, J.; Sole, A.; Gutierrez, L.; del Barrio, G.; Boer, M. Scales and processes of water and sediment redistribution in drylands: Results from the Rambla Honda field site in southeast Spain. Earth-Sci. Rev. 1999, 48, 39–70. [Google Scholar] [CrossRef]
  8. Bartley, R.; Roth, C.H.; Ludwig, J.; McJannet, D.; Liedloff, A.; Corfield, J.; Hawdon, A.; Abbott, B. Runoff and erosion from australia’s tropical semi-arid rangelands: Influence of ground cover for differing space and time scales. Hydrol. Process. 2006, 20, 3317–3333. [Google Scholar] [CrossRef]
  9. Puigdefábregas, J. The role of vegetation patterns in structuring runoff and sediment fluxes in drylands. Earth Surf. Process. Landf. 2005, 30, 133–147. [Google Scholar] [CrossRef]
  10. Zhang, L.; Wang, J.; Bai, Z.; Lv, C. Effects of vegetation on runoff and soil erosion on reclaimed land in an opencast coal-mine dump in a loess area. Catena 2015, 128, 44–53. [Google Scholar] [CrossRef]
  11. Zuazo, V.D.; Pleguezuelo, C.R.R. Soil-erosion and runoff prevention by plant covers: A review. In Sustainable Agriculture; Springer: Dordrecht, The Netherlands, 2009; pp. 785–811. [Google Scholar]
  12. Duan, G.; Wen, Z.; Xue, W.; Bu, Y.; Lu, J.; Wen, B.; Wang, B.; Chen, S. Agents affecting the plant functional traits in national soil and water conservation demonstration park (China). Plants 2022, 11, 2891. [Google Scholar] [CrossRef] [PubMed]
  13. Chen, L.; Wang, J.; Wei, W.; Fu, B.; Wu, D. Effects of landscape restoration on soil water storage and water use in the Loess Plateau Region, China. For. Ecol. Manag. 2010, 259, 1291–1298. [Google Scholar] [CrossRef]
  14. Jian, S.; Zhao, C.; Fang, S.; Yu, K. Effects of different vegetation restoration on soil water storage and water balance in the Chinese Loess Plateau. Agric. For. Meteorol. 2015, 206, 85–96. [Google Scholar] [CrossRef]
  15. Walkley, A.; Black, I.A. An examination of the Degtjareff method for determining soil organic matter, and a proposed modification of the chromic acid titration method. Soil Sci. 1934, 37, 29–38. [Google Scholar] [CrossRef]
  16. Ho, T.K. Random decision forests. In Proceedings of the 3rd International Conference on Document Analysis and Recognition, Montreal, QC, Canada, 14–16 August 1995; Volume 1, pp. 278–282. [Google Scholar]
  17. Duan, G.; Zhou, R.; Wang, L.; Zheng, C.; Liu, Y.; Chai, X.; Zhou, C.; Wen, Z. Effects of different soil and water conservation measures on plant diversity and productivity in loess plateau. J. Environ. Manag. 2023, 348, 119330. [Google Scholar] [CrossRef] [PubMed]
  18. Wang, B.; Bu, Y.; Li, Y.; Li, W.; Zhao, P.; Yang, Y.; Qi, N.; Gou, R. Quantifying the relationship among impact factors of shrub layer diversity in Chinese pine plantation forest ecosystems. Forests 2019, 10, 781. [Google Scholar] [CrossRef]
  19. Fornell, C.; Larcker, D.F. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 1981, 18, 39–50. [Google Scholar] [CrossRef]
  20. Hair, J.F.; Ringle, C.M.; Sarstedt, M. PLS-SEM: Indeed a silver bullet. J. Mark. Theory Pract. 2011, 19, 139–152. [Google Scholar] [CrossRef]
  21. Henseler, J.; Dijkstra, T.K.; Sarstedt, M.; Ringle, C.M.; Diamantopoulos, A.; Straub, D.W.; Ketchen, D.J., Jr.; Hair, J.F.; Hult, G.T.; Calantone, R.J. Common beliefs and reality about pls: Comments on Rönkkö and Evermann (2013). Organ. Res. Methods 2014, 17, 182–209. [Google Scholar] [CrossRef]
  22. Hu, L.T.; Bentler, P.M. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct. Equ. Model. Multidiscip. J. 1999, 6, 1–55. [Google Scholar] [CrossRef]
  23. Wang, Z.J.; Jiao, J.Y.; Rayburg, S.; Wang, Q.L.; Su, Y. Soil erosion resistance of “Grain for Green” vegetation types under extreme rainfall conditions on the Loess Plateau, China. Catena 2016, 141, 109–116. [Google Scholar] [CrossRef]
  24. Fullen, M.; Zhi, W.B.; Brandsma, R. A comparison of the texture of grassland and eroded sandy soils from Shropshire, UK. Soil Tillage Res. 1998, 46, 301–305. [Google Scholar] [CrossRef]
  25. Grubb, P. Plant populations and vegetation in relation to habitat, disturbance and competition: Problems of generalization. In The Population Structure of Vegetation; Springer: Berlin/Heidelberg, Germany, 1985; pp. 595–621. [Google Scholar]
  26. Falge, E.; Reth, S.; Brüggemann, N.; Butterbach-Bahl, K.; Goldberg, V.; Oltchev, A.; Schaaf, S.; Spindler, G.; Stiller, B.; Queck, R.; et al. Comparison of surface energy exchange models with eddy flux data in forest and grassland ecosystems of Germany. Ecol. Model. 2005, 188, 174–216. [Google Scholar] [CrossRef]
  27. Crawley, M.J. The population biology of invaders. Philos. Trans. R. Soc. Lond. B Biol. Sci. 1986, 314, 711–731. [Google Scholar]
  28. Xu, R.; Wang, L.; Zhang, J.; Zhou, J.; Cheng, S.; Tigabu, M.; Ma, X.; Wu, P.; Li, M. Growth rate and leaf functional traits of four broad-leaved species underplanted in Chinese fir plantations with different tree density levels. Forests 2022, 13, 308. [Google Scholar] [CrossRef]
  29. Feng, Q.; Zhao, W.; Fu, B.; Ding, J.; Wang, S. Ecosystem service trade-offs and their influencing factors: A case study in the loess plateau of China. Sci. Total Environ. 2017, 607, 1250–1263. [Google Scholar] [CrossRef]
  30. Mulumba, L.N.; Lal, R. Mulching effects on selected soil physical properties. Soil Tillage Res. 2008, 98, 106–111. [Google Scholar] [CrossRef]
  31. Jiang, C.; Zhang, H.; Wang, X.; Feng, Y.; Labzovskii, L. Challenging the land degradation in China’s Loess Plateau: Benefits, limitations, sustainability, and adaptive strategies of soil and water conservation. Ecol. Eng. 2019, 127, 135–150. [Google Scholar] [CrossRef]
  32. El-Ramady, H.R.; Alshaal, T.A.; Amer, M.; Domokos-Szabolcsy, É.; Elhawat, N.; Prokisch, J.; Fári, M. Soil quality and plant nutrition. In Sustainable Agriculture Reviews 14; Springer: Berlin/Heidelberg, Germany, 2014; pp. 345–447. [Google Scholar]
  33. Shi, P.; Zhang, Y.; Li, P.; Li, Z.; Yu, K.; Ren, Z.; Xu, G.; Cheng, S.; Wang, F.; Ma, Y. Distribution of soil organic carbon impacted by land-use changes in a hilly watershed of the Loess Plateau, China. Sci. Total Environ. 2018, 652, 505–512. [Google Scholar] [CrossRef]
  34. Zhao, Y.; Li, M.; Deng, J.; Wang, B. Afforestation affects soil seed banks by altering soil properties and understory plants on the eastern Loess Plateau, China. Ecol. Indic. 2021, 126, 107670. [Google Scholar] [CrossRef]
  35. Hu, F.; Du, H.; Zeng, F.; Song, T.; Peng, W.; Zhang, F. Dynamics of soil nutrient content and microbial diversity following vegetation restoration in a typical karst peak-cluster depression landscape. Acta Ecol. Sin. 2018, 38, 2170–2179. [Google Scholar]
  36. Wang, Z.; Hou, Y.; Fang, H.; Yu, D.; Zhang, M.; Xu, C.; Chen, M.; Sun, L. Effects of plant species diversity on soil conservation and stability in the secondary succession phases of a semihumid evergreen broadleaf forest in China. J. Soil Water Conserv. 2012, 67, 311–320. [Google Scholar] [CrossRef]
  37. De Deyn, G.B.; Cornelissen, J.H.C.; Bardgett, R.D. Plant functional traits and soil carbon sequestration in contrasting biomes. Ecol. Lett. 2008, 11, 516–531. [Google Scholar] [CrossRef] [PubMed]
  38. Shangguan, Z.P. Soil desiccation occurrence and its impact on forest vegetation in the Loess Plateau of China. Int. J. Sustain. Dev. World Ecol. 2007, 14, 299–306. [Google Scholar] [CrossRef]
  39. Bréda, N.; Huc, R.; Granier, A.; Dreyer, E. Temperate forest trees and stands under severe drought: A review of ecophysiological responses, adaptation processes and long-term consequences. Ann. For. Sci. 2006, 63, 625–644. [Google Scholar] [CrossRef]
  40. Duan, L.; Huang, M.; Zhang, L. Differences in hydrological responses for different vegetation types on a steep slope on the Loess Plateau, China. J. Hydrol. 2016, 537, 356–366. [Google Scholar] [CrossRef]
  41. Lavorel, S.; Grigulis, K. How fundamental plant functional trait relationships scale-up to trade-offs and synergies in ecosystem services. J. Ecol. 2012, 100, 128–140. [Google Scholar] [CrossRef]
  42. Passioura, J.B.; Munns, R. Rapid environmental changes that affect leaf water status induce transient surges or pauses in leaf expansion rate. Funct. Plant Biol. 2000, 27, 941–948. [Google Scholar] [CrossRef]
  43. Wei, X.; Liang, W. Multifactor relationships between stand structure and soil and water conservation functions of Robinia pseudoacacia L. in the Loess Region. PLoS ONE 2019, 14, e0219499. [Google Scholar] [CrossRef]
  44. Miura, S.; Yoshinaga, S.; Yamada, T. Protective effect of floor cover against soil erosion on steep slopes forested with Chamaecyparis obtusa (hinoki) and other species. J. For. Res. 2003, 8, 27–35. [Google Scholar] [CrossRef]
  45. Calder, I.R. Dependence of rainfall interception on drop size: 1. Development of the two-layer stochastic model. J. Hydrol. 1996, 185, 363–378. [Google Scholar] [CrossRef]
  46. Davidová, T.; Dostál, T.; David, V.; Strauss, P. Determining the protective effect of agricultural crops on the soil erosion process using a field rainfall simulator. Plant Soil Environ. 2015, 61, 109–115. [Google Scholar] [CrossRef]
  47. Zhou, P.; Luukkanen, O.; Tokola, T.; Nieminen, J. Effect of vegetation cover on soil erosion in a mountainous watershed. Catena 2008, 75, 319–325. [Google Scholar] [CrossRef]
  48. Bruun, E.W.; Petersen, C.T.; Hansen, E.H.; Holm, J.K.; Hauggaard-Nielsen, H. Biochar amendment to coarse sandy subsoil improves root growth and increases water retention. Soil Use Manag. 2014, 30, 109–118. [Google Scholar] [CrossRef]
  49. Marzen, M.; Iserloh, T.; de Lima, J.L.; Fister, W.; Ries, J.B. Impact of severe rain storms on soil erosion: Experimental evaluation of wind-driven rain and its implications for natural hazard management. Sci. Total Environ. 2017, 590, 502–513. [Google Scholar] [CrossRef] [PubMed]
  50. Evett, S.R. Water and energy balances at soil-plant-atmosphere interfaces. In Soil Physics Companion; CRC Press: Boca Raton, FL, USA, 2002; pp. 127–188. [Google Scholar]
  51. Sort, X.; Alcañiz, J.M. Contribution of sewage sludge to erosion control in the rehabilitation of limestone quarries. Land Degrad. Dev. 1996, 7, 69–76. [Google Scholar] [CrossRef]
  52. Brooker, R.W. Plant–plant interactions and environmental change. New Phytol. 2006, 171, 271–284. [Google Scholar] [CrossRef] [PubMed]
  53. Kobayashi, T.; Shimano, K.; Muraoka, H. Effect of light availability on the carbon gain of beech (Fagus crenata) seedlings with reference to the density of dwarf bamboo (Sasa kurilensis) in an understory of Japan sea type beech forest. Plant Species Biol. 2004, 19, 33–46. [Google Scholar] [CrossRef]
  54. Haxeltine, A.; Prentice, I.C. Biome3: An equilibrium terrestrial biosphere model based on ecophysiological constraints, resource availability, and competition among plant functional types. Glob. Biogeochem. Cycles 1996, 10, 693–709. [Google Scholar] [CrossRef]
  55. Wu, H.-F.; Gao, T.; Zhang, W.; Li, G.; Hao, W.-F. Understory vegetation composition and stand are mainly limited by soil moisture in black locust plantations of loess plateau. Forests 2021, 12, 195. [Google Scholar] [CrossRef]
  56. Silvertown, J. Plant coexistence and the niche. Trends Ecol. Evol. 2004, 19, 605–611. [Google Scholar] [CrossRef]
  57. Su, Y.-Z.; Lin, L.-Y.; Yuan, C.-J.; Zhi, Z.-W. Influences of continuous grazing and livestock exclusion on soil properties in a degraded sandy grassland, Inner Mongolia, Northern China. Catena 2005, 59, 267–278. [Google Scholar]
  58. Bloom, A.J.; Chapin, F.S.; Mooney, H.A. Mooney. Resource limitation in plants-an economic analogy. Annu. Rev. Ecol. Syst. 1985, 16, 363–392. [Google Scholar] [CrossRef]
  59. Håkansson, I.; Lipiec, J. A review of the usefulness of relative bulk density values in studies of soil structure and compaction. Soil Tillage Res. 2000, 53, 71–85. [Google Scholar] [CrossRef]
  60. Tilman, D. The ecological consequences of changes in biodiversity: A search for general principles. Ecology 1999, 80, 1455–1474. [Google Scholar] [CrossRef]
  61. Tilman, D. The resource-ratio hypothesis of plant succession. Am. Nat. 1985, 125, 827–852. [Google Scholar] [CrossRef]
  62. Dabney, S.M.; Delgado, J.A.; Reeves, D.W. Using winter cover crops to improve soil and water quality. Commun. Soil Sci. Plant Anal. 2001, 32, 1221–1250. [Google Scholar] [CrossRef]
Forests 15 01341 g001
Figure 1. Map of Xindiangou watershed, Yulin City, Shaanxi province. (a) Elevation and plot distribution map. (b) Survey design in plots.
Figure 1. Map of Xindiangou watershed, Yulin City, Shaanxi province. (a) Elevation and plot distribution map. (b) Survey design in plots.
Forests 15 01341 g001
Forests 15 01341 g002
Figure 2. Preliminary PLS-SEM model which reflects the correlation between variables. Leaf dry weight: FLT; height difference: HD; top layer of vegetation height: TVH; canopy width of top layer of vegetation: CTV; vegetation coverage of the top floor: VCT; number of canopies: CN; leaf thickness: LT; leaf tissue density: LTD; leaf area: LA; specific leaf area: SLA; vegetation coverage: VC; average plant height: PH; number of vegetation layers: VL; soil water content: SWC; soil bulk density: BD; maximum soil water capacity: BW; soil porosity: SP; Shannon–Wiener index: SHA; Simpson index: SIM; Margalef index: MAR; Pielon index: PIE; Gleason index: GLE; total runoff and sediment weight: TRS; total sediment weight: TS; clear water coefficient: CW; depth of clear water: DW; muddy water coefficient: MWC; depth of muddy water: DMW.
Figure 2. Preliminary PLS-SEM model which reflects the correlation between variables. Leaf dry weight: FLT; height difference: HD; top layer of vegetation height: TVH; canopy width of top layer of vegetation: CTV; vegetation coverage of the top floor: VCT; number of canopies: CN; leaf thickness: LT; leaf tissue density: LTD; leaf area: LA; specific leaf area: SLA; vegetation coverage: VC; average plant height: PH; number of vegetation layers: VL; soil water content: SWC; soil bulk density: BD; maximum soil water capacity: BW; soil porosity: SP; Shannon–Wiener index: SHA; Simpson index: SIM; Margalef index: MAR; Pielon index: PIE; Gleason index: GLE; total runoff and sediment weight: TRS; total sediment weight: TS; clear water coefficient: CW; depth of clear water: DW; muddy water coefficient: MWC; depth of muddy water: DMW.
Forests 15 01341 g002
Forests 15 01341 g003
Figure 3. Statistical picture of runoff and sediment variables. Total runoff and sediment weight: TRS; total sediment weight: TS; clear water coefficient: CW; depth of clear water: DW; muddy water coefficient: MWC; depth of muddy water quantity: DMW.
Figure 3. Statistical picture of runoff and sediment variables. Total runoff and sediment weight: TRS; total sediment weight: TS; clear water coefficient: CW; depth of clear water: DW; muddy water coefficient: MWC; depth of muddy water quantity: DMW.
Forests 15 01341 g003
Forests 15 01341 g004
Figure 4. Heat map of RF, red indicates accepted variables, blue indicates rejected variables, * indicate the significance of variables. Leaf dry weight: FLT; height difference: HD; top layer of vegetation height: TVH; canopy width of top layer of vegetation: CTV; vegetation coverage of the top floor: VCT; number of canopies: CN; leaf thickness: LT; leaf tissue density: LTD; leaf area: LA; specific leaf area: SLA; vegetation coverage: VC; average plant height: PH; number of vegetation layers: VL; soil water content: SWC; soil bulk density: BD; maximum soil water capacity: BW; soil porosity: SP; Shannon-Wiener index: SHA; Simpson index: SIM; Margalef index: MAR; Pielon index: PIE; Gleason index: GLE; total runoff and sediment weight: TRS; total sediment weight: TS; clear water coefficient: CW; depth of clear water: DW; muddy water coefficient: MWC; depth of muddy water quantity: DMW.
Figure 4. Heat map of RF, red indicates accepted variables, blue indicates rejected variables, * indicate the significance of variables. Leaf dry weight: FLT; height difference: HD; top layer of vegetation height: TVH; canopy width of top layer of vegetation: CTV; vegetation coverage of the top floor: VCT; number of canopies: CN; leaf thickness: LT; leaf tissue density: LTD; leaf area: LA; specific leaf area: SLA; vegetation coverage: VC; average plant height: PH; number of vegetation layers: VL; soil water content: SWC; soil bulk density: BD; maximum soil water capacity: BW; soil porosity: SP; Shannon-Wiener index: SHA; Simpson index: SIM; Margalef index: MAR; Pielon index: PIE; Gleason index: GLE; total runoff and sediment weight: TRS; total sediment weight: TS; clear water coefficient: CW; depth of clear water: DW; muddy water coefficient: MWC; depth of muddy water quantity: DMW.
Forests 15 01341 g004
Forests 15 01341 g005
Figure 5. Secondary model based on the CCA results.
Figure 5. Secondary model based on the CCA results.
Forests 15 01341 g005
Forests 15 01341 g006
Figure 6. The final fitting model describing the relationship between the variables. Red arrows indicate a positive impact and blue arrows indicate a negative impact; the thickness of the arrow indicates the weight. Leaf dry weight: FLT; height difference: HD; vegetation coverage of the top floor: VCT; leaf tissue density: LTD; specific leaf area: SLA; vegetation coverage: VC; average plant height: PH; number of vegetation layers: VL; soil bulk density: BD; maximum soil water capacity: BW; Simpson index: SIM; Gleason index: GLE; total runoff and sediment weight: TRS; depth of clear water: DW.
Figure 6. The final fitting model describing the relationship between the variables. Red arrows indicate a positive impact and blue arrows indicate a negative impact; the thickness of the arrow indicates the weight. Leaf dry weight: FLT; height difference: HD; vegetation coverage of the top floor: VCT; leaf tissue density: LTD; specific leaf area: SLA; vegetation coverage: VC; average plant height: PH; number of vegetation layers: VL; soil bulk density: BD; maximum soil water capacity: BW; Simpson index: SIM; Gleason index: GLE; total runoff and sediment weight: TRS; depth of clear water: DW.
Forests 15 01341 g006
Table 1. The details of the community structure.
Table 1. The details of the community structure.
ItemSpecies Level
Arboreal forestTrees, shrubs, and herbs
Mingled forestMulti-species trees, shrubs, and herbs
ShrubberyShrubs and herbs
Mingled shrubberyMulti-species shrubs and herbs
Arboreal forest (without shrubs)Trees and herbs
Table 2. CCA analytical results of each path.
Table 2. CCA analytical results of each path.
Variables GroupCanonical Correlations 1Eigenvaluep-ValueWilk’sDF
Vegetation diversity vs. Soil properties0.6980.950.1720.32732
Vegetation diversity vs. Grass properties0.96413.280 ***0.0335
Runoff and sediment vs. Grass properties0.6730.830.033 *0.33825
Canopy properties vs. Grass properties0.999426.280 ***080
Canopy properties vs. Runoff and sediment0.9458.340 ***0.05540
Canopy properties vs. Soil properties0.8733.210.001 ***0.03170
Soil properties vs. Grass properties0.7971.740.009 **0.0856
Soil properties vs. Runoff and sediment0.6860.880.023 *0.28528
* indicates p < 0.05; ** indicates p < 0.01; *** indicates p < 0.001.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Duan, G.; Leng, C.; Zhang, Z.; Zheng, C.; Wen, Z. Quantitative Study on the Effects of Vegetation and Soil on Runoff and Sediment in the Loess Plateau. Forests 2024, 15, 1341. https://doi.org/10.3390/f15081341

AMA Style

Duan G, Leng C, Zhang Z, Zheng C, Wen Z. Quantitative Study on the Effects of Vegetation and Soil on Runoff and Sediment in the Loess Plateau. Forests. 2024; 15(8):1341. https://doi.org/10.3390/f15081341

Chicago/Turabian Style

Duan, Gaohui, Chunqian Leng, Zeyu Zhang, Cheng Zheng, and Zhongming Wen. 2024. "Quantitative Study on the Effects of Vegetation and Soil on Runoff and Sediment in the Loess Plateau" Forests 15, no. 8: 1341. https://doi.org/10.3390/f15081341

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop