Global analysis of support practices in USLE-based soil erosion modeling

Abstract

Support practices (SPs) influence the magnitude of soil loss and can be readily influenced by human interventions to mitigate soil loss. The SPs factor is expressed as the P-factor in the widely used soil erosion model – the universal soil loss equation (USLE) – and its revised version. Although the effects of SPs on soil erosion are well recognized, the quantification of the P-factor for soil loss modeling remains challenging. This limitation of the P-factor particularly restricts the applicability of USLE-based models at large scales. Here, we analyzed the P-factor values in USLE-based models from 196 published articles. The results were as follows: (a) an increasing trend in the number of studies has been observed in recent years, especially at large scales; (b) the P-factor values for paddy fields, orchards, and croplands were 0.16 ± 0.15, 0.47 ± 0.12, and 0.49 ± 0.21, respectively, and in terms of different types of SPs, the P-factor values for terracing, contouring, and strip-cropping were 0.28 ± 0.18, 0.52 ± 0.24, and 0.49 ± 0.28, respectively; (c) various methods have been developed for P-factor qualification, although the methods that consider SP conditions were most frequently used in studies with relatively smaller areas (< 100 km²), suggesting that USLE-based models are in need of improvement via the quantification of the P-factor, particularly with respect to the regional and global scale; and (d) further improvements of the P-factor for soil erosion modeling should concentrate on building P-factor datasets at the regional level according to data on the effectiveness of SPs on soil loss control based on field experiments in published articles, using advanced image processing techniques based on higher-resolution satellite imagery and developing proxy indicators for P-factors at large scales.

Keywords

USLE RUSLE soil erosion support practices P-factor

I Introduction

Water-driven soil erosion is the most widespread form of soil degradation worldwide (Garcia-Ruiz et al., 2017; Maetens et al., 2012), and this erosion has been a major threat to soil quality and productivity, sustainable agriculture, crop production, and carbon stocks across the globe since the beginning of agriculture (Li and Fang, 2016). Soil erosion assessments and mapping of erosion-prone areas serve as a scientific base for soil conservation and watershed management (Patil, 2018). Therefore, the need has arisen to quantify the soil erosion rates, as they are of paramount importance to stakeholders and decision-makers during conservation planning processes.

Various soil erosion assessment models have been developed (De Vente et al., 2013; Karydas et al., 2014; Li et al., 2017a); however, certain disadvantages, such as model complexity and the high demand for input data by most models, reduce the applicability of these models to the large scales (Vente and Poesen, 2005). The universal soil loss equation (USLE) (Wischmeier and Smith, 1978) and its revised version (RUSLE) (Renard et al., 1997) have been widely applied to assess soil erosion in the past and predict soil erosion in the future at different spatial scales (Borrelli et al., 2017b; Panagos et al. 2015b) due to their simple structure and empirical basis (Naipal et al., 2015). Among the factors in the USLE and RUSLE, the support practice (SP) factor (P-factor) is important because many SPs have significant effects on reducing soil erosion (Wei et al., 2016; Xiong et al., 2018). The P-factor is defined as the ratio of soil loss with a specific SP to the corresponding loss with upslope and downslope tillage (Renard et al., 1997; Wischmeier and Smith, 1978). The P-factor accounts for control practices (e.g. contouring, strip-cropping, and terracing) that affect erosion potential of runoff by modifying the flow pattern, grade, or direction of surface runoff and by reducing the velocity, amount, and rate of runoff on the soil surface (Renard et al., 1997; Wischmeier and Smith, 1978). The lower the P-factor value is, the better the practice is for controlling soil erosion (Panagos et al., 2015a). The P-factor constitutes the most uncertain factor in the model (Arnhold et al., 2014; Ligonja and Shrestha, 2015; Panagos et al., 2015a; Renard et al., 1997).

The extensive use of the USLE-based model on large basins has been reported by many researchers (Correa et al., 2016; Terranova et al., 2009; Xiao et al., 2015). Moreover, in recent years, there has been growing interest in the assessment of soil erosion (Zhuang et al., 2015), including the effects of soil erosion on ecosystem services (Guerra et al., 2014; Jiang et al., 2018; Peng et al., 2017) and biogeochemical cycles, such as the carbon cycle (Doetterl et al., 2012; Van Oost et al., 2007) at large scales, and even the global scale. However, there is substantial uncertainty in the estimation of large-scale soil erosion rates by USLE-based modeling (Borrelli et al., 2017b; Doetterl et al., 2012; Naipal et al., 2015). Although many studies have quantified P-factor values for different SPs (Panagos et al., 2015a; Renard et al., 1997; Shin, 1999; Vezina et al., 2006; Wischmeier and Smith, 1978), little information can be found in the literature about the applicability and appropriate use of P-factor (Arnhold et al., 2014), and a global reference is not available because erosion control is a local activity (Panagos et al., 2015a). Moreover, the method of quantifying the P-factor for soil erosion assessment has not been clarified. To address this issue, a thorough review of different methods of determining the P-factor is needed. Such a study should compare different methods and consider the characteristics of each method. The results of such a method comparison should be beneficial for users to help them choose an appropriate method, and it can also promote a better understanding of future needs.

This paper aimed to compare methods of determining the P-factor value for soil erosion modeling based on a survey of the literature. We selected studies that used USLE-based models and considered the P-factors, then we compared and discussed them in terms of P-factor values, P-factor quantification methods, data requirements, and study area size. We discuss the limitations of P-factor quantification and provide a better understanding of P-factors worldwide.

II Data collection and analysis

To compile P-factor information found within scientific journal articles, conference papers, and books, we reviewed the Institute for Scientific Information (ISI) Web of Science and Google Scholar databases using two keywords: USLE and RUSLE. We retrieved a total of 1516 articles published worldwide, and the number of publications has increased annually (Figure 1). At the country/territory level, studies applying the USLE/RUSLE model were conducted in 95 countries. In the productivity ranking of countries, the USA ranked first (219 studies), China ranked second (124 studies), and Italy ranked third (102 studies). However, many studies focused on the analysis of other input factors, such as the rainfall erosivity factor (Capolongo et al., 2008; Sanches Oliveira et al., 2013) and the slope length factor (Liu et al., 2011), or the theoretical basis and main principles of the model (Todisco et al., 2015), which did not provide information about the P-factor.

Figure 1.

Worldwide distribution of the studies concerning USLE-based models.

The studies were organized within a database in which useful data for our analysis were registered. The data were selected in accordance with the following criteria: only data with clear sites, P-factor values, and practice descriptions were selected. If articles involved studies that focused on the same area and used the same P-factor value, only one study was selected and the others were excluded from the database. We retained 492 studies in the database (Figure 2). The USLE/RUSLE was used in eight studies to research soil erosion at the global scale, and other studies were conducted in 82 countries. The top three countries included China (82 studies), India (51 studies), and Italy (36 studies) (Figure 2).

Figure 2.

Distribution of the studies in this database.

The following information was collected for the studies: (a) spatial location (coordinates); (b) study surface area (km²); (c) land-use type (crop type); (d) P-factor value; (e) P-factor calculation method; and (f) corresponding SPs. In the 196 studies that had information about SPs, the corresponding P-factors were qualified, and in the other 296 studies in which SPs were not applied or lacking information about SPs, the P-factor values were set to 1. The geographical distribution of the 196 studies in which SPs were applied was mapped (Figure 2; Supplemental material). In the following, we analyze the P-factors based on these 196 studies.

Figure 2 shows the spatial distribution of the sites corresponding to the 196 studies except for the global-scale studies (Ito, 2007; Pham et al., 2001; Scherer and Pfister, 2015; Yang et al., 2003) and European-scale studies (Panagos et al., 2015b). The studies in most of these countries, including China (46 studies) and India (31 studies) as well as some European countries (22 studies), were conducted under conditions of conservation agriculture, in which the proportion of cropland area under conservation agriculture was reported to the Food and Agriculture Organization of the United Nations (FAO) (Borrelli et al., 2017b).

The unique feature of the data is the variability in environmental conditions, and the methods used to calculate the input factors of the USLE-based models were also important. For instance, the size of the study areas varied by more than hundreds of orders of magnitude. Similarly, the methods used to determine P-factor values had various levels of complexity. In addition, there was variability among studies in the application of the types of SPs, which increased the difficulty of analyzing the results and general conclusions. Many authors did not provide information on the slope gradient, distribution of SPs, or even the types of SPs. In the following, we report the size of the study areas, types of SPs, and methods for P-factor value calculation, and present the results of a multivariate analysis that included all factors.

III Methods for quantifying P-factor values

Various methods for quantifying P-factors have been developed based on different research objectives and data availability. The P-factor values in most studies were obtained from previous research works and assigned to land-cover classes, and empirical values were adopted for P-factors for different land uses (Lu et al., 2013; Xu et al., 2013; Yao et al., 2016; Zeng et al., 2017). Many studies have reported various tables (Cai et al., 2000; Hurni, 1982; Wischmeier, 1975; Wischmeier and Smith, 1978; Yang, 1999) and formulas (Medeiros et al., 2016; Renard et al., 1997; Villarreal et al., 2016; Wang et al., 2016; Wener, 1981) for P-factor values for the different SPs adopted under different environmental contexts (Panagos et al., 2015a). In addition, in some studies, the P-factor was treated either as a subfactor of the cover-management factor (C-factor) or as a new combined CP-factor (Bhandari et al., 2015; Fujaco et al., 2016; Hurni, 1982; Liu et al., 2016; Ozhan et al. 2005). Furthermore, other researchers have developed proxy indicators for the P-factor (Borrelli et al., 2017b; Scherer and Pfister, 2015).

1 P-factor values for different SPs based on field experiments

In the USLE, Wischmeier and Smith (1978) proposed P-factor values for three major SPs (contouring, strip-cropping, and terracing) in cultivated lands (Table 1), in which soil loss measurements on runoff plots in the USA were used to determine the effectiveness of conservation measures, and the results are the most widely used (Kumar et al., 2014; Plangoen et al., 2013; Yang et al., 2009). Certain studies used similar field experiments to calculate the P-factor values and developed tables that were more suitable for the local environmental conditions. The Korea Institute of Construction Technology also reported a table (Table 1) proposing P-factor values for the three major SPs based on a study in South Korea (KICT, 1992), and this table has been used in some studies for assessing soil erosion for cultivated lands under the three SPs (Jang et al., 2015; Karamage et al., 2016a, 2016b, 2016c, 2017; Lee et al., 2009). In addition, Yang (1999) provided the P-factor values for some SPs (Table 1) based on field experiments and studies in southwestern China, and these values were used in assessments of soil erosion in studies in China (Chen et al., 2017b; Xiao et al., 2015; Xu et al., 2008, 2011, 2017; Yang, 2002). Cai et al. (2000) calculated the P-factor values for terracing and contouring on croplands and orchards based on field experiments in China (Cai et al., 2000; Li and Wei, 2014; Tang et al., 2015). Similarly, other previous studies suggested the P-factor values of some SPs in Ethiopia (Hurni, 1985; Lema et al., 2016; Tamene et al., 2014) and Tanzania (Kabanza et al., 2013).

Table 1.

P-factor values under different conservation support practices.

Land slope (%)	Contouring	Strip- cropping	Terracing
Wischmeier and Smith, 1978 (USA)
1–2	0.60	0.30	0.12
3–8	0.50	0.25	0.10
9–12	0.60	0.30	0.12
13–16	0.70	0.35	0.14
17–20	0.80	0.40	0.16
21–25	0.90	0.45	0.18
KICT, 1992 (South Korea)
0–7	0.55	0.27	0.10
7–11.3	0.60	0.30	0.12
11.3–17.6	0.80	0.40	0.16
17.6–26.8	0.90	0.45	0.18
>26.8	1	0.50	0.20
Yang, 1999 (China)
Contouring		0.5532
Ridge		0.1836
Buffer strips + soil amendment		0.3745
Buffer strips		0.4962
Level terrace		0.0316
Land slope (%)	Arable land
Dunne and Leopold, 1978
2–7	0.25
8–12	0.3
13–18	0.4
19–24	0.45

2 P-factor values based on slopes

Many studies modified the P-factor values proposed by Wischmeier and Smith (1978) based on local SPs (Rawat et al., 2016). However, due to the difficulty of obtaining detailed information on SPs, some studies have calculated P-factor values based only on the slope gradient. For instance, Dunne and Leopold (1978) proposed P-factor values that were averaged for SPs and calculated based on slope ranges (Table 1).

However, the P-factor in the USLE may not be suitable for accurately describing the conservation effect because of the small size of runoff plots. In the RUSLE, the P-factor equations have been developed based on plot studies as well as watershed observations and calculations using detachment and transport theory (Renard et al., 1997); therefore, these equations have a broader applicability. In the RUSLE, many equations have been developed to calculate P-factor values for contouring, and they include combinations of strips, such as buffer and filter strips, and terracing. The effectiveness of contouring varied by ridge height and field slope steepness, the ridge height was classified into five groups. The P-factor values for contouring are calculated as follows (Renard et al., 1997):

P_{0} = {\begin{matrix} a \times {(sin m - sin θ)}^{4} + P_{m b} & sin θ < sin m \\ b \times {(sin θ - sin m)}^{1.5} + P_{m b} & sin θ \geq sin m \\ 1 & sin θ \geq sin e \end{matrix}

where P ₀ is the P-factor for on-grade contouring, θ is the slope angle for which a value of P ₀ is desired, m is the slope angle at which contouring has its greatest effectiveness, e is the slope angle above which contouring is ineffective, and P_mb is the minimum P-value for a given ridge height with base conditions. The coefficients a and b also vary with the ridge height. The values for coefficients in equation (1) are given.

The equation used to compute P-factor values for off-grade contouring in the RUSLE is:

P = P_{0} + (1 - P_{0}) \times {(sin θ_{f} / sin θ)}^{0.5}

where P is the P-factor for off-grade contouring, P ₀ is the P-factor for on-grade contouring (as computed by equation (1)), θ_f is the slope angle along the furrows, and θ is the average slope angle of the field (°).

In the RUSLE, the P-factor for strip-cropping considers the cropland cover-management conditions, and the P-factor for terracing considers the terrace spacing, horizontal terrace interval, and terrace grade. However, these formulas in the RUSLE are rarely applied because of the difficulty in obtaining detailed information on SPs. An alternative approach for approximating P-factor values based on empirical equations has been developed (Gao et al., 2016; Lufafa et al., 2003; Munir et al., 2000; Villarreal et al., 2016; Wang et al., 2016). For instance, the Wenner method assumes that P-factor values are linked to topographical features (Lufafa et al., 2003; Wener, 1981), and the equation is:

P = 0.2 + 0.03 \times S

where S is the mean slope grade (%). In large watershed scales, land-use maps do not show differences resulting from SPs, such as terracing and contouring (Napoli et al., 2016). Therefore, using slope gradients as inputs, this method is commonly used to determine P-factor values (Clark et al., 2015; Fu et al., 2005, 2011; Khosrokhani and Pradhan, 2014; Li et al., 2017b; Mutua and Klik, 2004; Napoli et al., 2016; Pope and Odhiambo, 2014; Ricker et al., 2008; Schnitzer et al., 2013; Wang et al., 2014; Wu et al., 2011, 2016).

Equation (4) presents another methodology to determine P-factor values based on the slope angle (θ) (Medeiros et al., 2016):

P = {\begin{array}{l} 0.6 & 0 \leq θ \leq 5 % \\ 0.69947 - 0.08991 θ + 0.01184 θ^{2} - 0.00035 θ^{3} & 5 % < θ \leq 20 % \\ 1 & θ > 20 % \end{array}

3 Combination of the C-factor and P-factor

The C-factor and P-factor were combined in a single value as previously performed (Brooks et al., 1996; Wischmeier, 1975). In the modified USLE (MUSLE), the C-factor and P-factor used in the USLE have been replaced by a vegetation management (VM) factor (Wischmeier, 1975). For forests, the VM-factor is calculated based on canopy height and cover, ground cover, and bare ground with fine roots. For plantation systems, the P-factor was mostly treated either as a subfactor of C-factor (Dissmeyer and Foster, 1980) or as a new combined CP-factor (Brooks et al., 1996; Hurni, 1982). Liu et al. (2016) found that the CP-factor value was highly correlated with plant cover in rubber plantations; therefore, they selected plant cover as an indicator to present the effects of the CP-factor. The relationship between plant cover and the CP-factor was described as follows (Liu et al., 2016):

CP = 0.04 e^{- 0.028 P C}

where PC = plant cover (%).

Furthermore, Andriyani et al. (2017) proposed that the C-factor and P-factor can be multiplied together to obtain the CP-factor as follows:

CP = (CC \times SC \times SR) \times CM

where CC is the crop cover subfactor, SC is the soil surface cover subfactor, SR is the soil roughness subfactor, and CM is the crop management subfactor (Andriyani et al., 2017).

4 Proxy indicators for the P-factor

Proxy indicators for the P-factor have been used in some studies because of the particular research objectives and data availability limitations. The regional population density and gross domestic product (GDP) can reflect the intensity of human activity, which is used to characterize the role of SPs in soil erosion research in the northern Loess Plateau (Liu and Liu, 2010). Scherer and Pfister (2015) used the high human development index (UNDP, 2014) and the score for pesticide regulation (Hsu et al., 2014) to reflect the role of SPs, and the USLE was used to model the spatial difference of phosphorus emissions in agriculture. Villarreal et al. (2016) refined the P-factor to express the vulnerability of soils to vehicle disturbances, and the P-factor was calculated as the difference in saturated hydrologic conductivity between disturbed and undisturbed soils, which was then scaled up to maps of vehicle disturbances digitized from aerial photography (Villarreal et al., 2016). Moreover, the Land Degradation Assessment in Drylands (LADA) project estimated P-factors using a management index related to overall crop performance and developed a global map for the parameter (Nachtergaele et al., 2010).

5 Quantification of P-factors based on an analysis of satellite imagery

P-factor values are commonly evaluated by experts via field observations or land-use maps. With the development of remote sensing (RS) and geographic information system (GIS) techniques, analysis of very high spatial resolution (VHR) satellite imagery, such as Pleiades imagery and QuickBird imagery, are also gaining attention (Karydas et al., 2009; Wang et al., 2016). Karydas et al. (2009) used an object-oriented classification method for terraces to quantify site-specific P-factors based on QuickBird imagery, and Wang et al.(2016) used the angle between the ridge direction and slope aspect to calculate the P-factor value for contouring according to the visual interpretation of the Pleiades imagery. Earth observation data obtained by VHR satellite imagery, such as QuickBird imagery, and advanced image processing techniques, such as object-oriented analysis, have provided new insights for mapping landscape features, such as terraces, and represents a basis for assessing P-factors (Karydas et al., 2009; Wang et al., 2016; Yang et al., 2009). P-factors can be derived via an object-oriented analysis of QuickBird imagery, which allows P-factors to be quantified via a thematic map instead of assigning uniform P-factor values. This method can further improve the resulting soil erosion mapping.

6 Application of the methods for P-factor quantification

The size of the study areas ranged from the plot scale (1–100 m²) to the global scale, and the area of most studies (76%) ranged from 10 km² to 1×10⁵ km², which is consistent with the typical size of many monitored catchments (García-Ruiz et al., 2015). Figure 3 shows the size of the studies combined with information on the year of the paper published based on the 196 studies in which SPs were applied. An increasing trend in the number of studies with P-factor values in the models has been observed, and more studies have focused on soil erosion estimates based on the USLE/RUSLE model at a large scale in recent years.

Figure 3.

Distribution of the number of studies with different study area sizes.

The methods of calculating P-factor values were classified into six categories: (a) P-factors calculated by a combination of C-factors and P-factors (M1); (b) P-factor values calculated with slope gradients based on SPs information, such as the type of SPs (M2); (c) P-factor values calculated with slope gradients without information about SPs (M3); (d) an empirical uniform P-factor value adopted based on the SPs according to previous research works or expert knowledge (M4); (e) an empirical uniform P-factor value adopted for different land use according to previous research works or expert knowledge (M5); and (f) P-factor values calculated based on proxy indicators (M6). Figure 4(a) shows the number of studies with different methods in every year. The P-factor values in the majority of studies adopted an empirical uniform value according to previous research works (M4 and M5) and were calculated using the slope gradients of land (M2 and M3) (Figure 4(a)). A greater number of studies that used the M2 and M4 methods, calculating the P-factor values according to the condition of SPs, had a smaller study area size (<100 km²) (Figure 4(b)), suggesting that the P-factors were considered more fully at smaller-scale studies.

Figure 4.

Distribution of the number of studies with (a) different P-factor-value calculation methods and (b) different study area sizes. M1-6 represent different categories of the methods of calculating P-factor values.

IV P-factor values for USLE-based soil loss modeling

The main methods for calculating the P-factor values used in this database were introduced in the previous section. Here, we analyzed the P-factor values in the database that did not include studies based on national and global scales. The P-factor values of SPs were reported in 103 case studies (Figure 5), and for the other cases, detailed information (the type and location of SPs) was lacking, and the P-factor values were mostly calculated by slope and land use. The SPs were applied mostly to arable lands, including croplands and paddy fields, and SPs were also applied for orchards and pasture in some studies. The main SPs included terracing (60 studies), contouring (59 studies), strip-cropping (24 studies), and mulching (nine studies) (Figure 5). The SPs were typically used in combination; for example, contouring almost always accompanied strip-cropping. The combination of more than one type of practice was applied in 44 studies.

Figure 5.

Distribution of case studies and the corresponding support practices (SPs).

The P-factor values in 67 studies adopted empirical uniform P-factor values for different land uses according to previous research works or expert knowledge (M5 method). We found that there was a high degree of variability in P-factor values because of the different land-use types (Figure 6(a)). The P-factor values were smaller for paddy fields (0.16 ± 0.15) than for orchards (0.47 ± 0.12) and croplands (0.49 ± 0.21) (Figure 6(a)). P-factor values were reported for paddy fields in 19 study areas located in China, Korea, Thailand, Malaysia, the Philippines, India, and Vietnam, which are the major rice-producing countries worldwide.

Figure 6.

Boxplots showing the range of P-factor values: (a) P-factor values for paddy fields, orchards, and croplands according to the M5 method; and (b) P-factor values of the main support practices (SPs) for croplands according to the M4 method.

The P-factor values in 49 studies were given uniform values for croplands under different SPs (M4 method). Boxplots (Figure 6(b)) show the range of P-factor values for croplands under terracing, contouring, and strip-cropping. The mean values for strip-cropping had the widest range (mean = 0.49, standard deviation (SD) = 0.28). The smallest P-factor value was recorded from terracing (mean = 0.28, SD = 0.18), and the mean value reported for contouring was 0.52 (SD = 0.24). The results show that terracing was more effective than contouring and strip-cropping and generally agree with those of published studies (Renard et al., 1997; Wischmeier and Smith, 1978).

Furthermore, the P-factor values were calculated for terracing, contouring, and strip-cropping based on the slope gradient of croplands according to the tables that included P-factor values obtained in previous studies (M2 method), such as Wischmeier and Smith (1978) and KICT (1992) (Table 1). Figure 7(a) shows that the mean P-factor values in the 20 studies vary based on the slope gradient. However, in some studies with a large scale, the information on the location, area, and type of SPs cannot be diagnosed via a land-use map, and the P-factor values for arable lands were calculated based on the slope gradient (M3 method), such as in the table generated using the Dunne and Leopold method (Dunne and Leopold, 1978; Litschert et al., 2014) (Table 1). Figure 7(b) shows the range of P-factor values for croplands whose slope gradients vary in 17 studies that lacked information. A high degree of variability in P-factor values was observed at the same slope gradient, although the effectiveness of SPs generally decreased as the terrain gradient increased (Figure 7(a) and (b)). The slope-based approach of the Wenner method (equation (3)) was also used to calculate the P-factor values for croplands in 15 studies.

Figure 7.

Range of P-factor values based on slope gradients: (a) P-factor values of the main support practices (SPs) for croplands according to the M2 method; and (b) P-factor values for croplands according to the M3 method.

V P-factor for USLE-based modeling at the national and global scale

USLE-based models have been widely used for soil erosion estimates at the catchment scale and regional scale based on the aforementioned studies. In recent years, USLE-based models have been applied to assess soil erosion at the national or global scale (Table 2). To estimate global soil erosion, the USLE model was first applied on a global scale by Pham et al. (2001); in their study, a P-factor value of 0.5 was introduced for complete agricultural lands (dry crops and paddies) and mixtures of agricultural land with forest or grassland. In addition, Yang et al. (2003) and Ito (2007) considered the importance of human influence on soil erosion control, and due to the lack of details concerning SPs on a global scale, they also identified P-factor values by land-cover type, with a value of 0.5 assigned to complete arable lands and 0.8 assigned to mixtures of arable land with forest or grassland. However, Doetterl et al. (2012) showed that compared with other RUSLE factors (LS-factor, R-factor, and C-factor), the P-factor did not contribute significantly to variations in soil erosion on scales ranging from continental to global; thus, the P-factor could be ignored. In a different study, Naipal et al. (2015) did not consider the P-factor as they had insufficient or no data concerning the P-factor on a global scale. Without considering the P-factor, Borrelli et al. (2017(b)) predicted values for reference years and estimated a conservative scenario for 54 countries in which information concerning the implementation of conservation agriculture was reported to the FAO. For these areas, Borrelli et al. (2017b) assumed a 45% reduction in soil erosion as opposed to conventional tillage (P-factor value = 0.45). P-factors have been poorly considered in soil erosion estimates at the global scale.

Table 2.

Overview of national and global soil erosion risk assessments based on the USLE/RUSLE models.

Study area	P-factor value	References
Europe	No P-factor	(Van der Knijff et al., 2000)
Italy	No P-factor	(Grimm et al., 2003; Van der Knijff et al., 1999; Van Rompaey et al., 2003)
Europe	No P-factor	(Grimm et al., 2001)
Australia	No P-factor	(Lu et al., 2003)
South Africa	No P-factor	(Le Roux et al., 2008)
Spain	No P-factor	(De Vente et al., 2008)
Slovakia	No P-factor	(Cebecauer and Hofierka, 2008)
Europe	No P-factor	(Podmanicky et al., 2011)
South Korea	Paddy field: 0.5; cropland: 0.5	(Park et al., 2011)
Spain	No P-factor	(Martin-Fernandez and Martinez-Nunez, 2011)
USA	No P-factor	(Segura et al., 2014)
China	No P-factor	(Rao et al., 2014)
Europe	No P-factor	(Panagos et al., 2014)
Europe	The P-factor is estimated based on observations in the European Union	(Panagos et al., 2015b)
Europe	No P-factor	(Bosco et al., 2015)
Mediterranean Europe	No P-factor	(Guerra et al., 2016)
Australia	P-factor values were calculated based on the slope	(Teng et al., 2016)
Hungary	No P-factor	(Pasztor et al., 2016)
Greece	No P-factor	(Panagos et al., 2016)
Italy	P-factor values were calculated based on the slope	(Borrelli et al., 2016)
Uganda	The P-factor value of 0.75 was retrieved from the LADA project	(Karamage et al., 2017)
Italy	The P-factor value of 0.75 were calculated based on data received through the observation of 500 clear-cut areas	(Borrelli et al., 2017a)
Global	Paddy field: 0.5; cropland: 0.5; mixture: 0.5	(Pham et al., 2001)
Global	Paddy field: 0.5; cropland: 0.5; mixture: 0.8	(Yang et al., 2003)
Global	Paddy field: 0.5; cropland: 0.5; mixture: 0.8	(Ito, 2007)
Global	No P-factor	(Van Oost et al., 2007)
Global	No P-factor	(Doetterl et al., 2012)
Global	P-factor values were calculated based on the slope and indicator	(Scherer and Pfister, 2015)
Global	No P-factor	(Naipal et al., 2015)
Global	P-factor = 0.45 for croplands in countries under conversation agriculture	(Borrelli et al., 2017b)

Previous studies have shown that engineering techniques can reduce the soil loss rate by up to 86% based on a global analysis of runoff plots (Xiong et al., 2018), and Chen et al. (2017a) reported that terracing resulted in a 53.0% mean reduction in sediment in China. These results show that these SPs contributed significantly to the reduction in soil erosion of local arable land, especially in erosion-sensitive regions, such as the Mediterranean and loess regions. Conservation agriculture covers approximately 15% of global cropland (1.6 million km²), and soil erosion of cropland accounts for approximately 50% of the total predicted soil erosion (Borrelli et al., 2017b). SPs are applied mostly in areas that are susceptible to soil erosion, such as loess regions, thereby resulting in significant reductions in absolute soil loss. If the influence of SPs is ignored, the calculated soil erosion rate will be much greater than its actual value on arable lands. As such, the influence of SPs should be considered in assessments of global soil erosion. However, determining differences in SPs using a land-use map is difficult, and the location of SPs cannot easily be identified based on land-use maps for global-scale studies; hence, the P-factor is poorly considered. This problem should be resolved to improve the results of soil erosion mapping.

At the national scale, the P-factors have not been fully considered for soil erosion estimates (Table 2). However, countries in the European Union have provided new insights for P-factor quantification. Panagos et al. (2015a) modeled SPs that reduce soil erosion and presented the areas in Europe where those SPs were implemented based on more than 226,000 observations from the land use/cover area frame statistical survey (LUCAS) carried out in the European Union. The results were presented at pixel level, then the data was aggregated at the regional level according to the application of SPs (Panagos et al., 2015a). Many studies have conducted field experiments to research the effectiveness of SPs on soil loss control (Chen et al., 2017a; Xiong et al., 2018); thus, at catchment or regional level, a larger number of field observations may be collected, which can be used to build a P-factor dataset at the regional level according to the method of Panagos et al. (2015a).

VI Limitations and future improvements for the quantification of P-factors

Our literature review and analysis of P-factor values showed that ignoring P-factors introduces uncertainty, and other limitations regarding current research were also identified. First, the effects of SPs on soil erosion had not been fully considered for other land-use types. SPs were applied to arable lands in a majority of studies using USLE-based models. However, Chen et al. (2017a) performed a meta-analysis of 636 observations involving terracing in China and found that terracing was used within different land-use types, including cropland, shrubland, and forestland. Panagos et al. (2015a) evaluated more than 226,000 observations in the European Union and found that stone walls were applied in all land-use types except artificial land and water bodies, and grass margins were widely used in agricultural areas, pastures, and heterogeneous agricultural areas. Second, studies that adopted an empirical uniform P-factor value for lands introduced inaccuracy to the P-factor values, and the SPs should be located on a land-use map. Third, P-factor values calculated according to previous research works or expert knowledge were not inaccurate. The effectiveness of the different SPs was impacted by many factors; for instance, the effects of terracing were limited by the climate, soil properties, topography, and land uses (Chen et al., 2017a). Contouring was less effective for areas that experience large storms and that have high ridge heights, short slope lengths, etc., and the effectiveness of strip-cropping begins to diminish when the strips become too wide or when the slope lengths become too long (Renard et al., 1997). The empirical relationships for P-factors given by Renard et al. (1997) and Shin (1999) were based on experiments in the Midwest USA and in South Korea. However, quantifying the effects of different SPs applied to other locations based on these empirical tables and formulas does not generate accurate results.

In terms of the analysis of P-factors in the previous sections, further improvements to the methods of quantifying P-factors should concentrate on the following three aspects:

building a P-factor dataset at the regional level according to the work of Panagos et al. (2015a) based on the SP field experiments’ data collection, as well as the P-factor dataset at the global scale via an integration of data from the US Department of Agriculture, the FAO, and the LUCAS of Europe;

developing proxy indicators for P-factors at the large scale – for example, the high human development index and scores for pesticide regulation (Scherer and Pfister, 2015), the GDP (Borrelli et al., 2017b), and climate region can serve as proxy indicators for P-factors or can be incorporated into P-factors; and

using Earth observation data from VHR satellite imagery, such as QuickBird imagery and advanced image-processing techniques, such as object-oriented analysis, to assess P-factors. In addition, higher-resolution input data correspond to increased computational needs and decreased practicability, which means there will be limitations for large-scale studies.

VII Conclusions

This study analyzed the P-factor values in USLE-based models that were applied to estimate soil erosion in 196 published articles. An increasing trend in the number of studies, with more studies focused on soil erosion estimates at large scales, has been observed in recent years. Various methods have been developed to quantify and propose P-factor values during recent years. However, the P-factor quantification methods, such as the M2 and M4 methods, which considers SPs more fully, were used in studies with relatively smaller study areas (< 100 km²). The results of our global synthesis suggested that SPs significantly contributed to reductions in the soil erosion of arable land, and the P-factor values for paddy fields, orchards, and croplands were 0.16 ± 0.15, 0.47 ± 0.12, and 0.49 ± 0.21, respectively. Therefore, those SPs and their local effectiveness should not be ignored in soil erosion modeling either at the small or large scale. In addition, it is more feasible to collect data on the effectiveness of SPs on soil loss control based on field experiments in publications or obtain the very high-resolution satellite imagery at the small scale, which can be used to calculate the P-factor value; thus, a small-scale study should try to quantify the P-factor to improve the resulted soil erosion mapping. According to these results, we make some recommendations to better quantify P-factors.

The results of this study provide useful implications for quantifying P-factors with respect to the assessment of soil erosion risks at large scales, and they can facilitate improved applications of USLE-based modeling at the regional and global scale. In addition, the results can be beneficial for model users to improve their knowledge about P-factors, and are also good for model users to choose or develop an appropriate method to calculate the P-factor value. A globally applicable USLE-based model can enable improved estimates of global soil erosion and soil organic carbon pools, including the effects of land-use changes and conservation agriculture in past, current, and future scenarios. In a scientific sense, the use of this model can aid policy decision-making with respect to sustainable soil protection.

Supplemental material

Supplementary_data - Global analysis of support practices in USLE-based soil erosion modeling

Supplementary_data for Global analysis of support practices in USLE-based soil erosion modeling by Muqi Xiong, Ranhao Sun and Liding Chen in Progress in Physical Geography: Earth and Environment

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the National Key R&D Program of China (Grant Number 2017YFA0604704).

ORCID iD

Muqi Xiong

Supplemental material

Supplemental material for this article is available online.

References

Andriyani

Jourdain

Lidon

. (2017) Upland farming system erosion yields and their constraints to change for sustainable agricultural conservation practices: A case study of land use and land cover (LULC) change in Indonesia. Land Degradation & Development 28(2): 421–430.

Arnhold

Lindner

Lee

. (2014) Conventional and organic farming: Soil erosion and conservation potential for row crop cultivation. Geoderma 219: 89–105.

Bhandari

Aryal

Darnsawasdi

(2015) A geospatial approach to assessing soil erosion in a watershed by integrating socio-economic determinants and the RUSLE model. Natural Hazards 75(1): 321–342.

Borrelli

Panagos

Maerker

. (2017a) Assessment of the impacts of clear-cutting on soil loss by water erosion in Italian forests: First comprehensive monitoring and modelling approach. Catena 149: 770–781.

Borrelli

Paustian

Panagos

. (2016) Effect of good agricultural and environmental conditions on erosion and soil organic carbon balance: A national case study. Land Use Policy 50: 408–421.

Borrelli

Robinson

Fleischer

. (2017b) An assessment of the global impact of 21st century land use change on soil erosion. Nature Communications 8(1): 2013.

Bosco

de Rigo

Dewitte

. (2015) Modelling soil erosion at European scale: Towards harmonization and reproducibility. Natural Hazards and Earth System Sciences 15(2): 225–245.

Brooks

Ffolliott

Gregersen

. (1996) Hydrology and the Management of Watersheds. Amess: Iowa State University Press.

Cai

Ding

Shi

. (2000) Study of applying USLE and geographical information system IDRISI to predict soil erosion in small watershed. Journal of Soil and Water Conservation 14(2): 19–24 (in Chinese).

10.

Capolongo

Diodato

Mannaerts

. (2008) Analyzing temporal changes in climate erosivity using a simplified rainfall erosivity model in Basilicata (southern Italy). Journal of Hydrology 356(1-2): 119–130.

11.

Cebecauer

Hofierka

(2008) The consequences of land-cover changes on soil erosion distribution in Slovakia. Geomorphology 98(3-4): 187–198.

12.

Chen

Wei

Chen

(2017a) Effects of terracing practices on water erosion control in China: A meta-analysis. Earth-Science Reviews 173: 109–121.

13.

Chen

Oguchi

(2017b) Assessment for soil loss by using a scheme of alterative sub-models based on the RUSLE in a Karst Basin of Southwest China. Journal of Integrative Agriculture 16(2): 377–388.

14.

Clark

Odhiambo

Yoon

. (2015) Hydroacoustic and spatial analysis of sediment fluxes and accumulation rates in two Virginia reservoirs, USA. Environmental Science and Pollution Research 22(11): 8659–8671.

15.

Correa

Mello

Chou

. (2016) Soil erosion risk associated with climate change at Mantaro River basin, Peruvian Andes. Catena 147: 110–124.

16.

De Vente

Poesen

Verstraeten

. (2008) Spatially distributed modelling of soil erosion and sediment yield at regional scales in Spain. Global and Planetary Change 60(3-4): 393–415.

17.

De Vente

Poesen

Verstraeten

. (2013) Predicting soil erosion and sediment yield at regional scales: Where do we stand? Earth-Science Reviews 127: 16–29.

18.

Dissmeyer

Foster

(1980) A Guide for Predicting Sheet and Rill Erosion on Forest Land. Atlanta: USDA, Forest Service, Southeastern Area.

19.

Doetterl

Van Oost

Six

(2012) Towards constraining the magnitude of global agricultural sediment and soil organic carbon fluxes. Earth Surface Processes and Landforms 37(6): 642–655.

20.

Dunne

Leopold

(1978) Water in Environmental Planning. San Francisco: W. H. Freeman & Company.

21.

Liu

. (2011) Assessing the soil erosion control service of ecosystems change in the Loess Plateau of China. Ecological Complexity 8(4): 284–293.

22.

Zhao

Chen

. (2005) Assessment of soil erosion at large watershed scale using RUSLE and GIS: A case study in the Loess Plateau of China. Land Degradation & Development 16(1): 73–85.

23.

Fujaco

MAG

Leite

MGP

Neves

AHCJ

(2016) A GIS-based tool for estimating soil loss in agricultural river basins. REM – International Engineering Journal 69(4): 417–424.

24.

Gao

Jia

. (2016) Capacity of soil loss control in the Loess Plateau based on soil erosion control degree. Journal of Geographical Sciences 26(4): 457–472.

25.

García-Ruiz

Begueria

Lana-Renault

. (2017) Ongoing and emerging questions in water erosion studies. Land Degradation & Development 28(1): 5–21.

26.

García-Ruiz

Beguería

Nadal-Romero

. (2015) A meta-analysis of soil erosion rates across the world. Geomorphology 239: 160–173.

27.

Grimm

Jones

RJA

Montanarella

(2001) Soil erosion risk in Europe. European Soil Bureau. EUR 19022 EN, 38 pp.

28.

Grimm

Jones

RJA

Rusco

. (2003) Soil erosion risk in Italy: A revised USLE approach. European Soil Bureau Research. Report no. 11, Encarnacion, J.

29.

Guerra

Maes

Geijzendorffer

. (2016) An assessment of soil erosion prevention by vegetation in Mediterranean Europe: Current trends of ecosystem service provision. Ecological Indicators 60: 213–222.

30.

Guerra

Pinto-Correia

Metzger

(2014) Mapping soil erosion prevention using an ecosystem service modeling framework for integrated land management and policy. Ecosystems 17(5): 878–889.

31.

Hsu

Emerson

Levy

. (2014) The 2014 Environmental Performance Index. Available at: http://sedac.ciesin.columbia.edu/data/set/epi-environmental-performance-index-2014 (accessed 3 February 2019).

32.

Hurni

(1982) Soil erosion in Huai Thung Choa-Northern Thailand concerns and constraints. Mountain Research and Development 2(2): 141–156.

33.

Hurni

(1985) Erosion productivity conservation systems in Ethiopia. In: Proceedings of 4th international conference on soil conservation (ed Pla

Sentis I

), Maracay, Venezuela, 3–9 November 1985, pp. 654–674.

34.

Ito

(2007) Simulated impacts of climate and land-cover change on soil erosion and implication for the carbon cycle, 1901 to 2100. Geophysical Research Letters 34(9): 1–5.

35.

Jang

Shin

Kum

. (2015) Assessment of soil loss in South Korea based on land-cover type. Stochastic Environmental Research and Risk Assessment 29(8): 2127–2141.

36.

Jiang

Zhang

(2018) Spatially explicit assessment of ecosystem services in China’s Loess Plateau: Patterns, interactions, drivers, and implications. Global and Planetary Change 161: 41–52.

37.

Kabanza

Dondeyne

Kimaro

. (2013) Effectiveness of soil conservation measures in two contrasting landscape units of South Eastern Tanzania. Zeitschrift Fur Geomorphologie 57(3): 269–288.

38.

Karamage

Shao

Chen

. (2016a) Deforestation effects on soil erosion in the Lake Kivu Basin, DR Congo-Rwanda. Forests 7(11): 1–17.

39.

Karamage

Zhang

Kayiranga

. (2016b) USLE-based assessment of soil erosion by water in the Nyabarongo River Catchment, Rwanda. International Journal of Environmental Research and Public Health 13(7).

40.

Karamage

Zhang

Liu

. (2017) Soil erosion risk assessment in Uganda. Forests 8(7): 1–20.

41.

Karamage

Zhang

Ndayisaba

. (2016c) Extent of cropland and related soil erosion risk in Rwanda. Sustainability 8(7): 1–19.

42.

Karydas

Panagos

Gitas

(2014) A classification of water erosion models according to their geospatial characteristics. International Journal of Digital Earth 7(3): 229–250.

43.

Karydas

Sekuloska

Silleos

(2009) Quantification and site-specification of the support practice factor when mapping soil erosion risk associated with olive plantations in the Mediterranean island of Crete. Environmental Monitoring and Assessment 149(1-4): 19–28.

44.

Khosrokhani

Pradhan

(2014) Spatio-temporal assessment of soil erosion at Kuala Lumpur metropolitan city using remote sensing data and GIS. Geomatics Natural Hazards & Risk 5(3): 252–270.

45.

KICT (1992) The development of selection standard for calculation method of unit sediment yield in river. Korea Institute of Construction Technology (KICT) 89-WR-113 Research Paper (in Korean).

46.

Kumar

Devi

Deshmukh

(2014) Integrated remote sensing and geographic information system based RUSLE modelling for estimation of soil loss in Western Himalaya, India. Water Resources Management 28(10): 3307–3317.

47.

Le Roux

Morgenthal

Malherbe

. (2008) Water erosion prediction at a national scale for South Africa. Water SA 34(3): 305–314.

48.

Lee

Jeong

(2009) A strategy for quantifying turbid-water occurrence possibility based on geologic characteristics and soil erosion in hydrologic basins. Environmental Earth Sciences 59(4): 821–835.

49.

Lema

Kebede

Mesfin

. (2016) Use of the revised universal soil loss equation (RUSLE) for soil and nutrient loss estimation in long-used rainfed agricultural lands, North Ethiopia. Physical Geography 37(3-4): 276–290.

50.

Holden

. (2017a) Comparison of soil erosion models used to study the Chinese Loess Plateau. Earth-Science Reviews 170: 17–30.

51.

Wei

(2014) Analysis of the relationship between soil erosion risk and surplus floodwater during flood season. Journal of Hydrologic Engineering 19(7): 1294–1311.

52.

Zhang

Yan

. (2017b) Mapping the hotspots and coldspots of ecosystem services in conservation priority setting. Journal of Geographical Sciences 27(6): 681–696.

53.

Fang

(2016) Impacts of climate change on water erosion: A review. Earth-Science Reviews 163: 94–117.

54.

Ligonja

Shrestha

(2015) Soil erosion assessment in Kondoa eroded area in Tanzania using universal soil loss equation, geographic information systems and socioeconomic approach. Land Degradation & Development 26(4): 367–379.

55.

Litschert

Theobald

Brown

(2014) Effects of climate change and wildfire on soil loss in the Southern Rockies Ecoregion. Catena 118: 206–219.

56.

Liu

Blagodatsky

Giese

. (2016) Impact of herbicide application on soil erosion and induced carbon loss in a rubber plantation of Southwest China. Catena 145: 180–192.

57.

Liu

Kiesel

Hoermann

. (2011) Effects of DEM horizontal resolution and methods on calculating the slope length factor in gently rolling landscapes. Catena 87(3): 368–375.

58.

Liu

(2010) Sensitivity analysis of soil erosion in the Northern Loess Plateau. In: Procedia Environmental Sciences 2: 134–148.

59.

Prosser

Moran

. (2003) Predicting sheetwash and rill erosion over the Australian continent. Australian Journal of Soil Research 41(6): 1037–1062.

60.

Liang

. (2013) Influences of the Grain-for-Green project on grain security in southern China. Ecological Indicators 34: 616–622.

61.

Lufafa

Tenywa

Isabirye

. (2003) Prediction of soil erosion in a Lake Victoria basin catchment using a GIS-based Universal Soil Loss model. Agricultural Systems 76(3): 883–894.

62.

Maetens

Poesen

Vanmaercke

(2012) How effective are soil conservation techniques in reducing plot runoff and soil loss in Europe and the Mediterranean? Earth-Science Reviews 115(1-2): 21–36.

63.

Martin-Fernandez

Martinez-Nunez

(2011) An empirical approach to estimate soil erosion risk in Spain. Science of the Total Environment 409(17): 3114–3123.

64.

Medeiros

GdOR

Giarolla

Sampaio

. (2016) Estimates of annual soil loss rates in the state of São Paulo, Brazil. Revista Brasileira de Ciência do Solo 40: 1–18.

65.

Munir

Suripin

Abdullah MN

. (2000) Application of geographic information system (GIS-IDRISI) for assessing land use risks on sediment yields. Journal of the Faculty of Agriculture Kyushu University 44(3-4): 463–471.

66.

Mutua

Klik

(2004) Soil erosion management at a large catchment scale using the RUSLE-GIS: The case of Masinga catchment, Kenya. In: Brebbia

(ed) Management Information Systems 2004: Incorporating GIS and Remote Sensing. Southampton: WIT press, pp. 287–296.

67.

Nachtergaele

Petri

Biancalani

. (2010) Global Land Degradation Information System (GLADIS), beta version. An information database for land degradation assessment at global level. Land Degradation Assessment in Drylands Technical Report, Vol. 17. Food and Agriculture Organization of the United Nations (FAO).

68.

Naipal

Reick

Pongratz

. (2015) Improving the global applicability of the RUSLE model – adjustment of the topographical and rainfall erosivity factors. Geoscientific Model Development 8(9): 2893–2913.

69.

Napoli

Cecchi

Orlandini

. (2016) Simulation of field-measured soil loss in Mediterranean hilly areas (Chianti, Italy) with RUSLE. Catena 145: 246–256.

70.

Ozhan

Balci

Ozyuvaci

. (2005) Cover and management factors for the Universal Soil-Loss equation for forest ecosystems in the Marmara region, Turkey. Forest Ecology and Management 214(1-3): 118–123.

71.

Panagos

Ballabio

Borrelli

. (2016) Spatio-temporal analysis of rainfall erosivity and erosivity density in Greece. Catena 137: 161–172.

72.

Panagos

Borrelli

Meusburger

. (2015a) Modelling the effect of support practices (P-factor) on the reduction of soil erosion by water at European scale. Environmental Science & Policy 51: 23–34.

73.

Panagos

Borrelli

Poesen

. (2015b) The new assessment of soil loss by water erosion in Europe. Environmental Science & Policy 54: 438–447.

74.

Panagos

Meusburger

Van Liedekerke

. (2014) Assessing soil erosion in Europe based on data collected through a European network. Soil Science and Plant Nutrition 60(1): 15–29.

75.

Park

Jeon

. (2011) Soil erosion risk in Korean watersheds, assessed using the revised universal soil loss equation. Journal of Hydrology 399(3-4): 263–273.

76.

Pasztor

Waltner

Centeri

. (2016) Soil erosion of Hungary assessed by spatially explicit modelling. Journal of Maps 12: 407–414.

77.

Patil

(2018) Spatial Techniques for Soil Erosion Estimation: Remote Sensing and GIS Approach. Cham: Springer International Publishing AG part of Springer Nature.

78.

Peng

Liu

. (2017) Mapping spatial non-stationarity of human-natural factors associated with agricultural landscape multifunctionality in Beijing–Tianjin–Hebei region, China. Agriculture, Ecosystems & Environment 246: 221–233.

79.

Pham

Yang

Kanae

. (2001) Application of RUSLE model on global soil erosion estimate. Annual Journal of Hydraulic Engineering 45: 811–816.

80.

Plangoen

Babel

Clemente

. (2013) Simulating the impact of future land use and climate change on soil erosion and deposition in the Mae Nam Nan sub-catchment, Thailand. Sustainability 5(8): 3244–3274.

81.

Podmanicky

Balazs

Belenyesi

. (2011) Modelling soil quality changes in Europe. An impact assessment of land use change on soil quality in Europe. Ecological Indicators 11(1): 4–15.

82.

Pope

Odhiambo

(2014) Soil erosion and sediment fluxes analysis: A watershed study of the Ni Reservoir, Spotsylvania County, VA, USA. Environmental Monitoring and Assessment 186(3): 1719–1733.

83.

Rao

Ouyang

. (2014) Spatial patterns and impacts of soil conservation service in China. Geomorphology 207: 64–70.

84.

Rawat

Mishra

Bhattacharyya

(2016) Soil erosion risk assessment and spatial mapping using LANDSAT-7 ETM+, RUSLE, and GIS – a case study. Arabian Journal of Geosciences 9(4): 1–22.

85.

Renard

Foster

McCool

WDK

. (1997) Predicting soil erosion by water: A guide to conservation planning with the Revised Universal Soil Loss equation. Washington, DC: US Department of Agriculture, Agriculture Handbook No. 703.

86.

Ricker

Odhiambo

Church

(2008) Spatial analysis of soil erosion and sediment fluxes: A paired watershed study of two Rappahannock River tributaries, Stafford County, Virginia. Environmental Management 41(5): 766–778.

87.

Sanches Oliveira

Wendland

Nearing

(2013) Rainfall erosivity in Brazil: A review. Catena 100: 139–147.

88.

Scherer

Pfister

(2015) Modelling spatially explicit impacts from phosphorus emissions in agriculture. International Journal of Life Cycle Assessment 20(6): 785–795.

89.

Schnitzer

Seitz

Eicker

. (2013) Estimation of soil loss by water erosion in the Chinese Loess Plateau using Universal Soil Loss Equation and GRACE. Geophysical Journal International 193(3): 1283–1290.

90.

Segura

Sun

McNulty

. (2014) Potential impacts of climate change on soil erosion vulnerability across the conterminous United States. Journal of Soil and Water Conservation 69(2): 171–181.

91.

Shin

(1999) The soil loss analysis using GSIS in watershed. Dissertation, Kangwon National University, South Korea.

92.

Tamene

Vlek

PLG

(2014) A landscape planning and management tool for land and water resources management: An example application in Northern Ethiopia. Water Resources Management 28(2): 407–424.

93.

Tang

Bennett

. (2015) Assessment of soil erosion using RUSLE and GIS: A case study of the Yangou watershed in the Loess Plateau, China. Environmental Earth Sciences 73(4): 1715–1724.

94.

Teng

Rossel

RAV

Shi

. (2016) Assimilating satellite imagery and visible-near infrared spectroscopy to model and map soil loss by water erosion in Australia. Environmental Modelling & Software 77: 156–167.

95.

Terranova

Antronico

Coscarelli

. (2009) Soil erosion risk scenarios in the Mediterranean environment using RUSLE and GIS: An application model for Calabria (southern Italy). Geomorphology 112(3-4): 228–245.

96.

Todisco

Brocca

Termite

. (2015) Use of satellite and modeled soil moisture data for predicting event soil loss at plot scale. Hydrology and Earth System Sciences 19(1): 3845–3856.

97.

UNDP (2014) Sustaining human progress. Reducing vulnerabilities and building resilience. Human Development Report (UNDP). United Nations Development Programme, New York.

98.

Van der Knijff

Jones

RJA

Montanarella

(1999) Soil erosion risk assessment in Italy. European Soil Bureau. EUR 19022 EN, 55.

99.

Van der Knijff

Jones

RJA

Montanarella

(2000) Soil erosion risk assessment in Europe. European Soil Bureau. EUR 19044 EN, 36.

100.

Van Oost

Quine

Govers

. (2007) The impact of agricultural soil erosion on the global carbon cycle. Science 318: 626–629.

101.

Van Rompaey

Bazzoffi

Jones

. (2003) Validation of soil erosion risk assessments in Italy. Report no. 12, EUR 20676 EN. Luxembourg: Office for Official Publications of the European Communities.

102.

Vente

Poesen

(2005) Predicting soil erosion and sediment yield at the basin scale: Scale issues and semi-quantitative models. Earth-Science Reviews 71(1-2): 95–125.

103.

Vezina

Bonn

Van

(2006) Agricultural land-use patterns and soil erosion vulnerability of watershed units in Vietnam’s northern highlands. Landscape Ecology 21(8): 1311–1325.

104.

Villarreal

Webb

Norman

. (2016) Modeling landscape-scale erosion potential related to vehicle disturbances along the USA–Mexico border. Land Degradation & Development 27(4): 1106–1121.

105.

Wang

Zeng

. (2014) Spatial heterogeneous response of land use and landscape functions to ecological restoration: The case of the Chinese loess hilly region. Environmental Earth Sciences 72(7): 2683–2696.

106.

Wang

Zhang

Yang

. (2016) Integrated use of GCM, RS, and GIS for the assessment of hillslope and gully erosion in the Mushi River sub-catchment, Northeast China. Sustainability 8(4): 1–20.

107.

Wei

Chen

Wang

. (2016) Global synthesis of the classifications, distributions, benefits and issues of terracing. Earth-Science Reviews 159: 388–403.

108.

Wener

(1981) Soil Conservation in Kenya. Nairobi: Ministry of Agriculture, Soil Conservation Extension Unit.

109.

Wischmeier

(1975) Estimating the soil loss equation’s cover and management factor for undisturbed areas. In: Present and prospective technology for predicting sediment yields and sources. Agricultural Research Service Report ARS-S-40. US Department of Agriculture, Oxford, Mississippi, pp. 118–124.

110.

Wischmeier

Smith

(1978) Predicting Rainfall Erosion Losses. A guide to conservation planning. Washington, DC: US Department of Agriculture, Agriculture Handbook No. 537.

111.

Zhou

Xiao

. (2011) Estimation of soil erosion in the Three Gorges Reservoir Area of China using RUSLE, remote sensing and GIS. Journal of Food Agriculture & Environment 9(2): 728–734.

112.

Liu

X-y

(2016) Tracking soil erosion changes in an easily-eroded watershed of the Chinese Loess Plateau. Polish Journal of Environmental Studies 25(1): 351–363.

113.

Xiao

Yang

Chen

. (2015) An assessment of erosivity distribution and its influence on the effectiveness of land use conversion for reducing soil erosion in Jiangxi, China. Catena 125: 50–60.

114.

Xiong

Sun

Chen

(2018) Effects of soil conservation techniques on water erosion control: A global analysis. Science of the Total Environment 645: 753–760.

115.

Meng

(2013) Risk assessment of soil erosion in different rainfall scenarios by RUSLE model coupled with Information Diffusion Model: A case study of Bohai Rim, China. Catena 100: 74–82.

116.

Tang

Wang

. (2017) Effects of landscape patterns on soil erosion processes in a mountain-basin system in the North China. Natural Hazards 87(3): 1567–1585.

117.

Luo

Peng

(2011) Land use change and soil erosion in the Maotiao River watershed of Guizhou Province. Journal of Geographical Sciences 21(6): 1138–1152.

118.

Shao

Kong

. (2008) Adapting the RUSLE and GIS to model soil erosion risk in a mountains karst watershed, Guizhou Province, China. Environmental Monitoring and Assessment 141(1-3): 275–286.

119.

Yang

Kanae

Oki

. (2003) Global potential soil erosion with reference to land use and climate changes. Hydrological Processes 17(14): 2913–2928.

120.

Yang

Zhao

Chow

. (2009) Using GIS and a digital elevation model to assess the effectiveness of variable grade flow diversion terraces in reducing soil erosion in northwestern New Brunswick, Canada. Hydrological Processes 23(23): 3271–3280.

121.

Yang

(1999) Study on soil loss equation of cultivated slope land in northeast mountain region of Yunnan province. Bulletin of Soil and Water Conservation 19: 1–9.

122.

Yang

(2002) Study on soil loss equation in Jinsha river basin of Yunnan Province. Journal of Mountain Science 20: 1–9 (in Chinese with English abstract).

123.

Yao

Shi

Shao

. (2016) Changes and influencing factors of the sediment load in the Xiliugou basin of the upper Yellow River, China. Catena 142: 1–10.

124.

Zeng

Wang

Bai

. (2017) Soil erosion evolution and spatial correlation analysis in a typical karst geomorphology using RUSLE with GIS. Solid Earth 8(4): 721–736.

125.

Zhuang

Zhang

. (2015) Research trends and hotspots in soil erosion from 1932 to 2013: A literature review. Scientometrics 105(2): 743–758.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB