Sage Journals: Discover world-class research

Abstract

Soil erosion is among the most pressing environmental problems worldwide, driven by agricultural intensification, climate change, and other uncontrolled human activities. Understanding how soil erosion threatens food productivity is important for preventing, reducing, and restoring land degradation, thereby contributing to attaining key sustainable development goals. This study employed the Revised Universal Soil Loss Equation (RUSLE) model and Support Vector Machine (SVM) algorithm, coupled with Geographical Information System (GIS), to estimate soil loss in the Federal University of Agriculture Abeokuta, Nigeria. Land use cover maps for 2002, 2012, and 2022 were generated through SVM. RUSLE parameters model was derived from remote sensing data, and erosion vulnerability zones were identified using GIS analysis. The results showed significant land use changes, including a substantial increase in built-up areas (42.7 %), and barelands (1.6 %), alongside a decrease in farmlands (30.9 %), and vegetation (23.7 %), respectively. Annual soil loss estimates of the studied area ranged in the order of 11 > 13 > 17 (t/ha/yr) with vulnerability soil loss classed as low (731 ha), moderate (421 ha), and high (491 ha). The findings from this study recommend adopting SVM and RUSLE for erosion assessment to inform sustainable land use and soil management practices. Future studies could integrate real-time erosion monitoring for improved conservation practices.

Keywords

Soil degradation RUSLE model support vector machine remote sensing geospatial FUNAAB—landscape

Introduction

Land use cover change represents a major driver of soil degradation and terrestrial ecosystem imbalance (FAO, 2020). The relationship between land use degradation, ecosystem services, and food insecurity has gained increasing attention, often exacerbated by anthropogenic activities (Lal, 2001). Proper understanding of soil resources and land use cover management are important to Life on Land (Amundson et al., 2015; FAO and ITPS, 2015). Consequently, degradation of the physical, chemical, and biological condition of soil caused by erosion is a worldwide problem e.g., depletion of nutrient rich surface soil, decreasing water availability for plants, and reducing the effective rooting depths (Borrelli et al., 2021; Ebabu et al., 2019; Oicha et al., 2010). Mitigating soil loss, a priority that aligns with global top-down initiatives and societal values—for example, the European Union’s 2021 policies—has become increasingly essential due to many factors like soil heterogeneity (Lal et al., 2021), relentless land cover changes (Bünemann et al., 2018), and challenges that seem progressively insurmountable in emerging countries like Nigeria (Tobore and Samuel, 2022).

Studies on soil erosion have a long scientific history and remain a critical area of research with increasing focus on detailing erosion processes and modeling landscape terrain (Kapur and Akça, 2020). Unfortunately, the rate of soil erosion is changing the landscape terrain. Moreover, soil erosion process is triggered or modified by biophysical factors like steep slopes, extreme climate events, abusive land use practices and interactions between them (Boufeldja et al., 2020). Though numerous erosion models have been developed through different methods and modeling approaches in the past, the concepts governing each erosion models differ widely due to changes in land use and land cover across different regions (Flanagan and Laflen, 1997). For instance, appropriate land use cover managements, e.g., steady increase in vegetative cover space, can gradually decrease soil erosion. Therefore, assessing land use cover changes alongside with soil erosion vulnerability can provide information about serious environmental issues that threaten agricultural productivity, erosion trends and allow easy efficient scenario analysis at different levels (Flanagan and Laflen, 1997; Wischmeier and Smith, 1978). The most widely accepted and frequently empirical model for assessing soil erosion vulnerability is the Revised Universal Soil Loss Equation (RUSLE) (Tibebu et al., 2018). More importantly, RUSLE is particularly notable for its ability to account for most of the factors influencing both long and short term erosion vulnerability (Adefisan et al., 2015). To date, RUSLE is the most widely utilized for soil erosion vulnerability studies, with extensive application in soil erosion estimations (Nearing et al., 2005; Panagos et al., 2014). Although other erosion models exist with varying degrees of complexity, RUSLE has been successfully applied to estimate soil loss with cell-by-cell pixel methods (Lee, 2004; McCool et al., 1995; Shinde et al., 2010).

Combining RUSLE with geo-spatial techniques like the Geographic Information System (GIS) provides a robust and non-trivial approach of assessing temporal and spatial distribution of soil erosion vulnerabilities across various geo-spatial scales (Adefisan et al., 2015; Fistikoglu and Harmancioglu, 2002; Wischmeier and Smith, 1978). Consequently, Sun et al. (2014) and Makinde and Oyebanji (2020) analyzed the impacts of land use cover changes on soil erosion vulnerability using the GIS and RUSLE model. Their findings showed that undisturbed or intact vegetation reduces soil erosion vulnerability rates over time. Even though this study provided critical insights into the influence of anthropogenic climate shocks on natural landscape and geomorphology, it did not incorporate land use metrics like Number of Patches (NP), Largest Path (LP), and Effective Mean Shape Index metric (MESH). Moreover, employing the RUSLE model especially in a complex landscape to assess the impact of land use cover on soil erosion vulnerability may not be able to capture the effects of fragmentation e.g., sub-Saharan Africa—Nigeria. Therefore, a combination of landscape metrics along with the RUSLE model is needed. Interestingly, these metrics remain invaluable for assessing landscape complexities and have been highlighted as effective proxies for studying and monitoring the dynamics of shape, patch connectivity, and ecological alterations while leading to soil erosion (Masroor et al., 2022; Nguemezi et al., 2020). Further, NP, LP, and MESH have demonstrated significant utility in quantifying spatial changes in landscape composition and structure caused by increasing climate variability (Li et al., 2019).

The emphasis on fostering a green environment to enhance food production underscores the importance of accurately assessing shifts in land use and land cover. The Support Vector Machine (SVM) algorithm has emerged as a potent tool for this purpose (Gashaw et al., 2017). RUSLE model utilized alongside SVM has been proven effective for modeling soil erosion vulnerability with high precision (Gashaw et al., 2017; McBratney et al., 2014). Nevertheless, SVM, a non-parametric supervised statistical learning approach, is particularly well-suited for analyzing land use cover dynamics (Tully et al., 2015). Therefore, the RUSLE and SVM algorithm used with land use metrics—NP, LP, and MESH—are timely and can accurately assist in identifying drivers responsible for gradual to sudden shifts in land use cover while leading to soil quality depletion (Kidd et al., 2018). Timeliness is key as some of the intrinsic features of soil could make it more susceptible to erosion particularly in countries like China and Australia, along with Nigeria (Lal et al., 2021). This further highlights the immediate concern and need for site-specific studies to curb soil vulnerability losses (Kim et al., 2005) relating to varied agro-climatic conditions especially in complex landscapes such as Federal University of Agriculture, Abeokuta (FUNAAB), Ogun State, Nigeria (Tobore et al., 2024).

Interestingly, there has been no complete assessment of land use cover changes on soil erosion vulnerability through geospatial techniques in the FUNAAB landscape. By employing the RUSLE and SVM algorithm, this research aims to provide critical information for better decision-making, benefiting society, stakeholders, and setting a benchmark for future studies in similar environments. The intellectual merit of this study lies in its potential capability to synthesize remote sensing data and algorithmic analysis into actionable insights. Ultimately, this research proposes strategies for improving vegetative cover and contributes to achieving the sustainable development goal on healthy environment and Life on Land.

The objectives of this study are to:

Analyze land use cover changes through Landsat images from 2002, 2012, and 2022.

Assess the landscape changes through NP, LP, and MESH metrics.

Develop a methodology that utilizes SVM and RUSLE to estimate soil erosion vulnerability.

Materials and methods

Study area

The study area with a total area of 1643 hectares (ha) was geographically selected between Latitudes 7° 19’ to 7° 26’ and Longitudes of 3° 42’ to 3° 46’ at Federal University of Agriculture, Abeokuta, (FUNAAB), Ogun State, Nigeria (Figure 1). The geology of the area overlies basement complex (Smyth and Montgomery, 1962) and is governed by a seasonal humid tropical climate which is divided into dry and rainy seasons (Ufoegbune et al., 2010). The topography of the area ranges between 346 to 41 meters above sea level and is consequently drained into the Ogun River basin (Nkwunonwo et al., 2024; Ufoegbune et al., 2010). Dominant soil types classified for the area (Sotona et al., 2013; Tobore et al., 2021) are Ferric-Fluvisols (29.9%), Ferric-Cambisols (53.1%) and Eutric-Fluvisols (17%). Annual Standardized Precipitation Index (SPI) showed moderate wet years ranged from 1.0 to 1.49 for the area (Tobore et al., 2021). More importantly, agricultural expansion and gradual to sudden increase in human population positions the slopes as vulnerable to erosion (Ufoegbune et al., 2010).

Figure 1.

Map of Nigeria indicating the study state boundary (a) Nigeria map enclosing the study state boundary (b) Ogun state showing the FUNAAB boundary (c) study area DEM (d) extracted within FUNAAB boundary.

Soil mapping

We employed a combination of semi-detailed random sampling and soil survey manual techniques (FAO, 2007) to map and create a homogenous sampling framework for the study area. To ensure the sampled data accurately represented the various land use cover classes, 25 sampling points were randomly selected within each land use cover category—built-up areas, vegetation, barelands, and farmlands—resulting in a total of 100 samples. Subsequently, 50 soil samples were grouped based on the on-field textural class examination to evaluate the soil properties of the area. The sampling point locations were georeferenced using a handheld Global Positioning System (GPS) device with accuracy less than 5 meters. The collected on-field textural samples were then stored in well-labeled polyethylene bags, maintained at controlled room temperature, and air-dried to preserve their properties for laboratory analysis.

Laboratory analysis

Air-dried soils were sieved with 2 mm diameter sieve to test for soil particle size (Bouyoucos, 1962), soil pH were tested in water suspensions (McLean, 1982), while total nitrogen and organic carbon were assessed by macro-Kjeldahl and chromic-acid oxidation.

Spatial remote sensing analysis

In this study, geoprocessing and remote sensing methods were used to assess land use cover changes in the area (Kafy et al., 2021). Landsat imageries from 2002, 2012, and 2022 were obtained from United States Geological Survey (USGS) repository, with a threshold of <5% cloud cover (Chavez, 1996), to analyze land use cover of the area. The images were selected based on the vegetation phenology of the area (Orimoloye et al., 2018), focusing on the four primary land use classes: built-up, vegetation, barelands and farmlands (Anderson et al., 1976). Pixel-cells and class separability were extracted from the satellite images to classify land use cover of the studied area through SVM algorithm. SVM was utilized for its simplicity, clarity, and higher accuracy in land use cover classification (Kafy et al., 2021). Additionally, statistics under this research such as User and Producer accuracies were employed through land use map categorization based on matrix analysis (Foody, 2002). On-screen zoom window digitization in a GIS environment was used to merge spectrally similar pixels to specific Areas of Interest (AOI). The AOI derived from the images was utilized to produce land use cover maps for 2002, 2012, and 2022.

Ecological assessment plays a critical role in managing landscape heterogeneity and mitigating threats posed by un-checked human actions (Aguilera et al., 2011). Spatial land use metrics (Table 1) were applied to evaluate landscape fragmentation in the study area using land use cover maps for 2002, 2012 and 2022 through FRAGSTATS (McGarigal, 2002). FRAGSTATS is a key scientific software used to measure the arrangement of landscape patterns and ecological sustainability (McGarigal, 2002). Therefore, the FRAGSTATS was applied as an extension in Q-GIS v3.18.3 to assess and measure ecological quality and structure characteristics of the area using NP, LP and MESH metrics. The analyzed land use cover maps between 2002 and 2022 were used as predictors to delineate NP, LPI and MESH pattern of the area (Aguilera et al., 2011).

Table 1.

Land use metrics used in the study.

Metric name	Formula	Explanation	References
Number patches	$N P \geq 1$	Spots for particular class	(Gevaert and Belgiu, 2022)
Largest path	$L P = \frac{{M A X}^{n}_{i = 1} (a i j)}{A} \times 100$	Biggest spot index	(Li et al., 2019)
Effective MESH	$M E S H = \frac{\sum_{j = 1}^{n} a i j (\frac{1}{10, 000})}{A}$	Fragmented areas	(McGarigal and Marks, 1995)

RUSLE factors estimation

RUSLE is a recognized conservation model (Wischmeier and Smith, 1978) that is widely utilized to evaluate the magnitude and extent of soil losses at both regional and global scales (Zerihun et al., 2018). In the study, Shuttle Radar Topographic Mapper (SRTM) with a resolution of 30 meters acquired from the US Geological Survey (USGS) repository was employed to analyze the drainage pattern of the area. Consequently, the RUSLE, as described by Renard et al. (1997) was applied to estimate annual soil loss of the area. The equation for estimating annual soil loss (Eq. (1)) was applied.

A = R \times K \times L S \times C \times P

(1)

where A: mean annual soil loss (t ha ^-1 yr ^-1) rate, R : rainfall runoff erosivity (MJ mm t ha ^-1 yr ^-1) factor, K represent soil erodibility (t ha ^-1 yr ^-1), LS : slope length and steepness, C refers to land use factor (dimensionless), P is the erosion control measure practices.

Rainfall erosivity (R)

Erosivity is the most important parameter in the RUSLE-base estimation which refers to the ability or power of rainfall intensity to erode the soil (Amsalu and Mengaw, 2014) or after a forceful rainfall event leading to soil surface detachment and depositions (Chuenchum et al., 2020). In this study, rainfall data obtained from Department of Water Resources Management – FUNAAB meteorological station was subjected to spatial Inverse Distance Weighted (IDW) technique to assess mean annual rainfall pattern (Bekele, 2021). The IDW method is a simple, easy to understand, intuitive and efficient interpolation that is globally accepted (Tobore et al., 2024). Therefore, the monthly rainfall data of 20 years (2002 –2022) was used to compute the erosivity (R) for the studied area through Eq. (2) developed by Hagos (2020). The computation was carried out in ArcGIS 10.5 software using spatial analysis toolbox.

R = - 8.12 + 0.562 \times P

(2)

P refers to the mean annual precipitation (mm)

Soil erodibility (K)

K-factor measures susceptibility of eroded soil particles without conservation practices and measures the function of particle size distribution, organic matter content, structure and permeability relating to the vulnerability of detachment and transport by either rainfall or runoff processes (Renard et al., 1997). In this study, suggested field soil color by Yesuph and Dagnew (2019) for soil qualities (Bewket and Teferi, 2009) was applied to determine the K (t/ha/MJ/mm) with aid of Munsell color chart (Table 2).

Table 2.

Soil erodibility.

Soil color	Black	Red	Brown	Yellow
K-factor	0.2	0.3	0.2	0.4

Slope length and steepness (LS)

In this study, the effects of topography on soil erosion were assessed using LS factor derived from Digital Elevation Model (DEM) with a 30 × 30 meter resolution (Renard et al., 1997; Yesuph and Dagnew, 2019). The LS factor, as described by Wischmeier and Smith (1978) was calculated using the Raster calculator tool in ArcGIS 10 based on equations 3 - 5

L i j = \frac{[{(A i j - i n + D^{2})}^{M + 1} - (A i j - i n^{M + 1})]}{(C^{M + 2}) \times (X i j m) \times (22.13) m}

(3)

where Lij represents the slope length, A: flow accumulation, C: grid cell size, m highlights slope exponent (Mccool et al., 1989) length.

m = \frac{B}{(B + 1)}

(4)

B = \frac{(\frac{\sin θ}{0.0896})}{3 [{\sin θ}^{0.8} + 0.56]}

(5)

sinθ is the angle of slope degree.

Cover management (C)

C – factor for cover management was applied to determine the agricultural land use cover of the area using classified images of the years 2002, 2012, and 2022. Each land use cover feature (Table 3) was assigned to its corresponding C value, as proposed by Hurni (1989).

Table 3.

C-values for land use cover.

Land classes	C
Forestland/Vegetation	0.02
Farmland	0.17
Bareland	0.60
Buildings/Built-up	0.09

Conservation practice (P)

P refers to the ratio of soil loss under certain conservation practices such as contour farming, strip-cropping, or terracing compared to the soil loss from conventional up-slope and down-slope cultivations. In this study, the conservation practices for both agricultural and non-agricultural landscapes, as adopted by Girma and Gebre (2020) from Wischmeier and Smith (1978), were applied in Table 4. A stepwise workflow outlining the methodology is presented in Figure 2.

Table 4.

Erosion-control practice.

Land use	Slope class	P
Land use	(%)	P
Agricultural	< 5	0.10
	5-10	0.12
	10-20	0.14
	20 -30	0.19
	30- 50	0.25
	50 -100	0.33
Other land use	0 – 100	1

Figure 2.

Stepwise workflow of the study.

Results and discussion

Land use and accuracy assessment

Assessing land use cover changes in the area using landscape metrics is crucial for accurate ecological and biodiversity protections (Tahmoures et al., 2022). In the study, results of the land use cover showed that built-up areas (16 to 42.7 %) and barelands (0.6 to 1.6 %) increased leading to a decrease in vegetation (39.1 to 23.7 %), and farmlands (42.8 to 30.9 %) in the area (Table 5; Figure 3). To evaluate the classification accuracy, statistical metrics such as User’s Accuracy (UA) and Producer’s Accuracy (PA) were employed (Rousta et al., 2018). UA reflects the probability of accurately identifying land use features on the terrestial ground, while PA assesses the overall fitness of the land use classification. The results (Table 6) met the standards set by previous land use cover classification schemes (Anderson et al., 1976; Rousta et al., 2018). The analysis pointed to a continuous shifts from vegetation cover to built-up environments, indicating that land use cover changes could be driven by variations in climate and geomorphology, combined with human-induced modifications. These existential threats accelerate degradation of land resources through processes such as soil erosion and nutrient depletion (Abdelkarim et al., 2023). Senjobi et al. (2019) mentioned that improper land use cover changes are critical issues facing the area, thereby contributing to landscape fragmentation (Tobore and Anoke, 2025). Therefore, the findings from the present study can provide a proper and fuller understanding for addressing the landscape fragmentation of the studied area.

Table 5.

Land use cover of FUNAAB.

Land useclasses	Number of pixels	Class separability	2002		2012		2022
Land useclasses	Number of pixels	Class separability	Ha	%	Ha	%	Ha	%
Built-up	87,098	198235	266.8	16.0	564.6	33.9	711.5	42.7
Vegetation	79,345	195235	652.8	39.1	530.7	31.8	394.5	23.7
Barelands	88,359	176865	9.4	0.6	23.0	1.4	27.0	1.6
Farmlands	78,059	178135	714.4	42.8	527.6	31.6	515.1	30.9

Figure 3.

Table 6.

Land use cover accuracy of FUNAAB.

	User Accuracy (%)						Producer Accuracy (%)
Land use	BU	VG	BL	FL	AUA	BU	VG	BL	WL	APA
2002	87.1	90.2	89.5	87.5	88.6	84.1	78.9	87.1	78.1	82.1
2012	92.1	88.3	87.3	88.6	89.1	88.1	77.9	85.6	81.7	83.3
2022	79.9	81.9	80.2	79.1	80.3	78.2	70.9	85.6	83.9	79.7.

Land use classes: BU: Built-up; VG: Vegetation; BL: Bareland, FL: Farmland, AUA; Average user accuracy, APA; Average producer accuracy.

Analysis of the land use cover maps from 2002, 2012 and 2022 showed changes in the NP, LP, and MESH metrics (Table 7), indicating habitat deterioration of the studied area (Pazúr et al., 2021). The results of LP indicated that built-up areas were the most significant land use type, followed by farmlands, and barelands. Though, a notable increase was found in built-up areas reflecting the growing proportion of the landscape occupied by the largest patches index. Additionally, significant changes in the MESH, i.e., Mean Shape Index metric, were observed for vegetation, barelands, and farmlands. Overall, the NP, LP, and MESH metrics played a crucial role in revealing the landscape pattern and characteristics of the study area. Indeed, these metrics provided useful in assessing the interaction between spatial and eco-biodiversity changes. Specifically, the obtained results demonstrate that landscape metrics can effectively assess spatial landscape patterns and heterogeneity issues (Mondal and Jeganathan, 2022).

Table 7.

LP, NP and MESH of FUNAAB.

	2002			2012			2022
LUC	NP	LP	MESH	NP	LP	MESH	NP	LP	MESH
Vegetation	132	14.1	3.1	85	8.0	2.2	50	6.9	2.0
Built-up	121	31.5	3.6	134	6.1	1.8	160	10.5	1.5
Barelands	160	12.27	5.6	61	8.5	2.2	89	7.6	1.9
Farmland	90	10.23	2.9	120	4.9	1.9	102	5.8	2.1

Soil condition of FUNAAB

The results of the selected soil nutrients distribution showed distinct pattern concentrations with values categorized as high, middle and low values for the study area. The soil texture is predominantly sandy-clay-loam according to USDA (United States Department of Agriculture) criteria. The soil pH ranges from acidic (<6.5) to neutral (7.3). Total nitrogen content was notably low in over 90% of the samples, with only a small proportion showing higher levels. Organic carbon (OC) content varied across the study area, being low in the northern part and moderate in the southern part. Variability obtained for soil pH possibly depicts presence of leaching, like basic cations, as well as varying nature of the topographic positions of the area. According to Brady and Weil (2014) and Oldeman (1998) and Weil et al. (2016), soil pH is a master variable that influences other chemical reactions in the soil. For example, high levels of toxic elements like cadmium and zinc degrade soil nutrients, especially in areas close to industries or factories where waste-water or effluents are indiscriminately discharged to the environment, potentially contributing to the acidic nature of the soil. Additionally, the loss of vegetative cover and the effects of slope positions have been identified as significant predictors contributing to runoff, nutrient leaching, and low OC content in the soils (Safadoust et al., 2016; Widyatmanti and Umarhadi, 2022). Li et al. (2009) further emphasized that land use cover degradation has led to detrimental effects on soil health through the mobility of traffic-related toxic metals and irrigation water containing high heavy metals. Excessive water retention in agricultural soils, when drainage is ineffective can also result in a decline in soil nutrients and stunted crop growth (Safadoust et al., 2016). Further, Dubrovsky et al. (2010) highlighted a 58 % loss in soil nutrients due to the substantial loss of vegetative cover exacerbated by direct sunlight heating the Earth’s surface which can cause a decrease in soil properties distribution. Ultimately, the Earth is experiencing a “fever” (Johnson et al., 2011) due to the decline in global soil carbon, and the low soil OC observed in the present study can be traced to various emissions of greenhouse gases, leading to ecological and biodiversity instability. The over-exploitation of soil properties may eventually lead to food scarcity and decreased crop quality and growth, particularly for vulnerable populations and future generations.

Soil severity loss

Erosion is a global threat to biodiversity conservation and agricultural yields (Arabameri et al., 2020). The RUSLE model was applied to assess the rate of soil degradation and highlights the soil condition of the studied area. In this study, the five major factors of the RUSLE model were super-imposed to compute annual soil loss for the area. A result for each RUSLE parameters was discussed as follows.

Rainfall (R) erosivity

In this study, the R factor IDW interpolation technique showed that the value ranges from 8 - 124, 10 - 118 and 18 -179 MJmm/ha/h/year, respectively (Figure 4). These results confirmed the variations in rainfall erosivity within the study area, highlighting R-factor as an important criterion for estimating soil loss over different temporal periods. Further, the variability in erosivity factors across the study area may be attributed more to changes in climatic conditions and soil management practices (Salako et al., 2013) than to the inherent erodibility of the soils. The multiple erosivity R values utilized for the study satisfactorily justify the varying distribution of rainfall pattern within the area which may be influenced by land use cover changes especially under changing anthropogenic climatic shocks.

Figure 4.

R-factor FUNAAB for 2002, 2012 and 2022.

Soil (K) erodibility

K values for the study were calculated using soil color data (Yesuph and Dagnew, 2019). The dominant soil types were Ferric-Fluvisols, Eutric-Fluvisols and Ferric-Cambisols. The resulting values of the erodibility factor revealed that the studied soils varied between 0 and 0.25 t/ha/MJ/mm, with Ferric-Cambisols being the most prevalent, having an erodibility value of 0.25 t/ha/MJ/mm and covering approximately 953.8 ha of the study area (Figure 5). In contrast, erodibility values for Ferric-Fluvisols and Eutric-Fluvisols were lower, at 0.15 t/ha/MJ/mm, covering approximately 534.4 ha and 179.5 ha, respectively (Table 8).

Figure 5.

Soil type and K of the area.

Table 8.

Coverage K values of the studied area.

Soil type	Soil color	Area-coverage		K value
Soil type	Soil color	(Ha)	(%)	K value
Ferric-Fluvisols	Black	534.4	32.0	0.15
Eutric-Fluvisols	Black	179.5	10.8	0.15
Ferric-Cambisols	Red	953.8	57.2	0.25

Ha: Hectare, %: Percentage.

The areas with low K-factor can be highlighted by the abundance of clay in the studied soils which provides resistance to erosion vulnerability. Even though soil type plays an essential role in water retention and nutrient transport via runoff, the interplay between slope steepness and slope length affects nutrient runoff (Dabral et al., 2008). Globally, 75 billion metric tons of soil every year are eroded due to the anthropogenic disturbances traced to increase population growth (Dabral et al., 2008; United Nations, 2015). Indeed, soil erosion is globally responsible for 84 % of the land degradation (Opeyemi et al., 2019), and thus negatively impacting the wellbeing of more than 3.2 billion people (Borrelli et al., 2021). The K values obtained for our study could highlight the ease of soil detachment by splash during rainfall or by surface flow due to external force of energy (Adegboyega, 2019). So far, our findings agree with the studies conducted by Adediji et al. (2010) in Katsina (Nigeria), Lorentz and Schulze (1994) in Pretoria (South Africa), and Reshma and Uday (2012) in Jharkhand (India).

Slope length and steepness (LS)

The studied slope ranged from 0 to 30 %, while the LS factor fluctuated between 1 and 55 % (Figure 6). The results reflect the influence of the slope length and slope steepness on the erosion susceptibility of the study area. The combination of these two sub-factor maps highlighted that the steeper areas are significantly dominated by higher LS factors covering southwest, central, and northeastern parts, indicating more susceptibility to soil erosion compared to the lower portions of the area. The higher LS values confirmed the increased vulnerability of the area to erosion, which is consistent with findings from other tropical research studies (Adediji et al., 2010; Fagbohun et al., 2016). So far, extensive cultivation and little to no conservation measures coincided with the areas where the LS factor is high in the studied area.

Figure 6.

Slope class and LS factor of FUNAAB.

Cover management (C) factor

In this study, built-up areas and barelands have increased over the period under study. The corresponding C-factor values ranging from 0.01 to 0.6 (Figure 7) were derived from Hurni (1989). The lower C factor values were observed in the western and south eastern part especially, which have land areas characterized by undisturbed vegetative cover. Contrarily, the higher C factor values can be seen in the southern part of the area covering a large portion of the land characterized by barelands and built-up areas which could be traced to no vegetative cover, thereby leading to direct rain-drop impacts on the bare soil. Although recent studies have confirmed that disturbed or unhealthy vegetative cover remain a significant driver that can aid soil erosion vulnerability, contributing to eco-biodiversity instability at different temporal and spatial scales (Negese et al., 2021; Elnashar et al., 2021).

Figure 7.

C factor map of FUNAAB for 2002, 2012 and 2022.

Erosion control practice

P factor ranges from 0.1 to 1 based on non-agricultural and agricultural land use. The results of the land use cover revealed land containing bareland (27.0 ha), built-up areas (711.5 ha) and vegetation (394.5 ha) were classified as non-agricultural and assigned P value of 1, whereas farmlands (30.9 ha) were identified as agricultural land use. The built-up areas, i.e., non-agricultural areas, were significantly occupied with a slope less than 20% and thus having P-value of 0.1 (Figure 8).

Figure 8.

P factor map of FUNAAB for 2002, 2012 and 2022.

Soil severity loss along different land cover

Annual soil loss values of the studied area varied from 0 to 132, 157 and 260 t/ha/year, respectively (Figure 9). These vulnerabilities could be associated with inappropriate land use practices. The high LS-factor and abrupt changes in slope positions of the area especially in the middle and some lower parts were identified as key contributors to the observed soil loss. Consequently, the average annual soil loss for the area was estimated at approximately 11, 13, and 17 t/ha/year respectively (Table 9) according to İrvem et al. (2007). Degife et al. (2021) noted that soil erosion beyond 10 t/ha/year may be considered irreversible over a 100-year period. Though, average soil loss rate in this study align with findings from other research in the studied region (Adediji et al., 2010; Dike et al., 2018; Mesfin et al., 2019). However, estimated annual soil loss within the identified land use cover classes shown in Table 10, highlight that farmlands and barelands experienced an increase trend soil loss of 41 t/ha/year to 59 and 63 t/ha/year, respectively.

Figure 9.

Soil loss of FUNAAB for 2002, 2012 and 2022.

Table 9.

Soil severity loss of FUNAAB for 2002, 2012 and 2022.

	t/ha/yr
	Min	Max	Mean
2002	0	132	11
2012	0	157	13
2022	0	260	17

Min: Minimum, Max: Maximum.

Table 10.

Mean annual soil loss rate.

t/ha/yr
Land use cover	2002	2012	2022
Built-up	28	22	9
Vegetation	7	5	3
Barelands	33	41	54
Farmlands	41	59	63

Vegetative cover loss in the area could be driver for erosion effects leading to soil loss of the studied area. Approximately 44% of the area experienced low soil loss values (<10 t/ha/year), categorizing it as a minor risk area (FAO and UNEP, 1984; Wolka et al., 2015). Nevertheless, the remaining areas, as shown in Table 11, included 18% classified as moderate and 31% as high severity zones. Disturbingly, the soil loss rate in the area warrants immediate concern and intervention to mitigate the depletion of soil caused by changes in the land use cover and inappropriate land management practices. Moreover, the significant conversion of vegetated lands into built-up areas, including agricultural farmlands, has reduced the protective functions of the vegetation, further exacerbating soil erosion. More importantly, anthropogenic activities, particularly deforestation, adversely contributed to the major effects of soil erosion severity while leading to landscapes fragmentation of the studied area (Adediji et al., 2010).

Table 11.

Soil severity classes (t/ha/year) 2022.

Mean	Severity class	Area (Ha)	Area (%)
< 10	low	731	44
10 -20	moderate	421	25
> 20	high	491	31
Total		1643	100

Conclusion and recommendation

The study successfully demonstrated the integration of the Support Vector Machine (SVM) and the Revised Universal Soil Loss Equation (RUSLE) models to map soil loss both qualitatively and quantitatively in the Federal University of Agriculture Abeokuta, Nigeria (FUNAAB). The findings from this study highlighted that from 2000–2022, farmlands and vegetation decreased, whereas built-up and barelands increased respectively. These changes have triggered severe soil erosion, depleted soil properties, and has deteriorated both ecosystem health and landscape fragmentation in the area. By identifying the long-term effects of land use cover change and its potential impacts, the study underscored the utility of the RUSLE model, combined with remotely sensed data and GIS techniques, to enhance vulnerability assessments of the studied soils.

The analysis revealed significant temporal and spatial changes in land use cover, which have escalated the risk of soil erosion in the studied area. The result shows that the soil loss rates increased from 11 to 17 tons per hectare per year, with annual soil loss ranging from 10 to 30 tons per hectare for the area. The soil losses estimated for the studied area can be attributed to shifts in land use cover particularly due to the removal of vegetation cover, expansion of built-up areas, and intensive farming practices.

The findings of this study demonstrate that our approach can effectively assess the impacts of land use cover changes on soil loss. To mitigate soil loss, tree plantations or reforestation and rehabilitation of native vegetation can play significant role in stabilizing soils and reducing soil erosion. Whilst adopting reduced tillage practices, erosion control structures and maintaining planting systems, such as row planting without trampling, are essential for protecting vulnerable areas and reclaiming damaged lands. Future studies should also incorporate socio-economic factors to better understand the full scope of human activities and impact on land use changes. Furthermore, exploring remote sensing spectral indices, like Normalized Difference Water Index (NDWI) and Land Surface Temperature (LST) are critical, given the global concerns about environmental issues related to surface energy heat. These factors will provide a more comprehensive understanding of soil erosion dynamics.

Studies of this nature are vital for developing sustainable land management strategies. They contribute to predicting and mitigating the effects of land use cover changes on soil erosion, which is crucial for maintaining agricultural productivity and environmental health. Future research could refine the RUSLE model by integrating real-time erosion estimations to monitor conservation practices while performing global sensitivity and uncertainty analyses. This would not only support local communities in preserving natural resources but also greatly contribute to environmental protection, sustainable soil management and promote Life on Land. Incorporating socio-economic and environmental data in future studies will enable more targeted, effective policymaking, ensuring interventions that are both environmentally sustainable and socially equitable.

Footnotes

CRediT authorship contribution statement

Tobore Anthony: Conceptualization, data collection, formal analysis, data curation, supervision, resources, software, methodology, investigation, writing – original draft, validation, writing – review & editing. Adedeji Hakeem: Methodology, investigation, validation, writing – review & editing. Jones Similoluwa (MSc student): Review & editing. Bamidele Samuel (MSc student): Review & editing.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Tobore Anthony

Data availability statement

The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Author biographies

Tobore Anthony has a Ph.D. in pedology which focused on addressing the challenges of coupling quantitative and qualitative soil spatial information from Digital soil mapping with environmental process modeling to refine our understanding of the biophysical and processes that influence and managed natural ecosystems. He also has two M. Sc. degrees In (1) Pedology, and (2) Cartography and Geo-information Science. His research focuses on numerical and stochastic modeling of soil properties, soil survey and land use planning, vulnerability assessment, earth system sciences, digital soil maps and soil databases of Nigeria. He is currently lecturing at the department of Soil Science and Land Management Federal University of Agriculture, Abeokuta (FUNAAB), Nigeria, and adjunct lecturer at the department of Geology FUNAAB.

Adedeji Hakeem is a full Professor of Earth system sciences and Environmental Management expert at the department of Environmental Management and Toxicology, Federal University of Agriculture, Abeokuta, Ogun State, Nigeria. His research is in the areas of environmental impact assessment, Soil Survey and Land Use Planning, Sustainable Agriculture, and Earth System researches.

Jones Similoluwa is a trained environmental management toxicologist supervised by ADEDEJI Hakeem and Co-assisted by TOBORE Anthony. He also has an M. Sc. in environmental management toxicology (EMT), after obtaining a first degree in the same discipline, Federal University of Agriculture, Abeokuta, Nigeria. He is currently lecturing at the Federal College of Animal health and livestock production Moor Plantation Ibadan, Nigeria.

Bamidele Samuel is a trained pedologist with a first degree in Soil Science and Land Management supervised by Senjobi Bolarinwa and Co-assisted by TOBORE Anthony at Federal University of Agriculture, Abeokuta, Nigeria. He also has an M. Sc. degree in Plant and Crop Science, University of Delaware, United States of America. He is currently working at KerTec LLC, Lubbock, Texas, United State of America.

References

Abdelkarim

Antunes

Missaoui

, et al. (2023) Assessment and modeling of the spatio-temporal variability of recharge in arid zones: The case of the Oued Zegzaou watershed (Southern Tunisia). In: 1st international virtual seminar on geosciences, Constantine, Algeria, 7–9 March 2023.

Adediji

Tukur

Adepoju

(2010) Assessment of revised universal soil loss equation (RUSLE) in Katsina area, Katsina state of Nigeria using remote sensing (RS) and geographic information system (GIS). Iranica Journal of Energy & Environment 1(3): 255–264.

Adefisan

Bayo

Ropo

(2015) Application of geo-spatial technology in identifying areas vulnerable to flooding in Ibadan metropolis. Journal of Environment and Earth Science 5: 153–166.

Adegboyega

(2019) The impact of soil erosion on agricultural land and productivity in Efon Alaaye, Ekiti State. International Journal of Agricultural Policy and Research 7(2): 32–40.

Aguilera

Valenzuela

Botequilha-Leitão

(2011) Landscape metrics in the analysis of urban land use patterns: A case study in a Spanish metropolitan area. Landscape and Urban Planning 99(3–4): 226–238.

Amsalu

Mengaw

(2014) GIS based soil loss estimation using RUSLE model: The case of Jabi Tehinan Woreda, ANRS, Ethiopia. Natural Resources 5(11): 616–626.

Amundson

Berhe

Hopmans

, et al. (2015) Soil and human security in the 21st century. Science 348: 1261071.

Anderson

Hardy

Roach

, et al. (1976) A land use and land cover classification system for use with remote sensor data. US Geological Survey Professional Paper 964. Washington, DC: US Government Printing Office USGS.

Arabameri

Chen

Loche

, et al. (2020) Comparison of machine learning models for gully erosion susceptibility mapping. Geoscience Frontiers 11: 1609–1620.

10.

Bekele

(2021) Geographic information system (GIS) based soil loss estimation using RUSLE model for soil and water conservation planning in Anka_Shashara watershed, southern Ethiopia. International Journal of Hydrology 5(1): 9–21.

11.

Bewket

Teferi

(2009) Assessment of soil erosion hazard and prioritization for treatment at the watershed level: Case study in the Chemoga watershed, Blue Nile basin, Ethiopia. Land Degradation & Development 20(6): 609–622.

12.

Borrelli

Alewell

Alvarez

, et al. (2021) Soil erosion modelling: A global review and statistical analysis. The Science of the Total Environment 780: Article 146494.

13.

Boufeldja

Baba Hamed

Bouanani

, et al. (2020) Identification of zones at risk of erosion by the combination of a digital model and the method of multi-criteria analysis in the arid regions: Case of the Bechar Wadi watershed. Applied Water Science 10: 121.

14.

Bouyoucos

(1962) Hydrometer method improved for making particle size analyses of soils. Agronomy Journal 54(5): 464–465.

15.

Bünemann

Bongiorno

Bai

, et al. (2018) Soil quality–A critical review. Soil Biology & Biochemistry 120: 105–125.

16.

Chavez

(1996) An improved dark-object subtraction technique for atmospheric scattering correction of multispectral data. Remote Sensing of Environment 24(3): 459–479.

17.

Chuenchum

Tang

(2020) Estimation of soil erosion and sediment yield in the Lancang–Mekong river using the modified revised universal soil loss equation and GIS techniques. Water 12(1): 1–24.

18.

Dabral

Baithuri

Pandey

(2008) Soil erosion assessment in a hilly catchment of north eastern India using USLE, GIS and remote sensing. Water Resources Management 22(12): 1783–1798.

19.

Degife

Worku

Gizaw

(2021) Environmental implications of soil erosion and sediment yield in Lake Hawassa watershed, south-central Ethiopia. Environmental Systems Research 10: 28.

20.

Dike

Alakwem

Nwoke

, et al. (2018) Potential soil loss rates in Urualla, Nigeria using Rusle. Global Journal of Science Frontier Research: H Environment and Earth Science 18(2): 1–7.

21.

Dubrovsky

Burow

Clark

, et al. (2010) The quality of our nation’s water – nutrients in the nation’s stream and groundwater, 1992-2004. U. S Geological Survey Circular 1350. Washington, DC: US Geological Survey.

22.

Ebabu

Tsunekawa

Haregeweyn

, et al. (2019) Effects of land use and sustainable land management practices on runoff and soil loss in the Upper Blue Nile basin, Ethiopia. The Science of the Total Environment 648: 1462–1475.

23.

Elnashar

Zeng

, et al. (2021) Soil erosion assessment in the Blue Nile Basin driven by a novel RUSLE-GEE framework. Science of the Total Environment 793: 148466.

24.

Fagbohun

Adeleye

AYB

Odeyemi

, et al. (2016) GIS-based estimation of soil erosion rates and identification of critical areas in Anambra sub-basin, Nigeria. Modeling Earth Systems and Environment 2: 159.

25.

FAO 2007. Soil and Terrain Database of Central Africa – DR of Congo, Burundi and Rwanda (SOTERCAF 1.0).

26.

FAO (2020) Committee on World food security. Available at: http://www.fao.org/cfs/en/ (accessed 23 September 2020).

27.

FAO and ITPS (2015) Status of the world’s soil resources (SWSR) remain report. Rome, Italy: Food and Agriculture Organization of the United Nations and Intergovernmental Technical Panel on Soil, 1–650.

28.

FAO and UNEP (1984) Provisional Methodology for Assessment and Mapping of Desertification. Rome, Italy: Food and Agriculture Organization of the United Nations (FAO), 84.

29.

Fistikoglu

Harmancioglu

(2002) Integration of GIS with USLE in assessment of soil erosion. Water Resources Management 16: 447–467.

30.

Flanagan

Laflen

(1997) The USDA water erosion prediction project (WEPP). Eurasian Soil Science 30: 524–530.

31.

Foody

(2002) Status of land cover classification accuracy assessment. Remote Sensing of Environment 80: 185–201.

32.

Gashaw

Tulu

Argaw

(2017) Erosion risk assessment for prioritization of conservation measures in Geleda watershed, Blue Nile basin, Ethiopia. Environmental Systems Research 6(1): 1–14.

33.

Gevaert

Belgiu

(2022) Assessing the generalization capability of deep learning networks for aerial image classification using landscape metrics. International Journal of Applied Earth Observation and Geoinformation 114: 103054.

34.

Girma

Gebre

(2020) Spatial modeling of erosion hotspots using GIS-RUSLE interface in Omo-Gibe river basin, Southern Ethiopia: Implication for soil and water conservation planning Environmental System Research 9: 19. DOI: 10.1186/s40068-020-00180-7.

35.

Hagos

( 2020). Estimating landscape vulnerability to soil erosion by RUSLE model using GIS and remote sensing: A case of Zariema watershed, Northern Ethiopia. Research Square. Epub ahead of print 31 December 2020. DOI: 10.21203/rs.3.rs-136586/v1.

36.

Hurni

(1989) Ecological issues in the creation of famines in Ethiopia. Paper presented on the national conference on disaster prevention and preparedness for Ethiopia, Addis Ababa.

37.

İrvem

Topaloğ lu

Uygur

(2007) Estimating spatial distribution of soil loss over Seyhan River Basin in Turkey. Journal of Hydrology 336: 30–37.

38.

Johnson

JMF

Archer

Weyers

, et al. (2011) Do mitigation strategies reduce global warming potential in the northern U.S. Corn Belt? Journal of Environmental Quality 40: 1551–1559.

39.

Kafy

Foyezur Rahman

ANM

Rakib

, et al. (2021). Assessment and prediction of seasonal land surface temperature change using multi-temporal Landsat images and their impacts on agricultural yields in Rajshahi, Bangladesh. Environmental Challenges 4: 100147.

40.

Kapur

Akça

(2020) Soil degradation: Global assessment. In: Wang

(ed.) Landscape and Land Capacity. Boca Raton: CRC Press, 199–210.

41.

Kidd

Field

McBratney

, et al. (2018). A preliminary spatial quantification of the soil security dimensions for Tasmania. Geoderma 322: 184–200.

42.

Kim

Saunders

Finn

(2005) Rapid assessment of soil erosion in the Rio Lempa Basin, Central America, using the universal soil loss equation and geographic information systems. Environmental Management 36(6): 872–885.

43.

Lal

(2001) Soil degradation by erosion. Land Degradation & Development 12: 519–539.

44.

Lal

Bouma

Brevik

, et al. (2021) Soils and sustainable development goals of the United Nations: An IUSS perspective. Geoderma Regional 25: e00398.

45.

Lee

(2004) Soil erosion assessment and its verification using the Universal soil loss equation and geographic information system: A case study at Boun, Korea. Environmental Geology 45: 457–465.

46.

Wang

Cao

, et al. (2009) Research on the distribution characteristics of Zn, Cd in the soil of Jinding Pb-Zn deposit, Lanping County. Geological Review 55: 126–133.

47.

Zhang

, et al. (2019) Anisotropic characteristic of artificial light at night–Systematic investigation with VIIRS DNB multi-temporal observations. Remote Sensing Environment 233: 111357.

48.

Lorentz

Schulze

(1994) Sediment yield. In: Schulze

(ed.) Hydrology and Agrohydrology: A Text to Accompany the ACRU 3.00 Agro Hydrological Modeling System, TT69/95. Pretoria: Water Research Commission, AT16-1 to ATI6-34.

49.

McBratney

Field

Koch

(2014) The dimensions of soil security. Geoderma 213: 203–213.

50.

McCool

Foster

Mutchler

, et al. (1989) Revised slope length factor for the Universal Soil Loss Equation. Transactions of the ASAE 32: 1571–1576.

51.

McCool

Foster

Renard

, et al. (1995) The Revised Universal Soil Loss Equation, vol. 9. San Antonio, TX: Department of Defense/Interagency Workshop on Technologies to Address Soil Erosion on Department of Defense Lands.

52.

McGarigal

(2002) FRAGSTATS: Spatial pattern analysis program for categorical maps. Computer Software Program Produced by the Authors at the University of Massachusetts, Amherst: University of Massachusetts. https://www.umass.edu/landeco/research/fragstats/fragstats.html

53.

McGarigal

Marks

(1995) FRAGSTATS: Spatial pattern analysis program for quantifying landscape structure. Report PNW-351. Corvallis, OR: US Department of Agriculture, Forest Service, Pacific Northwest Research Station, 1–122.

54.

McLean

(1982) PH and lime requirements. In: Page

ALE

(ed.) Methods of Soil Analysis. Poster session 3, GIS development proceedings, ACRS. Madison, WI: American Society of Agronomists Inc.

55.

Makinde

Oyebanji

(2020) Remote sensing and GIS application to erosion risk mapping in Lagos. Nigerian Journal of Environmental Sciences and Technology 4(1): 40–53.

56.

Masroor

Sajjad

Rehman

, et al. (2022) Analyzing the relationship between drought and soil erosion using vegetation health index and RUSLE models in Godavari middle sub-basin, India. Geoscience Frontiers 13(2): 101312.

57.

Mesfin

Saito

Fürst

, et al. (2019) Quantifying and mapping of water-related ecosystem services for enhancing the security of the food-water-energy nexus in tropical data – sparse catchment. Science of the Total Environment 646: 573–586.

58.

Mondal

Jeganathan

(2022) Effect of scale, landscape heterogeneity and terrain complexity on agriculture mapping accuracy from time-series NDVI in the Western-Himalaya region. Landscape Ecology 37: 1–25.

59.

Nearing

Jetten

Baffaut

, et al. (2005) Modeling response of soil erosion and runoff to changes in precipitation and cover. Catena 61: 131–154.

60.

Negese

Fekadu

Getnet

(2021) Potential soil loss estimation and erosion-prone area prioritization using RUSLE, GIS, and remote sensing in Chereti Watershed, Northeastern Ethiopia. Air, Soil and Water Research 14: 1–17.

61.

Nguemezi

Tematio

Yemefack

, et al. (2020) Soil quality and soil fertility status in major soil groups at the Tombel area, South-West Cameroon. Heliyon 6(2):e03432.

62.

Nkwunonwo

Tobore

Nwaka

(2024) Ecosystem-based approach to local flood risk management in Ogun State, Nigeria: Knowledge, and pathway to Actualisation. Natural Hazard Research 4(3): 357–373.

63.

Oicha

Cornelis

Verplancke

, et al. (2010) Short-term effects of conservation agriculture on Vertisols under tef (Eragrostis tef (Zucc.) Trotter) in the northern Ethiopian highlands. Soil and Tillage Research 106: 294–302.

64.

Oldeman

(1998) Soil degradation: A threat to food security? Report 98/01. Wageningen, the Netherlands: International Soil Reference and Information Centre.

65.

Opeyemi

Abidemi

Victor

(2019) Assessing the impact of soil erosion on residential areas of Efon-Alaaye Ekiti, Ekiti-State, Nigeria. International Journal of Environmental Planning and Management 5: 23–31.

66.

Orimoloye

Mazinyo

Nel

, et al. (2018) Spatiotemporal monitoring of land surface temperature and estimated radiation using remote sensing: Human health implications for East London, South Africa. Environmental Earth Sciences 77(3): 77.

67.

Panagos

Borrelli

Meusburger

, et al. (2014) Estimating the soil erosion cover-management factor at the European scale. Land Use Policy 48: 38–50.

68.

Pazúr

Price

Atkinson

(2021) Fine temporal resolution satellite sensors with global coverage: An opportunity for landscape ecologists. Landscape Ecological 36: 2199–2213.

69.

Renard

Foster

Weesies

, et al. (1997) Predicting Soil Erosion by Water: A Guide to Conservation Planning with the Revised Universal Soil Loss Equation (RUSLE). Agriculture Handbook 703. Washington DC: USDA, 404.

70.

Reshma

Uday

(2012) Integrated approach of universal soil loss equation (USLE) and geographical information system (GIS) for soil loss risk assessment in Upper South Koel Basin, Jharkhand. Journal of Geographic Information System 4: 588–596.

71.

Rousta

Sarif

Gupta

, et al. (2018) Spatiotemporal analysis of land use/land cover and its effects on surface urban heat island using Landsat data: A case study of Metropolitan City Tehran (1988-2018). Sustainability (Switzerland) 10(12): 4433.

72.

Safadoust

Doaei

Mahboubi

, et al. (2016) Long-term cultivation and landscape position effects on aggregate size and organic carbon fractionation on surface soil properties in the semi-arid region of Iran. Arid Land Research and Management 30(4): 345–361.

73.

Salako

Obi

Lal

(2013) Comparative assessment of several rainfall erosivity indices in Southern Nigeria. Soil Technology 4(1): 93–97.

74.

Senjobi

Tobore

Oyerinde

(2019) Decreasing suitable land for Maize production in Nigeria; assessments with the synchrony of conventional and geospatial techniques. Discovery Agriculture 5: 136–145.

75.

Shinde

Tiwari

Singh

(2010) Prioritization of micro watersheds on the basis of soil erosion hazard using remote sensing and geographic information system. International Journal of Water Resources and Environmental Engineering 2(3): 130–136.

76.

Smyth

Montgomery

(1962) Soils and Land Use in Central Western Nigeria. Ibadan, Nigeria: Western Nigerian Government Press, 217.

77.

Sotona

Salako

Adesodun

(2013) Soil physical properties of selected soil series in relation to compaction and erosion on farmers’ fields at Abeokuta, southwestern Nigeria. Archives of Agronomy and Soil Science 60(6): 841–857.

78.

Sun

Shao

Liu

(2014) Assessing the effects of land use and topography on soil erosion on the Loess Plateau in China. Catena 121: 151–163.

79.

Tahmoures

Honarbakhsh

Afzali

, et al. (2022) Fractal features of soil particles as an indicator of land degradation under different types of land use at the watershed scale in southern Iran. Land 11: 2093.

80.

Tibebu

Gizaw

Gete

, et al. (2018) Assessing the soil erosion control efficiency of land management practices implemented through free community labor mobilization in Ethiopia. International Soil and Water Conservation Research 6: 87–98.

81.

Tobore

Anoke

(2025) Geomodeling soil quality using remote sensing and land use metrics in Abeokuta, Nigeria. Discover Geoscience 3: 3.

82.

Tobore

Nkwunonwo

Senjobi

(2024) Combined remote sensing and multi-criteria analysis of wetland soil potential for rice production: Case study of Ogun River basin, Nigeria. African Geographical Review 43(1): 32–59.

83.

Tobore

Samuel

(2022) Wetland change prediction of Ogun-River Basin, Nigeria: Application of cellular automata Markov and remote sensing techniques. Watershed Ecology and the Environment 4: 158–168.

84.

Tobore

Senjobi

Oyerinde

(2021) Spatio temporal analysis and simulation pattern of land-use and land cover change in odeda peri-urban of Ogun State. Nigeria Jordan Journal of Earth, and Environmental Sciences 12(4): 326–336.

85.

Tully

Sullivan

Weil

, et al. (2015) The state of soil degradation in sub-Saharan Africa: Baselines, trajectories, and solutions. Sustainability 7: 6523–6552.

86.

Ufoegbune

Oyedepo

Awomeso

, et al. (2010) Spatial analysis of municipal water supply in Abeokuta metropolis, South western Nigeria. In: REAL CORP 2010 proceedings (eds Schrenk

Vasily

Popovich

, et al.) Vienna, Austria, 18–20 May 2010. Available at: http://www.corp.at (accessed 03 May 2010).

87.

United Nations (2015) Transforming our world. The 2030 Agenda for Sustainable Development. https://sustainabledevelopment.un.org/post2015/transformingourworld/publication

88.

Weil

Brady

Weil

(2016) The Nature and Properties of Soils. Fifteenth Edition, Columbus: Pearson.

89.

Widyatmanti

Umarhadi

(2022) Spatial modeling of soil security in agricultural land of Central Java, Indonesia: A preliminary study on capability, condition, and capital dimensions. Soil Security 8: 100070.

90.

Wischmeier

Smith

(1978) Predicting Rainfall Erosion Losses: A Guide to Conservation Planning. The USDA Agricultural Handbook No. 537, Maryland.

91.

Wolka

Tadesse

Garedew

, et al. (2015) Soil erosion risk assessment in the Chaleleka wetland watershed, Central Rift Valley of Ethiopia. Environmental Systems Research 4: 5.

92.

Yesuph

Dagnew

(2019) Soil erosion mapping and severity analysis based on RUSLE model and local perception in the Beshillo Catchment of the Blue Nile Basin, Ethiopia. Environmental Systems Research 8(1): 1–21.

93.

Zerihun

Mohammedyasin

Sewnet

, et al. (2018) Assessment of soil erosion using RUSLE, GIS and remote sensing in NW Ethiopia. Geoderma Regional 12: 83–90.

Assessing land use/cover changes to soil erosion vulnerability using machine learning and RUSLE model

Abstract

Keywords

Introduction

Materials and methods

Study area

Soil mapping

Laboratory analysis

Spatial remote sensing analysis

RUSLE factors estimation

Rainfall erosivity (R)

Soil erodibility (K)

Slope length and steepness (LS)

Cover management (C)

Conservation practice (P)

Results and discussion

Land use and accuracy assessment

Soil condition of FUNAAB

Soil severity loss

Rainfall (R) erosivity

Soil (K) erodibility

Slope length and steepness (LS)

Cover management (C) factor

Erosion control practice

Soil severity loss along different land cover

Conclusion and recommendation

Footnotes

CRediT authorship contribution statement

Declaration of conflicting interests

Funding

ORCID iD

Data availability statement

Author biographies

References