Determine Soil Loss Response to Agroforestry Land Use Change (ALUC) Using RUSLE Based GIS and RS in Gilgel Gibe I Catchment, Southwestern, Ethiopia

doi:10.21203/rs.3.rs-4865311/v1

Download PDF

Research Article

Determine Soil Loss Response to Agroforestry Land Use Change (ALUC) Using RUSLE Based GIS and RS in Gilgel Gibe I Catchment, Southwestern, Ethiopia

https://doi.org/10.21203/rs.3.rs-4865311/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Soil erosion is a serious environmental problem, reducing land productivity and endangering people's livelihoods. Land use land cover (LU/LC) has long been thought to be the primary cause of soil erosion, but ALUC has received little attention. The aim of this study is to compare the soil loss response to ALUC with other LU/LC from 1988 to 2021 in Nadi and Yedi micro watersheds of Gilgel Gibe I catchment. Data was obtained from Landsat images (TM, ETM+, and OLI/TIRS) of 1988, 2004, and 2021 and hybrid algorithm with maximum livelihoods was used to classify it. Besides, rainfall, soil digital map, and digital elevation model were used in this study. The data was analyzed using RUSLE model in ArcGIS. As a result cultivated land was increased between 1988 and 2021, followed by agroforestry and settlement in both micro watersheds. From the entire watershed, 1027.9 and 607.7 tons/ha/year total soil loss was computed between 1988 and 2021 in Nadi and Yedi watersheds, respectively. In the Nadi and Yedi watersheds, the average annual soil loss was (64; 46.43), (66.71; 48.51), and (68.10; 45.88) tons/ha/year in 1988, 2004, and 2021, respectively. The findings showed that approximately (9%; 10%), (11%; 12%) and (12%; 11%) of the land were exposed to very severe erosion during 1988, 2004, and 2021 in Nadi and Yedi micro watersheds, respectively. The soil loss rate can be significantly decreased with the use of ALUC. Compare to other LU/LC, ALUC provides a significant impact on reducing soil loss rate. Thus, to sustainably restore degraded land and limit the rate of soil erosion enhancing agroforestry land use practices in the study area.

LU/LC

Agroforestry

Soil erosion

RUSLE

Gilgel Gibe

Soil erosion is a continuous environmental and ecological problems that undermines land productivity capacity and deteriorates people's socioeconomic standing (Endalamaw et al., 2021; Moisa et al., 2022), and it is an endless and gradual process where water and wind are its agents (Hassen et al., 2021). It has both on-site and off-site effects. On-site effect on land use from where topsoil is detached and taken away by run-off that directly damages soil's biological and physiochemical properties, resulting in decline crop production and boosting food insecurity (Guadie et al., 2020; Jiru & Wegari, 2022). The off-site effect is the result of on-site effect that is almost marked at downstream and destroy infrastructure, eutrophication, and diminish the life span of hydroelectric dam storage capacity (Adugna & Cherie, 2021; Getachew et al., 2021; Wolka et al., 2018). For instance, it reduced the storage capacity of Aswan High Reservoir by 4% within 48 years in Egypt; Sennar reservoir by 85% within 85 years in Sudan, and deteriorated Lake Cheleleka (Degife et al., 2021) and Angereb Reservoir by 46% within 19 years in Ethiopia (Moussa, 2019). Besides, Gilgel Gibe I hydroelectric power is also the victim of this issue where this study fixated (Demissie et al., 2013; Estifanos & Gebremariam, 2019).

Accelerated soil erosion is erupted due to population pressures couple with land shortage that provoked LU/LC. Obviously LU/LC occurred when people modified land to meet their requirements (Borrelli et al., 2017; Derebe et al., 2022; Dibaba et al., 2020; Temesgen et al., 2021) and physical factors like topography, soil type and climate change that exacerbates soil erosion, soil loss, and sedimentation (Aneseyee et al., 2019; Belay & Mengistu, 2021; Gong et al., 2022; Teshome et al., 2022). This could be a well-known issue that has been overlooked it’s positively. However, ground truth ratified the existence of negative and positive responses of LU/LC that occurred simultaneously during lost or gained. For example, changing land use to cultivated land has a negative impact on the environment and land production capacity, whereas switching to agroforestry has a comparatively beneficial impact (Belay & Mengistu, 2021; Desalegn et al., 2023; Habte et al., 2021). It is a modern alternative land use management system that is chiefly used by smallholder farmers on their farmland because of its multiple benefits like improving soil fertility and quality, carbon sequestration, increasing biodiversity, socioeconomic benefits, and demonstrates the efficiency of a small unit of land (Endale et al., 2024; Jiru et al., 2020; Mbow et al., 2014; Ndalama et al., 2015; Wari et al., 2019). Several scholars evaluated LU/LC, including ALUC; nevertheless, insufficient research was undertaken on the impacts of LU/LC on soil erosion (Beyene et al., 2019; Desalegn et al., 2023; Meragiaw et al., 2022; Temesgen et al., 2021).

In Ethiopia, quantifying soil loss and identifying the most degraded area were undertaken at different places by various models, such as Universal Soil Loss Equation (USLE), revised universal Soil Loss Equation (RUSLE), and Soil and Water Assessment Tool (SWAT) (Williams, 1975; Wischmeier & Smith, 1978). Amongst, RUSLE model is an empirical model that frequently used for more than a decade (Gelagay & Minale, 2016; Moisa et al., 2022). It’s well advanced, accepted, and applied in the globe including Sub-Saharan African countries (Ghosal & Bhattacharya, 2020). This is due to its compatibility, applicable, flexibility and cost-effective and it’s more suitable with Geographic Information Systems (GIS) and remote sensing (RS) for small sizes specifically at watershed level (Moisa et al., 2022). Many scholars quantified soil loss by using this model for a single period. For instance, 47.4 t/ha/yr. from Koga watershed (Gelagay & Minale, 2016); 34.26 t/ha/yr. from Gobele watershed (Woldemariam et al., 2018); 32.84 t/ha/yr. from Megech watershed (Getu et al., 2022); 64.2 t/ha/yr. from Tashat watershed (Mengie et al., 2022) and 60 t/ha/yr. from the Omo-Gibe basin (Girma & Gebre, 2020). Besides, except ALUC, effects of LU/LC on soil loss were estimated (Masha et al., 2021; Moisa et al., 2022; Weldu & Harka, 2020).

Quantifying soil loss response to the spatial distribution of LU/LC and identifying highly prone land use plays an incredible role not only for introducing appropriate land management tools but also to scaling up land use that less vulnerable to soil erosion. Thus, studying the trends of LU/LC influences on soil loss, including ALUC, is a fundamental in the Gilgel Gibe I hydroelectric power plant. Because it is increasingly today than in the previous decade across the country (Desalegn et al., 2023; Endale et al., 2024; Meragiaw et al., 2022; Temesgen et al., 2021) particular in the study area. Information on the spatial distribution and trends of soil loss status response to LU/LC, comparing with ALUC, is crucial for policymakers, environmentalists, local communities and governments to make strategic decisions about land use planning and sustainable land management practices to mitigate land degradation. The objectives of this study were: to assess the trends of ALUC in comparison with other LU/LC between 1988 and 2021 years and to quantity soil loss accordingly in the Gilgel Gibe I catchment in southwest Ethiopia.

2.1 Study area description

The study was conducted at Nadi and Yedi micro watersheds of Gilgel Gibe I catchment of the Oromia Regional state of Ethiopia. Nadi and Yedi micro watersheds are geographically located between 7^o 20' 0'' to 8^o10' 0'' N latitudes and 36o 50' 0'' to 37^o 40' 0'' E longitudes (Fig. 1). Nadi micro watershed altitude ranges from 1640 to 2950 meter above sea level (masl). Its annual temperature ranges from 20 to 25℃ with a mean rainfall of 1650 mm. Yedi elevation varies from 1160 to 2840 masl and its mean annual rainfall of 1550 mm with a mean annual temperature of 19℃. From the Gilgel Gibe I catchment total area of 5125 km² Nadi micro-watershed covers 115.55 km² (3.2%) whereas Yedi micro-watershed covers 89.54 km² (3.3%) area (Demissie et al., 2013). The catchment is commonly known for its bimodal rain season, which includes unpredictable short rains from March to April and the main rain season from June to September. Sometimes, the rainy season of the catchment may extend from mid-February to early October and receive a mean annual rainfall of 1347mm with a mean annual temperature of 19 to 24.78 degrees (Tesfaye & Ameyu, 2021). The Gilgel Gibe I catchment landscape is dotted with height-relief hills and mountains, particularly around the powerhouse intake. The landscape topography slightly increases from the outlet of each watershed to the inlet it’s more rugged in Nadi than Yedi. Cultivated, shrubs/bushes, settlements, forest (both planted and natural), degraded land, and grazing/grass are the common land uses in the catchment. Major soil of the basin includes Nitosols, Cambisols, Fluvisols, Acrisols, Vertisols, and Planosols with the Eutric Nitosols being dominant (Tesfaye & Ameyu, 2021).

2.2 Farming systems of the catchment

Crop-livestock mixed farming practices are commonly practiced by smallholder farmers in both micro watersheds. The socio-economic and livelihoods of each micro watershed community are almost similar. Annual and perennial crops such as cereals, pulses, fruit crops, and vegetables are produced for household consumption and income in the area. Sorghum, barley, wheat, and maize are accounted as staple food crops (Broothaerts et al., 2012), while coffee, maize, and pepper are used as a source of income (Tolla et al., 2016). Kinds of livestock herds are cattle, small ruminants, poultry, equines, and beehives. Indigenous livestock and traditional animal-keeping systems (free grazing) are mainly practiced in the area, particularly in Nadi micro watershed. Soil erosion was so very high that more than 28% of the districts were prone to mean soil loss (> 50 tons/ha/year) in the areas where both micro watersheds were located (Lencha & Moges, 2015). In Yedi, but not in Nadi micro watersheds, various watershed management strategies (e.g., bund constructions, water harvesting structural, and gully treatment, biological and area exclosure) have been undertaken since 2011. Even though Nadi micro watershed does not have an officially launched watershed management intervention, interested smallholder farmers have applied a few soil and water conservation practices for mitigating land degradation. However, since 2019, most degraded land and shrubs/bush areas have been restored through green legacy collaboration with the local community. Consequently, this was considered as a milestone for sustainable land management and reducing soil erosion in both micro watersheds.

2.3 Sample design and selection

Topographic map (1:50,000 scales) reading and reconnaissance field survey were undertaken before the actual survey to capture information about the Gilgel Gibe I catchment, Gilgel Gibe I dam, and soil erosion. During this, the micro watershed per district and its boundary, land degradation, and land use types including traditional agroforestry practice were identified with district natural resource conservation experts, DA, and kebele's leader. Nadi and Yedi micro watersheds were selected purposively based on their proximity to the buffer zone of Gilgel Gibe I dam in that Nadi is relatively far away whereas Yedi approaches to Gilgel Gibe I dam/buffer zone of Gilgel Gibe I dam, land conservation intervention and dominant agroforestry based land management practice.

2.4 Data source and data analysis method

2.4.1 Data source

Both primary and secondary data were used. Data were collected from different sources, such as data for erosivity analysis (R-factor) of 33 years (1988–2021) rainfall from eleven stations were received from Ethiopia Metrological Agency (EMA) (Table 1). Data for soil erodibility analysis (K-factor) was retrieved from the Ministry of Water and Energy (MoWE) as a shape file. Data for topography (LS-factors), cover and management (C-factor), and supporting practice (P-factor) were analyzed from the remotely sensed Shuttle Radar Topographic Mission (30 × 30 m) Digital Elevation Model (SRTM-DEM) downloaded from the United States Geological Survey (USGS) (https://earthexplorer.usgs.gov/). For supervised LU/LC classification 250 reference points were received from the field using a Global Positioning System (GPS) based on recommended (Congalton & Green, 2019). Detailed source and descriptions of used data (Table 2). Gilgel Gibe I hydroelectric power building began in 1988, but stopped in 1994 and was resumed by a new construction firm in 1995, we used 1988 year as a benchmark for evaluating the status of soil loss response to LU/LC over the long year of 1988–2021.

Table 1

Data source and descriptions of data used in this study
Data category		Source	Descriptions				Aims
Data category		Source	Sensors	Path/Row	Spatial resolution (m)	Year	Aims
Landsat images	Landsat 5	USGS	T M	169/055	30*30	1988	For LU/LC changes and, P and C factor
	Landsat 7	USGS	ETM+	169/055	30*30	2004
	Landsat 8	USGS	OLI/TIRS	169/055	30*30	2021
Rainfall		EMA				1988–2021	R-factor
Soil data		MoWE		Soil shape file format			K-factor
DEM		USGS		ASTER 30			LS- factor

OLI is Operational Land Imager, TIRS is Thermal Infrared Sensor, ETM + is Enhanced Thematic Mapper Plus, and TM is Thematic Mapper

2.4.2 Data analysis method

A mean annual soil loss response to LU/LC was analyzed following the developed methodological framework (Fig. 2). Accordingly, we classified it into three procedures based on five RUSLE factors and their sources. (1) RUSLE model settle (Eq. 1), (2) analysis individual factor based on their features, and finally calculate mean soil loss.

2.4.2.1 RUSLE model

RUSLE is the most famous empirical model for predicting the long-term mean soil loss by water erosion (e.g., sheet and rill) (Prasannakumar et al., 2012; Zerihun et al., 2018) and modified for Ethiopian conditions (Hurni, 1985; Renard, 1997). Many scholars were applied it encompassing five parameters (rainfall, soil, topography, cover and management and supporting practice) (Bekele, 2021; Kebede et al., 2021; Tessema et al., 2020) using equation (Eq. 1).

A = R* K* LS* C* P (1)

Where: A represents annual soil loss (ton ha^− 1yr^− 1), R represents rainfall erosivity factor (MJmm/ha/hr/yr), K represents soil erodibility factor (Mg/ha/MJ/mm), LS represents topographic factor (slope length and steepness), C represents cover and management factor (dimensionless) and P represents support practice factor (dimensionless).

2.4.2.2 Erosivity factor (R)

This is a result of rainfall/raindrop kinetic energy and its 30-minute duration, quantity, and intensity level on soil erosion. However, the 30-minute rainfall intensity data is not available in less developed countries, including Ethiopia (particularly in the study area), because it requires automatic rain gauges (Hurni, 1985). To solve this gap alternative regression equation was used that developed by (Hurni, 1985) at the national level based on the climatic condition of the country (Eq. 2). Its acceptability was confirmed in different parts of Ethiopia, for instance, in Welmel watershed of Oromia region (Tessema et al., 2020); in Agewmariam watershed of Amhara region (Girmay et al., 2020); in Anka-shashara watershed of SNNP region (Bekele, 2021) and in Omo-Gibe basin (Girma & Gebre, 2020).

R = (0.562*P) − 8.12 (2)

Where: R represents the erosivity factor (MJmm ha/hr/yr) and P represents the mean annual rainfall (mm) of 33 years of 1988–2021. The most and simplest Arithmetic mean method was used for estimating a missed month rainfall before using it for analysis (Alem, 2022; De Silva et al., 2007). Its map was generated by using Inverse Distance Weighted (IDW) through interpolating the raster surface point in ArcGIS 10.8 spatial analyst tools.

2.4.2.3 Soil erodibility factor (K)

This indicates the vulnerability of soil to soil erosion (detachment and transport) and the ability of soil against to raindrop of the rate and amount of run-off. It is governed by soil properties (e.g. texture, structure, organic matter, porosity, color and permeability) (Panagos et al., 2015). Soil erodibility, detachability, infiltration rate, water storage capacity, and water flow paths all are influenced by soil characteristics (Alem, 2022; Assefa et al., 2015). Soil texture and organic carbon content were derived from shape file of MoWE. Its value was computed by using equation (Eq. 3), which was adopted from (Williams, 1995). Finally, soil type and K- factor map of each watershed was generated by using ArcGIS 10.8 spatial analysis tools (Fig. 6a & b).

K = f_sand * f_cl - _si * f_OC * f_{hi sand} (3)

Where: K represents the soil erodibility factor, f_sand represents a factor that gives low soil erodibility factors for soils with high coarse-sand contents and high values for soils with little sand, f_cl−si represents a factor that gives low soil erodibility factors for soils with high clay to silt ratios, f_OC represents a factor that reduces soil erodibility for soils with high organic carbon content, and f_{hi sand} represents a factor that reduces soil erodibility for soils with extremely high sand contents. According to (Neitsch et al., 2002), each factor was estimated as following (Eq. 4–7):

f_sand = $\:\left(0.2+0.3\text{*}\text{E}\text{x}\text{p}\left[-0.256{\text{m}}_{\text{s}\text{a}}\text{*}\left(1-\frac{{\text{m}}_{\text{s}\text{i}\text{l}\text{t}}}{100}\right)\right]\right)$ (4)

f_cl−si = $\:{\left(\frac{{\text{m}}_{\text{s}\text{i}\text{l}\text{t}}}{{\text{m}}_{\text{c}\text{l}}+{\text{m}}_{\text{s}\text{i}\text{l}\text{t}}}\right)}^{0.3}$ (5)

f_OC = $\:\left(1-\frac{0.25\text{*}\text{O}\text{C}}{\text{O}\text{C}+\text{E}\text{x}\text{p}\:[3.72-2.95\text{*}\text{O}\text{C}]}\right)$ (6)

f_hisand = $\:\left(1-\frac{0.7\text{*}\left(1-\frac{{\text{m}}_{\text{s}\text{a}}}{100}\right)}{\left(1-\frac{{\text{m}}_{\text{s}\text{a}}}{100}\right)+\text{E}\text{x}\text{p}\left[5.51+22.9\:\left(1-\frac{{\text{m}}_{\text{s}\text{a}}}{100}\right)\right]\:}\right)$ (7)

Where: $\:{\text{m}}_{\text{s}\text{a}}$ = the percent of sand content (0.05–2.00 mm diameter particles); $\:{\text{m}}_{\text{s}\text{i}\text{l}\text{t}}$= the percent of silt content (0.002–0.05 mm diameter particles); $\:{\text{m}}_{\text{c}\text{l}}$ = the percent of clay content (< 0.002 mm diameter particles), and OC = the percent of organic carbon content.

2.4.2.4 Slope length (L) and slope gradient (S) factor

Topographic factor (Slope length and slope steepness) is highly influence soil erosion (Lu et al., 2012). Slope length (L) is the distance of the slope that exists between the inlet at initial water inflow to the outlet at the soil becomes started to deposited or it is will out whereas slope gradient (S) reflects the inclination of land surface in which either it helps to swift or resists the flow velocity of run-off. Slope length can be calculated by considering longer slope as higher soil loss pressure without considering three dimensional complex nature of landscape (Robert & Hilborn, 2000). This ways of slope length calculation was undermined by a sort of scholars, that at presence of flow convergence and divergence soil loss was didn’t governed by slope length for three dimensional complex of terrain, but by upslope contribution area (Moore & Burch, 1986; Simms et al., 2003). Therefore, in this study, the reorganized and advanced approach factor (SL) that includes three dimensional complex terrain and upslope contribution area were applied (Gelagay & Minale, 2016; Moore & Burch, 1986; Simms et al., 2003) (Eq. 8). Practically, Yirgu (2022), Gelagay & Minale (2016) and Kebede et al. (2021) were used this equation in Ethiopia.

LS = $\:\left(\frac{\left(\text{F}\text{A}\right)\:\text{*}\:\left(\text{C}\text{e}\text{l}\text{l}\:\text{s}\text{i}\text{z}\text{e}\right)}{22.1}\right)$^0.4 _* $\:\left(\frac{sin\left(slope\right)}{0.0896}\right)$^1.3 (8)

Where: FA represents flow accumulation which derived from flow directions and fill while sin (slope) in degree was extracted by mask from the DEM with a cell size of 30*30m.

2.4.2.5 Land use and cover management factor (C)

LU/LC is interlocked with C-factor in estimating soil loss. This describes about land cover management through LU/LC classes that influence soil loss by water erosion in the study area. Therefore, based on land use classes, we assigned C-value through adopting it from early literatures (Table 4). This ways of assigning C-value was a commonly known by a sort of scholars (Bekele, 2021; Girma & Gebre, 2020; Yirgu, 2022).

2.4.2.6 Conservation practice factor (P)

Conservation practice factor (dimensionless) is the ratio of soil loss with limited erosion controlling supporting practice to soil loss consistency with up/down slope of cultivated land (Wischmeier & Smith, 1978). Its value ranged from 0 (good conservation practice) to 1 (no conservation practice). Due to a lack of enough data for each conservation measure (contour farming, terrace,) in the study area, we used a classified land use for determining P-value based on reclassifying land use into agricultural and non-agricultural classes. Accordingly, P-value was assigned for agricultural land through reclassifying slope up to six classes but governed by cropland slope and for non-agricultural land (e.g., grassland, settlement, shrubs/bush, and agroforestry) regardless of slope as recommended by (Wischmeier & Smith, 1978) (Table 6). In this study, the slope ranges from 0 to 104.39 percent and from 0 to 69.22 percent in Nadi and Yedi micro, respectively.

2.5 Land use classification and accuracy test

In this study, LU/LC was classified into five classes (Table 3). Prior remotely sensed image classification, geometric rectification, atmospheric correction, bad line detection, radiometric calibration and topographic correction were applied as image pre-processing technique (Gitima et al., 2022; Tewabe & Fentahun, 2020). Sensor noise, haze, corrective data loss, and missed lines which happened due to solar position and satellite calibration, were solved by radiometric and geometric corrections methods (Aneseyee et al., 2019). All satellite imagery of each watershed was projected into World Geodetic System_84, Universal Transversal Mercator zone of 37N (WGS_84_UTM_Zone_37N). Hybrid algorithms, unsupervised (visual interpretation) and supervised classification were applied. Supervised classification with maximum likelihood algorithm procedure was applied for classifying of LU/LC classes from three Landsat images (1988, 2004 and 2021). In this classification 500/2 reference points (50 points per LU/LC classes) were used, which excluded from accuracy assessment. This a maximum sample recommended for image classifications of less than 12 classes that 50 samples per class (Lillesand et al., 2015) reference point for 1988 and 2004 were collected through Google earth, and top maps, while original Landsat images used for 2021 by using GPS. Totally, 450 ground reference points were randomly selected (250 from Nadi and 200 from Yedi) to compared with classified image by confusion matrix or error matrix (Hassan et al., 2016).

Accuracy assessment was done by overall accuracy (OA) and Kappa coefficient (Girma & Gebre, 2020). OA is the ratio of correctly fixed number of pixels (diagonal value) (A) to the total numbers of reference pixels (N) used for assessing accuracy (Shao et al., 2016) in training site (Eq. 9). Kappa coefficient (Kc) was estimated following equation (Eq. 10) the difference between correctly classified reference pixels and randomly classified reference pixels in comparing reference data.

$\:\text{O}\text{A}=\left(\frac{\text{A}}{\text{N}}\right)$ *100 (9)

$$\:\text{K}\text{c}=\frac{\text{N}{\sum\:}_{\text{i}=1}^{\text{n}}\text{X}\text{i}\text{i}-{\sum\:}_{\text{i}=1}^{\text{n}}\left({\text{X}}_{{+\text{i}}^{\:\text{*}\:}}{\text{X}}_{\text{i}+}\right)}{{\text{N}}^{2}-{\sum\:}_{\text{i}=1}^{\text{n}}\left({\text{X}}_{{+\text{i}}^{\:\text{*}\:}}{\text{X}}_{\text{i}+}\right)}$$

Where; K_c represents Kappa coefficient, n represent number of row in the matrix, A represents a correctly fixed number of pixels, N represents the total number of reference pixels, x_ii is the number of pixels observed in row and column, x_+ i and x_i+ are marginal value of column and row, respectively.

Table 2

Description of LU/LC classes in the study area
LU/LC classes	Description
Cultivated land	Area of land used for Rainfed or/and irrigation crop production and fallow lands (Abiye et al., 2023; Yirgu, 2022)
Grassland/grazing	Land use which covered with diverse grasses uses for communal/individual and help for feeds animals (Aneseyee, 2020; Ogato et al., 2021)
Agroforestry land	A unit of land use in which trees/shrubs are integrated with crops/animals (e.g. Home garden, coffee farm, woodlot, and Khat) (Endale et al., 2024)
Settlement	Land use which occupied by built up (e.g. urban, schools, religions, health centers and rural homesteads (Abiye et al., 2023; Yirgu, 2022)
Shrubs/bush land	Land use which dominated by short trees, shrubs, poorly dispersedly remained forestry and a variety of grasses (Aneseyee, 2020; Degife et al., 2021; Yirgu, 2022)

Table 3

LULC change for 1988, 2004 and 2021 years in both micro watersheds
Micro Watershed	LU/LC type	Year						LU/LC (1988–2004)		LU/LC (2004–2021)		LU/LC (1988–2021)
		1988		2004		2021		LU/LC (1988–2004)		LU/LC (2004–2021)		LU/LC (1988–2021)
		(Ha)	(%)	(Ha)	(%)	(Ha)	(%)	(Ha)	(%)	(Ha)	(%)	(Ha)	(%)
Nadi	Grassland	2485.6	21.5	2140.3	18.5	1506.9	13.0	-345.3	-3.0	-633.5	-5.5	-978.7	-8.5
	Settlement	1249.2	10.8	1351.7	11.7	1706.7	14.8	+ 102.5	+ 0.9	+ 355.0	+ 3.1	+ 457.5	+ 4.0
	Cultivated land	3014.1	26.1	3365.0	29.1	3806.0	32.9	+ 350.8	+ 3.0	+ 441.0	+ 3.8	+ 791.8	+ 6.8
	Shrubs/Bush land	2863.4	24.8	2294.4	19.9	1630.4	14.1	-569.0	-4.9	-664.0	-5.8	-1233.0	-10.7
	Agro forestry	1942.7	16.8	2403.6	20.8	2905.1	25.1	+ 460.9	+ 4.0	+ 501.5	+ 4.3	+ 962.4	+ 8.3
	Total	11555	100.0	11555	100.0	11555	100.0
Yedi	Grassland	1837.00	20.51	1582.00	17.67	1390.30	15.53	-255.0	-2.9	-191.7	-2.1	-446.7	-5.0
	Settlement	1171.50	13.08	1225.40	13.68	1365.80	15.25	+ 53.9	+ 0.6	+ 140.4	+ 1.6	+ 194.3	+ 2.2
	Cultivated land	2175.80	24.30	2391.00	26.70	2620.60	29.26	+ 215.2	+ 2.4	+ 229.6	+ 2.6	+ 444.8	+ 5.0
	Shrubs/Bush land	2069.8	23.1	2050.1	22.9	1106.7	12.4	-19.7	-0.2	-943.4	-10.5	-963.1	-10.7
	Agro forestry	1700.9	19.0	1706.5	19.1	2471.6	27.6	+ 5.6	+ 0.1	+ 765.1	+ 8.5	+ 770.7	+ 8.6
	Total	8955	100	8955	100	8955	100

3.1 Land use land cover change

Trends of the land use land cover change for 1988, 2004, and 2021 were presented (Table 3 and Fig. 3). A hybrid algorithms method was applied for classifying LU/LC into five classes both in micro watersheds. Among, LU/LC classifications settlement, cultivated land, and agroforestry continuously increase in the cost of grassland and shrubs/bush land in both micro watersheds. Agroforestry and cultivated land were increased by 4.3% (501.5ha) and 3.8% (441ha) in Nadi whereas by 4.7% (421.5ha) and 2.6% (229.6ha) in Yedi micro watershed from 2004 to 2021 years, respectively. Cultivated land was the dominant change followed by agroforestry land use for 2004 and 2021 in both micro watersheds. This agrees with Temesgen et al. (2021); Desalegn et al. (2023); Meragiaw et al. (2022), who found that agroforestry land use continues increasing relative to other land use. Similarly, substantial scholars confirmed that cultivated land was high increased than other land use in varied parts of the country (Derebe et al., 2022; Habte et al., 2021; Moisa et al., 2022). Besides, Tewabe & Fentahun (2020) also confirmed that within 32 years, agriculture and settlement land was increased by 15.61% and 8.05% in Lake Tana Basin respectively. Many land use/land cover change was observed between 1988 and 2021, but relatively large size changes detected between 2004 and 2021. This could be due to three factors; one, at the beginning commissioning of Gilgel Gibe I hydropower, displaced the farmers to other, this may be reinforced them to use other land uses; two, increased the demand for agroforestry and cultivated land as a result of population pressure and finally, most farmers shifted into farming followed to project commissioned since 2004. In contrast, shrubs/bush and grazing land use continuously declined from 1988 to 2021 in both micro watersheds. This suggests that the farmer were mostly used their land for multipurpose and crop production to feed their family and lower the quantity of cattle to be produced. This agrees with Tewabe & Fentahun (2020), who reported cultivated and settlement increases at a cost of grassing, shrubs/bush and forest land.

Land use land cover classification overall accuracy was presented (Table 4). As the errors matrix result shows that the overall accuracy classified land use map was (92.4%; 91%), (92.4%; 84%) and (91.2%; 92.5%) for 1988, 2004, and 2021 in Nadi and Yedi micro watersheds, respectively. Besides, its Kappa coefficient (Kc) was (.90; .89), (.90; .83), and (.89; .90) for 1988, 2004, and 2021 in Nadi and Yedi micro watersheds, respectively. This result indicates that the classified LU/LC image had strong agreement and accuracy as earlier recommended (Monserud & Leemans, 1992; Viera & Garrett, 2005).

3.2 Rainfall erosivity (R factor)

A rainfall is a major influencing factor of soil erosion. In this study, an average annual rainfall of 33 years (1988–2021) ranged from 1403.04 to 2395.71, high mean rainfalls was registered around Jimma and Limu Kosa while low rainfall was registered around Gibe farm and Nonno Kondala stations (Fig. 4). R-value varied from 884 to 949 MJmm/ha/hr/yr in Nadi (Fig. 5a) and from 796 to 912 MJmm/ha/hr/yr in Yedi micro watersheds (Fig. 5b). The erosivity value decreased from west northern to south eastern part of Nadi while western to eastern part of Yedi micro watersheds, an area known in high rainfall located at high topography feature. Similarly, Kidane et al. (2019) confirmed the presence of high erosivity value at high topography than downstream in Guder sub-watershed. The estimated R-value of this study was much less than the global mean value that is 2000 MJmm/hr/ha/yr (Borrelli et al., 2017) and 1179.4 MJmm/ha/hr/yr Kidane et al. (2019) and greater than 690 MJmm/ha/hr/yr (Girma & Gebre, 2020). The spatial distribution of rainfall varied within and between two watersheds.

3.3 Soil erodibility (K factor)

Soil type and K-factor of both micro watersheds were displayed (Fig. 6). Predominantly Eutric Nitosols is available in the area, with a K-factor value of 0.028. This suggests that the soil vulnerability to erosion is too low as a result of good drainage, low surface run-off, and high buildup of sandy base cations, high water infiltration rate, good nutrient recycling, soil organic matter, and low silt concentrations. Low silt content deactivates the erodibility of soil to erosion (Mhangara et al., 2012).

Table 4

LU/LC classification accuracy for 1988, 2004 and 2021 in both micro watersheds
Micro Watershed	LU/LC type	1988		2004		2021
Micro Watershed	LU/LC type	Producer's	User's	Producer's	User's	Producer's	User's
Nadi	Grassland	0.929	0.963	0.886	0.886	0.841	0.881
	Settlement	0.886	0.912	0.900	0.947	0.911	0.911
	Cultivated land	0.957	0.931	0.967	0.967	0.968	0.923
	Shrubs/Bush land	0.920	0.920	0.882	0.882	0.837	0.878
	Agro forestry	0.897	0.875	0.964	0.930	0.964	0.947
	Overall (%)	92.4		92.40		91.2
	Kapp	0.90		0.90		0.89
Yedi	Grassland	0.951	0.907	0.871	0.900	0.929	0.853
	Settlement	0.833	0.926	0.931	0.964	0.942	0.951
	Cultivated land	0.958	0.920	0.935	0.956	0.955	0.942
	Shrubs/Bush land	0.911	0.911	0.900	0.857	0.897	0.897
	Agro forestry	0.861	0.886	0.975	0.951	0.955	0.955
	Overall (%)	91.00		86.00		92.50
	Kapp	0.88		0.83		0.90

3.4 Topographic factor (LS_factor)

The result of a topographic factor was displayed (Fig. 7). Surface run-off is a primary input in soil erosion that is determined by the slope length and gradient. LS-factor varied from 0 to 261 with a 6.25 mean in the Nadi (Fig. 7a) whereas ranged from 0 to 210 with a 3.55 mean in Yedi (Fig. 7c) micro watersheds. Both slope length and steepness played a significant role in influencing the rate of soil erosion by water and its value was high in the western part of Nadi (Fig. 7b), as well as from the middle to north and south-east of Yedi (Fig. 7d) micro watersheds.

3.5 Cover management factor (C factor)

Analysis results of C-factor and its generated map were depicted (Table 5 and Fig. 8). It is strongly correlated with land cover and a dimensionless parameter with values ranging between 0 (well covered) and 1 (less covered). In this study, C-value was ranged from 0.01 to 0.25; 0.01 for bush/shrub, agroforestry, and grassland, 0.1 for settlement, and 0.25 for cultivated land. Except agroforestry, all value for each land use cover was assigned based on previous literature (Asmamaw et al., 2011; Bekele, 2021; Hurni, 1985) (Table 5).

Table 5

LU/LC and their C- factor for both micro watersheds
Watershed	LU/LC	C-factor	Sources
Nadi/Yedi	Grassland	0.01	(Bekele, 2021; Hurni, 1985)
	Settlement	0.1	(Hurni, 1985)
	Cultivated land	0.25	(Hurni, 1985)
	Shrubs/Bush land	0.01	(Asmamaw et al., 2011; Yirgu, 2022)
	Agroforestry	0.01

Table 6

Conservation practices factor (P) value
Land use type	Slope (%)	P-Value
Agricultural land	0–5	0.1
>>	5–10	0.12
>>	10–20	0.14
>>	20–30	0.19
>>	30–50	0.25
>>	> 50	0.33
Non-agricultural	All	1

3.6 Conservation practice factor (P factor)

The result of P- factor was presented (Table 6 and Fig. 9). As ArcGIS analysis tools show that the P-factor value ranged from 0.1 to 0.33 in Nadi and 0.1 to 0.0.33 in Yedi micro watersheds for agriculture based on slope reclassification and 1 for non-agricultural land. It was assigned as recommended (Wischmeier & Smith, 1978).

3.7 Annual soil loss estimate

Estimated soil loss is depicted (Table 7 and Fig. 10). Annual soil loss varied in each micro watershed per year; its ranged from 0 to 339.2 with mean value of 64 ton/ha/year in 1988, from 0 to 343 ton/ha/year with mean value of 66.71 in 2004 and from 0 to 345.7 with mean value of 68.10 ton/ha/year in 2021 in Nadi micro watershed. Similarly, in Yedi micro watershed, an annual soil loss varied from 0 to 201.98 with mean of 46.43 ton/ha/year in 1988; from 0 to 207.55 with a mean of 48.51 in 2004 and from 0 to 198.17 with mean value of 45.88 ton/ha/year in 2021. This result was less than half of the annual soil loss of 867.3, 905.7 and 958.2 ton/ha/year in 1986, 2003, and 2020 in Muger sub-basin (Teshome et al., 2022). In the same basin of this study (Omo-Gibe) but differ in sub-watershed, Girma & Gebre (2020) estimated total annual soil loss ranged from 0 to 279 tons/ha/year. Lencha & Moges (2015) reported 1.6 to 232.4 ton/ha/year, that is less than Nadi’s and relatively greater than Yedi’s value. In terms of mean annual soil loss, Nadi’s results were relatively less than the previous mean soil loss of 69 ton/ha/year from Omo-Gibe (Girma & Gebre, 2020) and 75.85 and 107.07 ton/ha/year in 2000 and 2018 in Erer Sub-Basin, respectively (Weldu & Harka, 2020) and greater than the mean value of 62.98 ton/ha/year central part of Gilgel Gibe I catchment (Tesfaye & Tibebe, 2018); 51.04 ton/ha/year of eastern part of the country (Woldemariam et al., 2018); 53.2, 63.6, and 64 ton/ha/year in 1986, 2003 and 2020 of Muger basin of upper Blue Nile basin, respectively (Teshome et al., 2022). Again, Yedi’s value was less than the above mentioned results and greater than the mean value of 13.2 ton/ha/year of Beles of Amhara region (Kebede et al., 2021), 37 ton/ha/year of Gelada watershed (Yesuph & Dagnew, 2019); 38.7 of Chereti watershed Amhara region (Negese et al., 2021) and 31 ton/ha/year of Welmel watershed of Oromia region (Tessema et al., 2020). Mean annual soil loss varied between Nadi and Yedi micro watersheds. This may happen due to a variation of parameters, landscape, watershed size, LU/LC and land management.

Mean soil loss value registered was continuously increasing from 1988 to 2021 in Nadi. This agrees with (Teshome et al., 2022), who reported a continuous enlarge soil loss mean from 1986 to 2020. Besides, Lencha & Moges (2015) also confirmed increments of mean soil loss from 48.1 in 2001 to 60.9 ton/ha/year in 2013. In contrast, Woldemariam et al. (2018) found a rapidly declination of soil loss from 51.04 in 2000 to 34.26 in 2016 from east Hararghe zone of Oromia. This may shows there is a continuous intervention of land management. High mean soil loss was registered between 1988 and 2004 in both micro watersheds when slightly increased in Nadi while swiftly decreased in Yedi between 2004 and 2021. This discrepancy may happen due to two reasons: the nature of land features (hilly and rugged landscape) and anthropogenic pressures that accelerate land use changes and land fragmentation due to the displacement of farmers between 1988 and 2004. Furthermore, LU/LC was increased between 2004 and 2021 than between 1988 and 2004; this may be due to the encroachment of farmers into the buffer zone coupled with the utilization of marginal land after a Gibe hydroelectric power commissioned. In contrast, land management measures were undertaken by the government integrated with others (NGOs and community after 2004 and further strengthen officially since 2011 in Yedi while alone by interest individual farmers in Nadi micro watershed. As a result soil loss has declined in Yedi micro watershed in second period. Moreover, varied agroforestry practiced Home garden, Coffee based, Boundary planting, Hedgerow, Woodlot and Khat based AF were relatively high in Yedi than Nadi micro watersheds (Endale et al., 2024).

Table 7

Soil loss (ton/ha/year) for 1988, 2004, and 2021 in both micro watersheds
Micro watershed	Year
Micro watershed	1988			2004			2021
	Min	Max	Mean	Min	Max	Mean	Min	Max	Mean
Nadi	0	336.20	64.01	0	343.00	66.71	0	345.70	68.10
Yedi	0	201.98	46.43	0	207.55	48.51	0	198.17	45.88

3.8 Soil erosion severity area classes

In this study, soil loss area was divided into five classes, namely very slight (0–5), slight (5–15), moderate (15–30), severe (30–50), and very severe (> 50) (Table 8). Its classification was accomplished following the previous studies (Kidane et al., 2019; Moisa et al., 2022; Teshome et al., 2022). You have to check a similarity of study area condition. Accordingly, about 4044.25 ha (35%) in 1988, 3582.05 ha (31%) in 2004, and 3697.6 ha (32%) in 2021 of land were classified under very slightly, slightly, and moderately soil erosion risk area in Nadi micro watershed, respectively. Similarly, about 3364.9 ha (38%), 2302 ha (26%), and 2833.6 ha (32%) were categorized under very slightly, slightly, and moderately soil erosion area in 1988, 2004, and 2021 Yedi micro watershed, respectively. From the entire area, relatively small land size of (9%; 10%), (11%; 12%) and (12%; 11%) was affected by very severe erosion risks in 1988, 2004, and 2021 in Nadi and Yedi micro watersheds, respectively. This result is much smaller than a previously examined Teshome et al. (2022), where 33, 38.1, and 38.9% of the total area categorized under very severe erosion risk in 1986, 2003, and 2020, respectively. Area of land exposed to soil erosion risk (at moderate, severe, and very severe classes) was increased from 1988 to 2004 whereas decreasing from 2004 to 2021 in Yedi, whereas almost inverse in Nadi micro watershed. This is due to the introduction of various land management methods and LU/LC shifted to agroforestry-based land use intensively after 2004 in Yedi. Besides, Endale et al. (2024) ratified that in Yedi micro watershed high smallholder farmer adopted various agroforestry practices than Nadi micro watersheds.

Table 8

Severity classes and soil loss area for 1988, 2004, and 2021 in both micro watersheds
Micro watershed	Severity classes	Soil erosion rate (ton/ha/year)	1988		2004		2021
Micro watershed	Severity classes	Soil erosion rate (ton/ha/year)	Area (km²)	%	Area (km²)	%	Area (km²)	%
Nadi	Very slight	0–5	4044.25	35	3235.4	28	2657.65	23
	Slight	5–15	3350.95	29	3582.05	31	3697.6	32
	Moderate	15–30	1848.8	16	1964.35	17	2195.45	19
	Severe	30–50	1271.05	11	1502.15	13	1617.7	14
	Very severe	> 50	1039.95	9	1271.05	11	1386.6	12
	Total		11555	100	11555	100	11555	100
Yedi	Very slight	0–5	3364.90	38	2302.30	26	2833.60	32
	Slight	5–15	2036.65	23	2125.20	24	2080.93	23.5
	Moderate	15–30	1505.35	17	2036.65	23	1771.00	20
	Severe	30–50	1062.60	12	1328.25	15	1195.43	13.5
	Very severe	> 50	885.50	10	1062.60	12	974.05	11
	Total		8855	100	8855	100	8855	100

3.9 LU/LC effects on soil erosion

Mean annual soil loss was computed across LU/LC for 1988, 2004, and 2021 in both watersheds (Table 9). The average annual soil loss registered under cultivated land was high followed by grassland in 1988, 2004, and 2021 in both micro watersheds. As a result, the anticipated mean soil loss in the Nadi micro watershed increased by 1.81 ton/ha/year from 1988 to 2004 and by 0.59 ton/ha/year from 2004 to 2021. Such that it was significant among LU/LC (F = 191.18; P = 0.00). Similarly, mean soil loss in the Yedi micro watershed increased by 2.22 ton/ha/year from 1988 to 2004 and decreased by 1.98 ton/ha/year between 2004 and 2021. The rate of soil loss continually rose from 1988 to 2004 to 2021 in Nadi whereas inconsistence in Yedi micro watershed specifically decreased between 2004 and 2021. This suggests a variance in land management practices, expansion of crop production and soil conservation intervention between two micro watersheds. Teshome et al. (2022) confirmed increasing of mean soil loss from cultivated land between 1986 and 2020. Besides, Borrelli et al. (2013) also deep-rooted the impacts of land use change on soil loss, that soil loss increased from cultivated between 2001 and 2012. In other views, among LU/LC identified, agroforestry plays a positive role in reducing soil erosion and the lowest mean soil loss was computed between 1988 and 2021 in both watersheds. Thus, cultivated and agroforestry land were the most increased land from 1988 to 2021 in line with this, cultivated land aggravates soil erosion and loss, whereas agroforestry was reverse. In both micro watersheds, as multiple comparisons indicated that except between shrubs/bush and grass estimated soil loss under others LU/LC was significant to each other.

Table 9

Land use change effect on soil loss in the study area
Micro watershed	Land use type	1988		2004		2021
Micro watershed	Land use type	MSL (ton/ha/year)	%	MSL (ton/ha/year)	%	MSL (ton/ha/year)	%
Nadi	Grassland	14.08	22	14.68	22	15.66	23
	Settlement	6.40	10	7.34	11	7.49	11
	Cultivated land	26.88	42	28.69	43	29.28	43
	Shrubs/Bush land	12.16	19	13.34	210	13.62	20
	Agro forestry	4.48	7	2.67	4	2.04	3
Yedi	Grassland	9.75	21	11.39	23	10.09	22
	Settlement	5.57	12	6.44	13	5.51	12
	Cultivated land	18.57	40	20.79	42	18.81	41
	Shrubs/Bush land	9.29	20	8.91	18	8.72	19
	Agro forestry	3.25	7	1.98	4	1.84	4

Though it dates back thousands of years, agroforestry is today widely used as an alternate method of land use. However, its contributions, particularly on soil loss reduction, were ignored in many LU/LC investigations. Thus, the impact LU/LC comprising ALUC on soil loss was examined in Nadi and Yedi micro watersheds of Gilgel Gibe I catchment. Nadi and Yedi micro watersheds were purposively selected. By primary and secondary data, five parameters, such as rainfall data of 33 years (1988–2021) from Ethiopia Metrological Agency, soil from Ministry of Water and Energy, data for LU/LC of 1988, 2004, and 2021 downloaded from the United States Geological Survey using Landsat images (TM, ETM+, and OLI/TIRS) and Digital elevation model were used. The data was analyzed using RUSLE model in ArcGIS. As results, cultivated land was increased between 1988 and 2021, followed by agroforestry and settlement in cost of grassing and shrubs/bush land in both micro watersheds. An average soil loss was (64; 46.43), (66.71; 48.51) and (68.10; 45.88) ton/ha/year in 1988, 2004 and 2021 in Nadi and Yedi micro watersheds, respectively. Result discovered that about (9%; 10%), (11%; 12%) and (12%; 11%) of land were affected by very severe soil erosion risks in 1988, 2004, and 2021 from Nadi and Yedi micro watersheds, respectively. Among LU/LC, ALUC was found to be exceedingly effective in reducing soil loss when cultivated land was increasing soil loss in both watersheds. To constantly mitigate soil erosion, identifying land uses that are more vulnerable to soil erosion and suggesting the optimal land management plan. Therefore, this finding recommend that increasing adopting agroforestry land use practices to sustainably restore degraded land and limit the rate of soil loss in the study area.

Ethical approval

All authors have read, understood, and have complied as applicable with the statement on “Ethical responsibilities of Authors” as found in the Instructions for Authors.

Competing interests

The authors declare no competing interests.

Author Contribution

Statement of Authors Endale Bekele has significantly contributed to the study's conception and design, data collection, analysis, and interpretation.Habtamu Temesgen had played a major role in providing concrete comments from title modification to manuscript completion, and the revision, all of which helped to bridge the gap in this work's quality at this stage.Girma Abera had played a major role in providing concrete comments from title modification to manuscript completion, and the revision, all of which helped to bridge the gap in this work's quality at this stage.

Acknowledgments

The authors are grateful to Dilla University's College of Agricultural and Natural Resources for providing an appropriate environment for learning and conducting this research by arranging a different presentation schedule. Bonga University is also highly regarded by the correspondent author for providing this learning opportunity and for sponsoring him. The authors also thank Dr. Habtamu Temesgen and Dr. Girma Abera, who provided advice and tireless assistance in improving methodology, the redesign study site plan, and the article revision, all of which helped to bridge the gap in this work's quality at this stage.

Data availability

All data that support the findings of this study are referenced in the article.

Abiye, W., Waltner, I., & Kindie, H. (2023). Spatiotemporal dynamics of soil erosion response to land use land cover dynamics and climate variability in Maybar watershed , Awash basin , Ethiopia. Geology, Ecology, and Landscapes, 00(00), 1–23. https://doi.org/10.1080/24749508.2023.2256542
Adugna, T. M., & Cherie, D. A. (2021). A Review on Reservoirs Sedimenation Problems in Ethiopia. Asian Journal of Advanced Research and Reports, 1–8. https://doi.org/10.9734/ajarr/2021/v15i330372
Alem, B. B. (2022). The nexus between land use land cover dynamics and soil erosion hotspot area of Girana Watershed , Awash Basin , Ethiopia. Heliyon, 8(February), e08916. https://doi.org/10.1016/j.heliyon.2022.e08916
Aneseyee, A. B. (2020). Land use / land cover change effect on Soil erosion and Sediment Delivery on Winike Watershed , Omo Gibe Basin , Ethiopia. https://doi.org/https://doi.org/10.1016/j.scitotenv.2020.138776.
Aneseyee, A. B., Soromessa, T., & Elias, E. (2019). The effect of land use/land cover changes on ecosystem services valuation of Winike watershed, Omo Gibe basin, Ethiopia. HUMAN AND ECOLOGICAL RISK ASSESSMENT: AN INTERNATIONAL JOURNAL, 26(10), 2608–2627. https://doi.org/10.1080/10807039.2019.1675139
Asmamaw, L. B., Mohammed, A. A., & Lulseged, T. D. (2011). Land use/cover dynamics and their effects in the Gerado catchment, northeastern Ethiopia. International Journal of Environmental Studies, 68(6), 883–900.
Assefa, T. T., Jha, M. K., Tilahun, S. A., Yetbarek, E., Adem, A. A., & Wale, A. (2015). Identification of erosion hotspot area using GIS and MCE technique for koga watershed in the upper blue Nile Basin, Ethiopia. American Journal of Environmental Sciences, 11(4), 245.
Bekele, M. (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), 9–27. https://doi.org/10.15406/ijh.2021.05.00260
Belay, T., & Mengistu, D. A. (2021). Impacts of land use / land cover and climate changes on soil erosion in Muga watershed , Upper Blue Nile basin ( Abay ), Ethiopia. Ecological Processes. https://doi.org/10.1186/s13717-021-00339-9
Beyene, A. D., Mekonnen, A., Randall, B., & Deribe, R. (2019). Household Level Determinants of Agroforestry Practices Adoption in Rural Ethiopia. Forests Trees and Livelihoods, 28(3), 194–213. https://doi.org/10.1080/14728028.2019.1620137
Borrelli, P., Robinson, D. A., Fleischer, L. R., Lugato, E., Ballabio, C., Alewell, C., Meusburger, K., Modugno, S., Schütt, B., Ferro, V., Bagarello, V., Oost, K. Van, Montanarella, L., & Panagos, P. (2017). An assessment of the global impact of 21st century land use change on soil erosion. Nature Communications, 8(2013). https://doi.org/10.1038/s41467-017-02142-7
Broothaerts, N., Kissi, E., Poesen, J., Van Rompaey, A., Getahun, K., Van Ranst, E., & Diels, J. (2012). Spatial patterns, causes and consequences of landslides in the Gilgel Gibe catchment, SW Ethiopia. Catena, 97, 127–136. https://doi.org/10.1016/j.catena.2012.05.011
Congalton, R. G., & Green, K. (2019). Assessing the accuracy of remotely sensed data: principles and practices. CRC press.
De Silva, R. P., Dayawansa, N. D. K., & Ratnasiri, M. D. (2007). A comparison of methods used in estimating missing rainfall data. Journal of Agricultural Sciences, 3(2), 101. https://doi.org/10.4038/jas.v3i2.8107
Degife, A., Worku, H., & Gizaw, S. (2021). Environmental implications of soil erosion and sediment yield in Lake Hawassa watershed, south-central Ethiopia. Environmental Systems Research, 10(28). https://doi.org/10.1186/s40068-021-00232-6
Demissie, T. A., Saathoff, F., Seleshi, Y., & Gebissa, A. (2013). Evaluating the Effectiveness of Best Management Practices in Gilgel Gibe Basin Watershed-Ethiopia. In Journal of Civil Engineering and Architecture (Vol. 7, Issue 10).
Derebe, M. A., Hatiye, S. D., & Asres, L. A. (2022). Dynamics and Prediction of Land Use and Land Cover Changes Using Geospatial Techniques in Abelti Watershed , Omo Gibe River. 2022. https://doi.org/10.1155/2022/1862461
Desalegn, M. Y., Miheretu, B. A., & Gobezie, T. (2023). Impact of land use / land cover changes on soil erosion risk in upper Mile River sub-watershed , North Eastern highlands of Ethiopia. Geology, Ecology, and Landscapes, 7(4), 1–13. https://doi.org/10.1080/24749508.2023.2206063
Dibaba, W. T., Demissie, T. A., & Miegel, K. (2020). Drivers and Implications of Land Use/Land Cover Dynamics in Finchaa Catchment, Northwestern Ethiopia. Land, 9(113), 1–20.
Endalamaw, N. T., Moges, M. A., Kebede, Y. S., Alehegn, B. M., & Sinshaw, B. G. (2021). Potential soil loss estimation for conservation planning, upper Blue Nile Basin, Ethiopia. Environmental Challenges, 5, 100224. https://doi.org/10.1016/j.envc.2021.100224
Endale, B., Girma, A., & Habtamu, T. (2024). Factors influencing adoption and intensity of agroforestry systems for mitigating land degradation (MLD) in Gilgel Gibe I catchment, southwestern Ethiopia. Cogent Food and Agriculture, 10(1), 1–18. https://doi.org/10.1080/23311932.2024.2380782
Estifanos, H. ., & Gebremariam, B. (2019). Modeling-impact of Land Use/Cover Change on Sediment Yield (Case Study on Omo-gibe Basin, Gilgel Gibe III Watershed, Ethiopia). American Journal of Modern Energy, 5(6), 84–93. https://doi.org/10.11648/j.ajme.20190506.11
Gelagay, H. S., & Minale, A. S. (2016). Soil loss estimation using GIS and Remote sensing techniques: A case of Koga watershed, Northwestern Ethiopia. International Soil and Water Conservation Research, 4(2), 126–136. https://doi.org/10.1016/j.iswcr.2016.01.002
Getachew, B., Manjunatha, B. R., & Bhat, G. H. (2021). Assessing current and projected soil loss under changing land use and climate using RUSLE with Remote sensing and GIS in the Lake Tana Basin, Upper Blue Nile River Basin, Ethiopia. Egyptian Journal of Remote Sensing and Space Science, 24(3), 907–918. https://doi.org/10.1016/j.ejrs.2021.10.001
Getu, L. A., Nagy, A., & Addis, H. K. (2022). Soil loss estimation and severity mapping using the RUSLE model and GIS in Megech watershed, Ethiopia. Environmental Challenges, 8(April). https://doi.org/10.1016/j.envc.2022.100560
Ghosal, K., & Bhattacharya, S. Das. (2020). A Review of RUSLE Model. Journal of the Indian Society of Remote Sensing, 48(4), 689–707.
Girma, R., & Gebre, E. (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 Systems Research, 9(19). https://doi.org/10.1186/s40068-020-00180-7
Girmay, G., Moges, A., & Muluneh, A. (2020). Estimation of soil loss rate using the USLE model for Agewmariayam Watershed, northern Ethiopia. Agriculture and Food Security, 9(1). https://doi.org/10.1186/s40066-020-00262-w
Gitima, G., Teshome, M., Kassie, M., & Jakubus, M. (2022). Heliyon Spatiotemporal land use and cover changes across agroecologies and slope gradients using geospatial technologies in Zoa watershed , Southwest Ethiopia. Heliyon, 8(June), e10696. https://doi.org/10.1016/j.heliyon.2022.e10696
Gong, W., Liu, T., Duan, X., Sun, Y., Zhang, Y., Tong, X., & Qiu, Z. (2022). Estimating the Soil Erosion Response to Land-Use Land-Cover. Water, 14(742), 1–17.
Guadie, M., Molla, E., Mekonnen, M., & Cerdà, A. (2020). Effects of soil bund and stone-faced soil bund on soil physicochemical properties and crop yield under rain-fed conditions of Northwest Ethiopia. Land, 9(1), 13.
Habte, D. G., Belliethathan, S., & Ayenew, T. (2021). Evaluation of the status of land use /land cover change using remote sensing and GIS in Jewha Watershed , Northeastern Ethiopia. SN Applied Sciences, 3(501), 1–10. https://doi.org/10.1007/s42452-021-04498-4
Hassan, A. R., Imamul, M., & Bhuiyan, H. (2016). A decision support system for automatic sleep staging from EEG signals using tunable Q-factor wavelet transform and spectral features. Journal of Neuroscience Methods, 271, 107–118. https://doi.org/10.1016/j.jneumeth.2016.07.012
Hassen, G., Bantider, A., Legesse, A., & Maimbo, M. (2021). Assessment of Design and Constraints of Physical Soil and Water Conservation Structures in respect to the standard in the case of Gidabo sub-basin, Ethiopia. Cogent Food and Agriculture, 7(1). https://doi.org/10.1080/23311932.2020.1855818
Hurni, H. (1985). Erosion-productivity-conservation systems in Ethiopia.
Jiru, E. B., & Wegari, H. T. (2022). Soil and water conservation practice effects on soil physicochemical properties and crop yield in Ethiopia : review and synthesis. Ecological Processes. https://doi.org/10.1186/s13717-022-00364-2
Jiru, E. B., Zerihun, K., & Ermias, M. (2020). Analysis of socio-economic contribution of agroforestry systems to smallholder farmers around Jimma town, Southwestern Ethiopia. International Journal of Biodiversity and Conservation, 12(1), 48–58. https://doi.org/10.5897/ijbc2018.1237
Kebede, Y. S., Endalamaw, N. T., Sinshaw, B. G., & Atinkut, H. B. (2021). Modeling soil erosion using RUSLE and GIS at watershed level in the upper beles, Ethiopia. Environmental Challenges, 2. https://doi.org/10.1016/j.envc.2020.100009
Kidane, M., Bezie, A., Kesete, N., & Tolessa, T. (2019). The impact of land use and land cover (LULC) dynamics on soil erosion and sediment yield in Ethiopia. Heliyon, 5, e02981. https://doi.org/10.1016/j.heliyon.2019.e02981
Lencha, B., & Moges, A. (2015). Identification of Soil Erosion Hotspots in Jimma Zone (Ethiopia) Using GIS Based Approach. Ethiopian Journal of Environmental Studies and Management, 8(2), 926. https://doi.org/10.4314/ejesm.v8i2.7s
Lillesand, T., Kiefer, R. W., & Chipman, J. (2015). Remote sensing and image interpretation. John Wiley & Sons.
Masha, M., Yirgu, T., & Debele, M. (2021). Impacts of Land Cover and Greenness Change on Soil Loss and Erosion Risk in Damota Area Districts , Southern Ethiopia. 2021.
Mbow, C., Smith, P., Skole, D., Duguma, L., & Bustamante, M. (2014). Achieving mitigation and adaptation to climate change through sustainable agroforestry practices in africa. Current Opinion in Environmental Sustainability, 6(1), 8–14. https://doi.org/10.1016/j.cosust.2013.09.002
Mengie, A. M., Hagos, Y. G., Malede, D. A., & Andualem, T. G. (2022). Assessment of soil loss rate using GIS – RUSLE interface in Tashat Watershed , Northwestern Ethiopia. Journal of Sedimentary Environments, 7(3), 617–631. https://doi.org/10.1007/s43217-022-00112-8
Meragiaw, M., Woldu, Z., & Singh, B. R. (2022). Land use and land cover dynamics and traditional agroforestry practices in Wonchi District, Ethiopia. PeerJ, 10(:e12898), 1–26. https://doi.org/10.7717/peerj.12898
Mhangara, P., Kakembo, V., & Lim, K. J. (2012). Soil erosion risk assessment of the Keiskamma catchment , South Africa using GIS and remote sensing. 2087–2102. https://doi.org/10.1007/s12665-011-1190-x
Moisa, M. B., Babu, A., & Getahun, K. (2022). Integration of geospatial technologies with RUSLE model for analysis of soil erosion in response to land use /land cover dynamics: a case of Jere watershed, Western Ethiopia. Sustainable Water Resources Management, 9(13), 21. https://doi.org/10.1007/s40899-022-00805-y
Monserud, R. A., & Leemans, R. (1992). Comparing global vegetation maps with the Kappa statistic. Ecological Modelling, 62(4), 275–293.
Moore, I. D., & Burch, G. J. (1986). Physical basis of the length‐slope factor in the universal soil loss equation. Soil Science Society of America Journal, 50(5), 1294–1298.
Moussa, A. M. A. (2019). Assessment of sediment deposition in Aswan high dam reservoir during 50 years (1964–2014). Grand Ethiopian Renaissance Dam Versus Aswan High Dam: A View from Egypt, 233–253.
Ndalama, E., Missanjo, E., & Kamanga-Thole, G. (2015). Agroforestry Contribution to the Improvement of Rural Community Livelihoods in Balaka , Malawi. International Journal of Forestry and Horticulture (IJFH), 1(1), 5–11.
Negese, A., Fekadu, E., & Getnet, H. (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. https://doi.org/10.1177/1178622120985814
Neitsch, S. L., Arnold, J. G., Kiniry, J. R. e a1, Srinivasan, R., & Williams, J. R. (2000). Soil and water assessment tool user’s manual. Blackland Research Center, Temple, TX.
Ogato, G. S., Bantider, A., & Geneletti, D. (2021). Dynamics of land use and land cover changes in Huluka watershed of Oromia Regional State , Ethiopia. Environmental Systems Research. https://doi.org/10.1186/s40068-021-00218-4
Panagos, P., Ballabio, C., Borrelli, P., Meusburger, K., Klik, A., Rousseva, S., Tadić, M. P., Michaelides, S., Hrabalíková, M., & Olsen, P. (2015). Rainfall erosivity in Europe. Science of the Total Environment, 511, 801–814.
Prasannakumar, V., Vijith, H., Abinod, S., & Geetha, N. (2012). Estimation of soil erosion risk within a small mountainous sub-watershed in Kerala, India, using Revised Universal Soil Loss Equation (RUSLE) and geo-information technology. Geoscience Frontiers, 3(2), 209–215. https://doi.org/10.1016/j.gsf.2011.11.003
Renard, K. G. (1997). Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). US Department of Agriculture, Agricultural Research Service.
Robert, P. S., & Hilborn, D. (2000). Factsheet: universal soil loss equation (USLE). Index No-572/751, Queen, s Printer for Ontario.
Shao, Z., Fu, H., Fu, P., & Yin, L. (2016). Mapping urban impervious surface by fusing optical and SAR data at the decision level. Remote Sensing, 8(11), 1–21. https://doi.org/10.3390/rs8110945
Simms, A. D., Woodroffe, C. D., & Jones, B. G. (2003). Application of RUSLE for erosion management in a coastal catchment, southern NSW. Https://Ro.Uow.Edu.Au/Scipapers/34.
Temesgen, H., Wu, W., Legesse, A., & Yirsaw, E. (2021). Modeling and prediction of effects of land use change in an agroforestry dominated southeastern Rift-Valley escarpment of Ethiopia. Remote Sensing Applications: Society and Environment, 21, 100469. https://doi.org/10.1016/j.rsase.2021.100469
Tesfaye, G., & Ameyu, T. (2021). Soil Erodibility Analysis and Mapping in Gilgel Gibe-I Catchment, Omo-Gibe River Basin, Ethiopia. Applied and Environmental Soil Science, 2021, 7. https://doi.org/10.1155/2021/8985783
Tesfaye, G., & Tibebe, D. (2018). Soil Erosion Modeling Using GIS Based RUSEL Model in Gilgel Gibe-1 Catchment, South West Ethiopia. Int J Environ Sci Nat Res., 15(5), 2–9. https://doi.org/10.19080/ijesnr.2018.15.555923
Teshome, D. S., Moisa, M. B., Gemeda, D. O., & You, S. (2022). Effect of Land Use-Land Cover Change on Soil Erosion and Sediment Yield in Muger Sub-Basin, Upper Blue Nile Basin, Ethiopia. Land, 11, 2173. https://doi.org/https://doi.org/10.3390/land11122173
Tessema, Y. M., Jasińska, J., Yadeta, L. T., Świtoniak, M., Puchałka, R., & Gebregeorgis, E. G. (2020). Soil loss estimation for conservation planning in the welmel watershed of the Genale Dawa Basin, Ethiopia. Agronomy, 10(6). https://doi.org/10.3390/agronomy10060777
Tewabe, D., & Fentahun, T. (2020). Assessing land use and land cover change detection using remote sensing in the Lake Tana Basin , Northwest Ethiopia Assessing land use and land cover change detection using remote sensing in the Lake Tana Basin , Northwest Ethiopia. Cogent Environmental Science, 6(1). https://doi.org/10.1080/23311843.2020.1778998
Tolla, S., Wasse, W., & Mulugeta, Y. (2016). Rural Household Vulnerability to Poverty in South West Ethiopia: The Case of Gilgel Gibe Hydraulic Dam Area of Sokoru and Tiro Afeta Woreda. Global Journal of Human-Social Sciences: E. Economics, 16(3), 44-54.
Viera, A. J., & Garrett, J. M. (2005). Anthony J. Viera, MD; Joanne M. Garrett, PhD (2005). Understanding interobserver agreement: the kappa statistic. Fam Med 2005;37(5):360-63. Family Medicine, 37(5), 360–363. http://www1.cs.columbia.edu/~julia/courses/CS6998/Interrater_agreement.Kappa_statistic.pdf
Wari, B. N., Feyssa, D. H., & Kebebew, Z. (2019). Assessment of woody species in agroforestry systems around Jimma Town , Southwestern Ethiopia. 11(January), 18–30. https://doi.org/10.5897/IJBC2018.1207
Weldu, G. W., & Harka, A. E. (2020). Eff ect of Land Use and Land Cover Change on Soil Erosion in Erer Sub-Basin, Northeast Wabi Shebelle Basin, Ethiopia. Land, 9(111), 1–25.
Williams, J. . (1975). Sediment-yield prediction with universal equation using runoff energy factor. In Present and Prospective Technology for Predicting Sediment Yield and Sources: Proceedings of the Sediment-Yield Workshop, USDA Sedimentation Laboratory, Oxford, Miss., Agricu. 40, 44.
Williams, J. . (1995). Chapter 25. The EPIC model. In computer models of watershed hydrology. Water Resources Publications, Highlands Ranch, pp 909–1000.
Wischmeier, W. ., & Smith, D. D. (1978). Predicting Rainfall Erosion Losses; A Guide to Corsevation Planning. U.S Department of Agriculture Handbook No.537.Washington D.C.
Woldemariam, G. W., Iguala, A. D., Tekalign, S., & Reddy, R. U. (2018). Spatial modeling of soil erosion risk and its implication for conservation planning: The case of the gobele watershed, east hararghe zone, ethiopia. Land, 7(1). https://doi.org/10.3390/LAND7010025
Wolka, K., Sterk, G., Biazin, B., & Negash, M. (2018). Benefits , limitations and sustainability of soil and water conservation structures in Omo-Gibe basin , Southwest Ethiopia. Land Use Policy, 73(2018), 1–10. https://doi.org/https://doi.org/10.1016/j.landusepol.2018.01.025
Yesuph, A. Y., & Dagnew, A. B. (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). https://doi.org/10.1186/s40068-019-0145-1
Yirgu, T. (2022). Assessment of soil erosion hazard and factors affecting farmers’ adoption of soil and water management measure: A case study from Upper Domba Watershed, Southern Ethiopia. Heliyon, 8(6), e09536. https://doi.org/10.1016/j.heliyon.2022.e09536
Zerihun, M., Mohammedyasin, M. S., Sewnet, D., Adem, A. A., & Lakew, M. (2018). Assessment of soil erosion using RUSLE, GIS and remote sensing in NW Ethiopia. Geoderma Regional, 12, 83–90. https://doi.org/10.1016/j.geodrs.2018.01.002

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Determine Soil Loss Response to Agroforestry Land Use Change (ALUC) Using RUSLE Based GIS and RS in Gilgel Gibe I Catchment, Southwestern, Ethiopia

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods and Materials

2.1 Study area description

2.2 Farming systems of the catchment

2.3 Sample design and selection

2.4 Data source and data analysis method

2.4.1 Data source

2.4.2 Data analysis method

2.4.2.1 RUSLE model

2.4.2.2 Erosivity factor (R)

2.4.2.3 Soil erodibility factor (K)

2.4.2.4 Slope length (L) and slope gradient (S) factor

2.4.2.5 Land use and cover management factor (C)

2.4.2.6 Conservation practice factor (P)

2.5 Land use classification and accuracy test

3. Results and discussions

3.1 Land use land cover change

3.2 Rainfall erosivity (R factor)

3.3 Soil erodibility (K factor)

3.4 Topographic factor (LS_factor)

3.5 Cover management factor (C factor)

3.6 Conservation practice factor (P factor)

3.7 Annual soil loss estimate

3.8 Soil erosion severity area classes

3.9 LU/LC effects on soil erosion

Conclusions

Declarations

Ethical approval

Author Contribution

Acknowledgments

Data availability

References

Additional Declarations

Status:

Version 1