Investigation of land use and land cover dynamics, drivers, and their consequences in south-central Ethiopia, with special emphasis on the contrasting agro-ecological settings of the Satame and Legabora watersheds

doi:10.21203/rs.3.rs-2962481/v1

Download PDF

Research Article

Investigation of land use and land cover dynamics, drivers, and their consequences in south-central Ethiopia, with special emphasis on the contrasting agro-ecological settings of the Satame and Legabora watersheds

https://doi.org/10.21203/rs.3.rs-2962481/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Identifying the causes and drivers of land change is one of the main objectives of the land use and cover change investigation. Land use and cover change data are crucial for managing water resources and land preservation. This study examined the trends, driving factors, and implications of land use and land cover dynamics in contrasting agro-ecologies to provide empirical evidence about increases and losses in land use classes. Landsat images of MSS (1976), TM (1991 and 2001), ETM+ (2011), and OLI (2022) were used to study LULC dynamics. To identify the potential driving forces and impacts, field observations, discussions with key informants, analysis of land use policy, and focus group discussions (FGDs) were undertaken. The analysis of remote sensing data revealed that there was an increasing tendency for croplands, settlements, and forest lands at the expense of grasslands and shrublands. In Satame, shrub land and grassland decreased annually by –128.7 and –139.9 ha per year, while grassland and shrub cover reduced annually by –248.2 and –144.5 ha per year in Legabora, respectively. The Local communities perceived agricultural expansion, population growth, settlement expansion, and poverty as the main drivers of LULC changes, while deforestation, land degradation, low production and loss of biodiversity were the major consequences. From this study, it can be inferred that LULC has undergone significant changes over the past 45 years, with both positive and negative impacts. Hence, this study provides a foundation for land management policies and strategies in this watershed and other watersheds facing similar problems.

Ethiopia

watersheds

LULC

Agro-ecologies

Drivers

Studies on LUCC are essential for global change research as they can reveal the effects of human activity on natural ecosystems (Wang et al., 2023). Data on changes in land use and land cover (LULC) is essential to understanding the relationship between people and the environment (Warner et al.). LULC changes are being prioritized globally, despite the fact that they have significant environmental effects at the local and regional levels (Foley et al., 2005). Anthropogenic modification of land has a significant impact on ecology (Garedew et al., 2009), and making it a top concern topic in many scientific research communities because of its impact on climate change (Elias et al., 2019), biodiversity loss (Phalan et al., 2011), biogeochemical processes (Houghton et al., 2012), social well-being (Dwivedi et al., 2007; Mengistu et al., 2012), hydrological cycles and groundwater tables at the local level (Lambin & Geist, 2006), and land degradation (Yohannes et al., 2020). The livelihoods of local societies, sustainable development, and food security may be significantly affected by these complex environmental effects (Lambin & Geist, 2006).

A substantial amount of literature has been published at regional and global levels, but much of the study focuses on highland tropical deforestation and its effects on ecosystems and biomes, such as savannahs and wetlands (Bewket & Abebe, 2013). Much remains to be studied about the direct and indirect factors that influence changes in land use and land cover at the local level (Lambin et al., 2003). According to Lambin et al. (2003), findings from local LULC research can help to predict and explain LULC changes that are likely to emerge at larger geographic scales. However, the degree, pattern, and speed of LULC changes are extremely severe and pervasive at the local level because of significant changes in social advancement and population growth (Gebreslassie, 2014). In numerous parts of the undeveloped nations, rapid population growth has often led to LULC changes triggered through natural resource degradation (Ramankutty & Foley, 1999) aimed at boosting agricultural production and other consumption materials. In developing countries such as Ethiopia, the deterioration of land resources, including LULC changes, has a significant impact on agricultural land and the livelihoods of around ten million farmers in the south-central highlands. Ethiopians still remember the 1984 and 2002/2003 droughts vividly, and in 2011, another severe drought occurred because of LULC changes (Haregeweyn et al., 2015). LULC modifications are also considered one of the main environmental challenges in south-central Ethiopia, as elsewhere (Gashaw et al., 2017; TekleL & Hedlund, 2000), where agricultural activity is the mainstay of the economic sector and 85% of the population is engaged as the main source of income generation (Berihun et al., 2019). In south-central Ethiopia, the primary utilization of land is for smallholder farming, leading to alterations in LULC due to deforestation (Tefera, 2011). In terms of natural resource endowment, Ethiopia is primarily renowned for its diverse vegetation. However, these resources are under pressure, such as population growth leading to an increasing demand for open land for agriculture and settlements, which has led to soil erosion and the reduction of the country's forest resources (Ariti et al., 2015). As a result of the conversion of many forest areas to cultivated lands and settlement areas, such challenges currently produce conflicts and have detrimental effects on equitable natural resource sharing among local communities (Ariti et al., 2015; Fentahun & Gashaw, 2014). Similar research on the Ojoje watershed in south-central Ethiopia was also done after studying LULC specifically related to the watershed (Mariye et al., 2022a). It is crucial to gather thorough, trustworthy, and current information that connects the patterns and main causes of the disturbance of natural resources, and this information is missing from earlier studies of a similar landscape at the study site.

According to recent studies, the eastern highland region has experienced substantial LULC dynamics, drives, and consequences (Desta & Fetene, 2020; Kidane et al., 2012). However, contrasting agro-ecological-based LULC changes have received little attention, and there is no comprehensive scientific evidence in the south-central highlands. Thus, contrasting agro-ecological LULC trends, drives, and consequences have been studied to fill these research gaps. Studying contrasted agro-ecological LULC changes and incorporating this knowledge into land-use science is an essential prerequisite for effective land use planning, biodiversity conservation, land management policies, and strategies by decision-makers for designing sustainable natural resource management (Ayalew et al., 2022). Moreover, the study watersheds, Satame and Lagabora, were once known for their cereal production, high population, diverse agroforestry, and livestock, but are now highly vulnerable to land degradation, deforestation, and drought (Belayneh & Eyasu, 2021). The extent, magnitude, and trends of LULC change in the study watersheds have not been adequately addressed, which has led to poor documentation of the changes, drivers, and their implications. Therefore, this study was initiated to map and analyze LULC dynamics between 1976 and 2022 and their possible consequences in the contrasting agro-ecological environments of the Sateme (midland) and Lagabora (highland) watersheds. It comprises categorizing the acquired image data, creating maps of the land cover, and computing changes that occurred over the study periods. Thus, the study should provide comprehensive, accurate, and current information to develop strategies to preserve and utilize natural resources in a sustainable way to meet the needs of future generations and protect the natural resource base. Additionally, this research can inform policy decisions aimed at mitigating the negative effects of human activity on the environment.

Description of the study areas

The study areas were situated in Kembata Tembaro district (Fig. 1) and are located within the rift valley and Belate basin of south-central Ethiopia. The watersheds are about 350 kilometers south-west of Addis Ababa (the capital of Ethiopia). Geographically, the district is positioned between 71⁰03’048’’–75⁰00’946’’ latitude and 37⁰341’711’’–38⁰ 078‘ 387’’longitude and the two watersheds cover about 31,080.2 ha of Legabora and 22,797.6 ha of Satame, respectively. The study landscape has a share of three woreda’s (the smallest administrative unit of the country). Lagabora watershed is in Damiboya and Kedida Gamella, while Satame is in Angecha Woreda. It is characterized by undulating mountainous topography with an elevation stretching from 2300–3080 m.a.s.l. (KTARDO, 2019). These two watersheds were chosen to represent two distinct agro-ecological environments, categorized on the basis of agro-ecology, vegetation class, rainfall, temperature, and farming system, as summarized in Table 1 (Hurni et al., 2016). The watershed was dominated by the dry evergreen Afromontane forest and grassland complex vegetation. Furthermore, there are several indigenous tree species that serve in the study site, including Hagenia abyssinica, Podocarpus falcatus, Juniperus procera, Cordia africana, and Olea europaea subsp (Hailu et al., 2020). There are also many other very important exotic trees found in the watershed, including Cupressus Lusitania Mill, Eucalyptus camaldulesis Dehnh., Eucalyptus globules Labill, which are currently replacing the need for wood purposes of the local community (Hailu et al., 2020). Due to its rapid growth and soci-economic benefits, Eucalyptus globulus Labill is also the predominant tree species that is frequently used for building and for supplying the energy needs of rural communities.

Table 1

Main biophysical features of Legabora and Satame watershed
Biophysical characteristics of watershed	Legabora (highland)	Satame (midland)
Area (ha)	31, 080.2	22, 797.6
Agro-ecological zone	Moist subtropical (wet Dega)	Humid subtropical (Moist Weyna Dega)
Major soil types	Vitric Andosols, Pellic Vertisols, Chromic Luvisols	Pellic vertisols, Luvisols, Andosols, and Eutric nitosols,
Dominant crops types ^d	Wheat, barely, teff, Ensete, chat, vegetables, fruit trees	Maize, wheat, potatoes, sorghum, teff, fruit trees
Precipitation pattern (mm)	Unimodal	Unimodal
Mean monthly temperature (⁰C)	12–19	19–25
Dominant livestock	Cattle, sheep, horse, donkeys	Goats, sheep, cattle, donkeys
Geology type	Quaternary extrusive and intrusive	Tertiary extrusive and intrusive
Major vegetation class	Dry ever green Afromontane forest and grassland complex	Dry ever green Afromontane forest and grassland complex
Commonly applied soil and water conservation structures/activities	Medium (Soil bund, fanya juu, trenches, micro-basin)	High (Soil bund, fanya juu)
^d Teff (Eragrostis tef), Wheat (Triticum aestivum), Maize (Zea mays), Enset ( Ensete ventricosum), Barely (Hordeum vulgare), sorghum (Sorghum bicolor).

Figure 1 (Inserting near here)

On the study site, there are two rainy seasons with bimodal rainfall. “Maxxe Sana” and “Glalichi Sana” The average annual rainfall (Houghton et al.) in the area ranges from 1200 to 1800mm. Locally, "Glalichi Sana" refers to a shorter rainy season that lasts from March to May and contributes 19% of the total amount of precipitation (RF). The primary agricultural activity in the study watershed is dominated by mixed farming. Sheep, cattle, horses, goats, and donkeys are the main livestock species that smallholder farmers raise as a secondary activity. Most farmers in the study area own at least two different types of land, including home gardens and primary land. Building houses, growing cereals, grazing livestock, and planting trees are all permitted on the mainland, while the planting of valuable crops like Enset (Ensete ventricosum) and fruit trees takes place in home gardens. Annual crops like teff (Eragrostis tef), wheat (Triticum aestivum), maize (Zea mays), barely (Hordeum vulgare), sorghum (Sorghum bicolor) (Wolancho, 2015), and potatoes are grown on lands that are far from homesteads. The dominant soil types are Leptosols, chromic vertisols, Haplic Luvisols, and Eutric Nitisols (Wolancho, 2015). According to Ethiopian agro-ecological zone classification, Woyina Dega (midland, 1800–2400) makes up 70.8% of the district, Dega (high land, 2400–3200) makes up 22.3%, and Kolla (lowland, 500–1800) makes up7% m. a. s. l (Maryo, 2020).

Table 1 (inserting near here)

Data sources and analysis methods

At the global scale, satellite images play a key role in monitoring landscape change both spatially and temporally (Mammo et al., 2018). A total of four multi-temporal Landsat images for the years of 1976, 1991, 2001, 2011, and 2022 were downloaded from the USGS data portal and EROS center, where image data sets are archived, to study trends and quantify the magnitude and dynamics of LULC change (Hailu et al., 2020). The following factors were considered when selecting satellite images: (i) the principal events that occurred in the area (Muriithi, 2016) and image quality, which reduces the impact of clouds on land cover mapping; (iii) anticipated major changes and year-to-year consistency between the study sites. We accessed satellite images with a minimum cloud cover of < 10% at the early and mid-crop harvest season between November and January, which is before the usual time of the existence of forest fires and the dry season under clear cloud cover conditions. Moreover, we employed multispectral satellite data (Landsat with sensors MSS (1976), TM (1991), ETM+ (2001, 2011), and OLI (2022), respectively) to identify and evaluate satellite images (Table 2).

Table 2

Explanation of imagery statistics and data sources used to produce LULC maps in Satame and Legabora watersheds
Watershed	Satellite sensors	Source	Acquisition date	Resolution scale (m)	Path/Row	Spectral resolution type
Satame	MSS	USGS	January–1976	57*57^a	169/55	Multi-spectral
	TM	USGS	February–1991	28.5*28.5	169/55	Multi-spectral
	ETM+	USGS	January–2001	30*30^c	169/55	Multi-spectral
	OLI-TIRS	USGS	April–2011	30*30^c	169/55	Multi-spectral
	OLI-TIRS	USGS	March–2022	28.5*28.5	169/55	Multi-spectral
Legabora	MSS	USGS	February–1976	57*57^a	169/55	Multi-spectral
	TM	USGS	March–1991	28.5*28.5	169/55	Multi-spectral
	ETM+	USGS	January–2001	30*30^c	169/55	Multi-spectral
	OLI-TIRS	USGS	January–2011	30*30^c	169/55	Multi-spectral
	OLI-TIRS	USGS	March–2022	28.5*28.5	169/55	Multi-spectral
^c panchromatic enhanced to 15m to facilitate visual interpretation, but 28.5 resolution images from 1991 to 2022 were used to conduct supervised classification and change analysis
^a Mosaic

We obtained DEM and a topographic map of the study area (1:50000) from the space science and geospatial information institute of Ethiopia, and then the images were orthorectified into Universal Transfer Mercator (UTM) Zone 37N, World Geodetic System (WGS) 1984. Remote sensing and GIS tools, including ERDAS Imagine 2014 and ArcGIS 10.6, were used for image processing and data analysis, respectively. At first, images were converted to UTM and geo-referenced using a datum selected by WGS-84 for Ethiopia. We digitized the demarcated study area in Arc GIS 10.6 to overlay the view on the spatial databases generated by the photograph and the satellite image. Aerial photographs from 1991 to 2022 were interpreted via mirror stereoscope for LULC type identification and classification, whereas post-classification change detection was used for the evaluation of LULC types on the aerial photographs (Mehari. et al., 2022). For the making of the latest land cover map, Google Earth Imagines, field surveys, and ground control points (GCPs) were used as major sources of data. LULC trends and dynamics were assessed using Landsat imagery, which provides a range of temporal, spectral, and multi-resolution capabilities for land cover analysis following Oettera et al. (2000). We coded and organized the household (HH) questionnaires before conducting analysis. Coding was established during the preparation of the questionnaires. While the qualitative data of the focus group discussions (FGDs), key informant interviews (KIIs), and observational notes were transcribed, arranged, and analyzed, then, we used Origin Pro to analyze and summarize the quantitative data gained from general informants during a formal survey. In order to enhance the quality of the image, sub-setting and radiometric enhancement techniques (haze reduction and histogram equalization) were applied to the raw data following Belete et al. (2021). The key drivers of LULC changes have been ranked based on their severity in the local context using a preference rating. Numerous changes in LULC drivers perceived by respondents in Satame and Legabaora watersheds were highlighted by references (Angessa et al., 2019). Nine significant factors were chosen by agreement of those who participated and ranked in random order. The rank was assigned by an integer number (1, 2, 3,... 9) with the greatest number (9) allocated for the most driving force and (1) given for the lowest score (Angessa et al., 2019). And finally, forty respondents who actively participated in the discussions and presented more details on driving factors were chosen to rate the nine identified drivers with respect to the degree of influence they had on watersheds (Angessa et al., 2019).

Table 2 (Inserting near here)

After the image was processed, signatures were distributed per pixel by identifying the land into eight classes between 1976 and 1991. Furthermore, eight LULC categories have been recognized for the years 2011 and 2022, respectively. Image classification was based totally on the reflectance characteristics (false-color composite) of the specific land cover classes and also supplemented using intensive field observation (Belete et al., 2021), KIIs, and FGDs. Each class was given a unique identification and assigned a selected color to differentiate one from another. For each of the predetermined LULC categories, training samples had been selected via a delimiting polygon around representative sites. During this time, the Google Earth extension was employed to reduce confusion in interpreting the pixel. Additionally, ArcGIS 10.6 was used to prepare the map layout, reclassify the results from the supervised class, and estimate the pixel values for all LULC classes after importing the results from ERDAS Imagine 2014 (Belete et al., 2021; Othow et al., 2017). The conversion matrices were created using ERDAS Imagine 2014, and the source and destination of each land cover value were evaluated in an Excel spreadsheet. In addition to calculating the area of LULC in hectares and percent, we also computed the percentage change in LULC between the stated period following Braimoh (2006) and Pontius et al. (2004).

$$\text{A}\text{r}\text{e}\text{a} \text{i}\text{n} \text{h}\text{e}\text{c}\text{t}\text{a}\text{r}= \frac{\text{C}\text{o}\text{u}\text{n}\text{t} \times 900 \text{o}\text{r} 3600}{10, 000}$$

………………………………………………………………………………..

$$\text{A}\text{r}\text{e}\text{a} \text{i}\text{n} \text{p}\text{e}\text{r}\text{c}\text{e}\text{n}\text{t} =\frac{\text{V}\text{a}\text{l}\text{u}\text{e} \text{o}\text{f} \text{t}\text{h}\text{e} \text{i}\text{d}\text{e}\text{n}\text{t}\text{i}\text{f}\text{i}\text{e}\text{d} \text{p}\text{i}\text{x}\text{e}\text{l} }{\text{N}\text{u}\text{m}\text{b}\text{e}\text{r} \text{o}\text{f} \text{p}\text{i}\text{x}\text{e}\text{l}\text{s} \text{i}\text{n} \text{t}\text{o}\text{t}\text{a}\text{l} }\times 100$$

………………………………………..………………………

Where Pt₁ is the area of LULC at the initial period and Pt₂ is the area of LULC at the final period. A positive result indicates that the extent of LULC has increased, whereas a negative result indicates that the amount has decreased. We validated the classification results by constructing a confusion matrix as a basis for determining accuracy assessments (Belete et al., 2021; Miheretu & Yimer, 2017). Furthermore, we used topographic maps obtained from the Ethiopian Geospatial Information Institute to visualize the landscape of the whole study area (Anchan et al.). We collected ground trothing points with the support of FGDs and KIIs during the transect walk. In addition, Google Earth Engine images were used as a supplementary tool for ground truth points (GTPs). Moreover, GTPs for 1976 and 1991 LULC maps were collected by using the false color composite of 1976 and 1991 satellite images in conjunction with elder knowledge based on previous studies (Betru et al., 2019). In all scenarios, members of the KIIs and FGDs were employed to collect GTPs by describing the dynamics of LULC changes occurring in and around their locality. This study has applied both unsupervised and supervised cataloging techniques, following previous researchers (Belay & Mengistu, 2019; Congalton, 2001; Congalton & Green, 2009 ). In supervised classification, specific land cover types are delineated using training sites, whereas in unsupervised classification, land cover classes are formed based on the number of classes requested. Moreover, we combined unsupervised classifications with visual signature editions based on the spectral values of recent images to determine the classification of historical images of 1976 (Andualem et al., 2018). A supervised signature extraction with the maximum likelihood algorithm was used to categorize the Landsat images (Amare, 2015; Congalton & Green, 2009 ). Finally, LULC maps for the respective years were produced for future analysis. Moreover, eight classes of LULC categories (Table 3) were identified and categorized into cropland, settlement, water body, wetland, grassland, bare land, forest, and shrub land to allow us to evaluate the tendencies of every factor in distinctive agro-ecologies and their upcoming consequences, and the description (Berihun et al., 2019) is provided from various supporting data (Kindu et al., 2013).

Table 3

LULC categories and their description used to measure the changes in periods from 1976 to 2022 (based on FAO classification)
Code	LULC class	Their descriptions
ST	Settlement	A land-use type that includes rural settlement area, educational, health, socio-economic facilities, residential houses, administrative buildings, small-scale industrial areas, etc.
WL	Wetland	Land use that is waterlogged and swampy during the wet season, which dries in the sunny season.
CL	Cropland	Smallholder farmers owned land that is used to grow cropland. It is characterized by tilled and planted, bare crop fields, and limited areas temporarily left as fallow.
SL	Shrub land	Areas covered by bushes, and shrubs > 20% cover, and mixed with grasses; less than 20% tree cover.
FL	Forest land	Land covered with dense trees > 80% canopy cover which includes evergreen forest land, mixed forest, and plantation forests
GL	Grassland	A land-use type where the land is dominated by grasses, forbs, and herbs with nil or little proportion of shrubs.
BL	Bare land	Areas with little or no vegetation cover consist of exposed soil quarries
WB	Water body	The area covered by water (ponds, streams and rivers)

Table 3 (Inserting near here)

Socio-economic data

LULC trends and priorities were evaluated through a socioeconomic survey of the local community regarding resource availability and usage, government decision-making procedures, and governance practices. To validate the data obtained from aerial photos and satellite imagery, a socioeconomic survey was undertaken with 134 randomly selected households. The district, administration, and household participants were chosen using a three-stage sample technique that included purposive and random sampling, while the household respondents are being selected through systematic random sampling following Wubie et al. (2016). Based on a household survey, we picked 34 households from two sample areas and 33 households from another two sample areas. Two phases of sampling were carried out for the socioeconomic survey. The first step involved choosing the sampling locations, but the second stage involved taking individual households from the selected kebeles (the lowest level of Ethiopian government organization after districts) using system sampling in accordance with earlier methods (Amare, 2015).

The schematic diagram represents the general flow chart of the methodology used for the LULC change analysis as depicted in Fig. 2. Additionally, focus group discussions (FGDs) with 35 participants (15 women and 20 men) were conducted to acquire additional information about the long-term LULC practices and laws that affect local land management. To gather information on the trends of LULC alteration, the effects of LULC modifications on the farmer's livelihoods and the environment, land degradation, and how the existing socioeconomic activities affected LULC over the last four and half a decades, 20 people (age > 65 years) were purposefully chosen as key informants (Danano et al.). The purpose of the discussion and interviews was to pinpoint the key factors that would influence LULC changes (Berihun et al., 2019) and ascertain how the LULC would disrupt the long-term socioeconomic welfare of the watershed community. We computed the study sample size following the research approach proposed by Kothari (2004).

$$\text{n}=\frac{{\text{Z}}^{2}. \text{p}. \text{q}. \text{N}}{{\text{e}}^{2}\left(\text{N}-1\right)+{\text{Z}}^{2}. \text{p}. \text{q} }$$

………………………………………………………………………………………………….

Where n is the sample size and Z denotes the 95 confidence limit (interval) under the normal curve, which would be 1.96. P = 0.1 (the percentage of the population to be included in the sample is 10 percent). q = none event occurrence = 1–0.1, which denotes (0.9). N = is the total number of households (4100). e = allowable error term (margin of error or degree of accuracy) (0.05).

Thus$\text{n}=\frac{{1.96}^{2}\text{*} 0.1 \text{*} 0.9 \text{*} 4100}{{0.05}^{2}\left(4100-1\right)+{1.96}^{2}\text{*} 0. 1\text{*} 0.9 }=134$

Figure 2 (Inserting near here)

Accuracy assessment for image classification

Accuracy assessment is a vital validation technique and a requirement in remote sensing (Okeke & Karnieli, 2007) because it tells us how well-generated classifications perform. One of the most popular techniques for displaying accuracy rating data is in the form of a confusion or error matrix (Congalton, 1991). Based on the findings of the post-classification accuracy assessment, it was determined that in 2022, wetland had the highest overall user accuracy evaluation (100%) and settlement had the lowest (75.3%) in the Legabora watershed, whereas in Satame, bare land had the highest overall user accuracy evaluation (100%) and forest had the lowest (81.7%). As shown in Table 4, the producer accuracy for the same study periods in Legabora was (85.7%) for bare land, (92.5%) for forest, (91.8%) for settlement, (89.8%) for cropland, and (83.6%) for water bodies, respectively.

Table 4

Accuracy assessment table for eight LULC classes in Satame and Legabora (2022)
Year	LULC	WSD	ST	WB	BL	SL	WL	CL	GL	FL	UA%	K
2022	ST	Legabora watershed	45	3	5	1	0	5	1	0	75.3	0.83
	WB		0	51	1	7	0	3	0	1	78.5
	BL		0	4	60	3	1	1	4	0	82.2
	SL		4	0	1	65	0	0	4	0	87.8
	WL		0	0	0	0	6	0	0	0	100
	CL		0	0	0	0	1	114	0	1	98.3
	GL		0	3	3	0	0	4	60	2	83.3
	FL		0	0	0	6	0	0	5	49	81.7
	PA%		91.8	83.6	85.7	79.3	75	89.8	81.1	92.5
	OAA	89.8
2022	ST	Satame watershed	56	2	0	0	0	3	0	3	87.5	0.85
	WB		0	77	0	5	0	5	1	0	87.5
	BL		0	0	1	0	0	0	0	0	100
	SL		2	3	0	89	0	2	0	6	87.3
	WL		0	0	0	0	9	2	0	0	81.8
	CL		0	2	0	2	0	82	0	2	91.1
	GL		2	2	0	0	0	2	80	0	93
	FL		1	4	0	0	0	5	0	49	81.7
	PA%		91.8	85.6	100	92.7	100	81.2	98.8	81.7
	OAA		84.2

The user accuracy was highest for bare land (100%) and lowest for forest (81.8%), while producer accuracy was highest for wetland and bare land (both at 100%) and lowest for cropland (81.2%) in the Satame watershed, respectively. Throughout our field observation and transaction walk with the help of Google Earth Engine, we gathered a total of 893 Ground Control Points (GCPs) from a GPS. Of these, 493 points were used for supervised classification, with the remaining 400 GCPs being used to calculate the accuracy of satellite images. We followed the approaches proposed by Lillesand et al. (2008): a minimum of 50 points for each LULC type with less than 4000km² of area and lower than 12 categories. Based on the confusion matrix, which includes both the producer's and user's accuracy, the accuracy metric was developed (Asokan & Anitha, 2019; Thakkar et al., 2017). In order to compute the user accuracy, we divided the number of correctly classified pixels in each class by the number of training set pixels per classified class, which indicates the percentage of correctly classified pixels per LULC class (Othow et al., 2017). A research study conducted by Belete et al. (2021), reported that producers' and users' low accuracy records may have been caused by the homogeneity of land cover classes at the spectral level.

The number of samples in each class was weighted to determine the overall classification accuracy for both watersheds, i.e., the sum of all samples on the diagonal divided by the total number of cells (45 + 51 + 60 + 65 + 6 + 114 + 60 + 49)/501, which is 89.8%. This result showed that in Legabora, 89.8% of LULC categories have been correctly classified, compared to Satame, where the percentage was (56 + 77 + 1 + 89 + 9 + 82 + 80 + 49)/526, or 84.2%. In total, kappa statistics of 0.83 were obtained for the classification period of 2022 images, indicating that there is 83% and 85% better agreement than would be predicted by chance alone for the Legabora and Satame watersheds, respectively, as shown in Table 4. The columns contain reference data, while the rows offer classifications produced from remotely sensed data. Bolded diagonal numbers demonstrate properly categorized sites based on reference data, whereas off-diagonals represent incorrectly classified data.

Table 4 (Inserting near here)

Note

CL; Crop land, GL = Grass land; FL = Forest land; SL = Shrub land; BL = Bare land; WL = Wetland; ST = Settlement land; WB = Water body; UA = User accuracy; PA = Producer accuracy; OAA = Overall accuracy; K = Kappa statistics; WSD = Watershed

In a study by Hassan et al. (2016), the kappa coefficient is associated with three possible groups of the covenant: the value over 0.80 shows strong agreement, the value between 0.40 and 0.80 denotes moderate covenant, and the value below 0.40 indicates poor covenant (Belete et al., 2021). Therefore, this kappa accuracy was generally regarded as reasonable for the subsequent analysis and change detection. For all LULC, as shown in Table 4, kappa statistics showed a strong correlation between the classification map and ground reference data. Finally, it was calculated following the previous study approach proposed by Tilahun (2015).

$$\text{K}= \frac{\text{N}{\sum }_{\text{i}=1}^{\text{n}}{\text{m}}_{\text{i}\text{i}}{-{\sum }_{\text{i}=1 }^{\text{n}}\left({\text{G}}_{\text{i}}{\text{C}}_{\text{i}}\right)}^{}}{{\text{N}}^{2}-{\sum }_{\text{i}=1}^{\text{n}}\left({\text{G}}_{\text{i}}{\text{C}}_{\text{i}}\right)}$$

…………………………………………………………………………….

Where: i denotes the class number, (N) represents the total number of classified values in comparison to truth values, (m_ii) is the number of values that fall into the truth class i and are also classified as values found along the diagonal of the confusion matrix, (Kløve et al.) is the sum of all predicted values belonging to class i, and (G_i ) indicates the total number of truth values in class

Spatio-temporal distributions of LULC change within the Satame and Legabora watersheds

During the 1976 study period, shrublands and grasslands were the most common land cover types, covering 32.9% and 32.4% of the Satame watershed, followed by croplands at 24.8%, the forest at 5.5%, and bare lands at the lowest percentage, respectively. Notably, settlement lands were entirely lost during this period, as per the data presented in Fig. 3. However, a shift in land cover was observed in the second study period conducted in 1991, with croplands being the dominant land use type at 37.7%, followed by grasslands at 26.7%, shrublands at 25.1%, forest lands at 3.9%, and wetlands at 3% (Fig. 3). Settlements constituted the smallest percentage of land usage on the study site during this period (Table 5).

Table 5

Spatial coverage and eight LULC classes distribution for each time phase
WSD	LULC	1976		1991		2001		2011		2022
WSD	LULC	(ha)	%	(ha)	%	(ha)	%	(ha)	%	(ha)	%
Satame watershed	ST	0	0	8.73	0.04	7.11	0.03	17.28	0.08	255.33	1.12
	WB	657.08	2.88	657.54	2.88	663.03	2.91	657.54	2.88	657.54	2.88
	BL	52.2	0.23	170.91	0.75	2.43	0.01	25.11	0.11	14.85	0.07
	SL	7498.11	32.89	5710.14	25.05	3855.24	16.91	3034.08	13.31	1707.66	7.49
	WL	292.6	1.28	686.7	3.01	605.43	2.66	196.2	0.86	71.1	0.31
	CL	5656.32	24.81	8591.67	37.69	13733.19	60.24	13659.66	59.92	16904.97	74.15
	GL	7379.12	32.37	6080.67	26.67	993.96	4.36	693.72	3.05	1080.27	4.74
	FL	1262.02	5.54	891.09	3.91	2937.06	12.88	4513.86	19.79	2105.73	9.24
Total		22797.45	100	22797.45	100	22797.45	100	22797.45	100	22797.45	100
Legabora watershed	ST	0	0	0	0	0.99	0	62.84	0.2	263.79	0.85
	WB	376.02	1.2	376.02	1.2	376.02	1.2	376.02	1.2	376.04	1.2
	BL	274.76	0.88	1177.38	3.79	33.93	0.11	78.77	0.25	389.97	1.25
	SL	7990.92	25.71	5549.04	17.9	3052.53	9.8	3638.36	11.73	1489.77	4.79
	WL	0	0	0	0	230.76	0.74	62.01	0.19	43.47	0.15
	CL	7440.92	23.91	10125.27	32.5	17424.03	56.1	17970.29	57.83	22969.25	73.9
	GL	14266.68	45.9	12491.13	40.2	7261.81	23.4	4897.15	15.75	3097.19	9.97
	FL	730.88	2.4	1361.34	4.4	2700.11	8.65	3994.74	12.85	2450.7	7.89
Total		31080.18	100	31080.18	100	31080.18	100	31080.18	100	31080.18	100

Croplands made up about 59.9% (13659.7 ha) of the stud area in 2011, followed by forests (19.8%) and shrubs (13.3%), or 3034.1 ha Fig. 4. Settlements cover the lowest amount of land area. Although wetlands and bare land accounted for the lowest proportion of the Satame watershed in the final study period (2022), approximately 74.2% of the watershed was covered by croplands, followed by forest at 9.2% and 7.5% of shrublands. In 1976, wetland and settlement lands were completely lost, and grassland made up 45.9% of the Legabora watershed, followed by shrublands (25.7%). About 73.9% of the total land area was covered by cropland during the last study period (2022) in Legabora, and the remaining share was covered by other LULC classes. (Table 5) provides an illustration of the spatial extent of LULC changes and their distribution for each time phase.

Table 5 (Inserting near here)

Figure 3 (Inserting near here)

Figure 4 (Inserting near here)

Trends and dynamics of LULC change in Satame and Legabora from 1976 to 2022

State of cropland LULC from 1976 to 2022

As elsewhere in Ethiopia, the majority of the rural populations in the Legabora and Satame watersheds depend heavily on agriculture, particularly crop cultivation. The study area underwent significant LULC changes, according to an analysis of LULC trends over the past five consecutive study periods (Table 6). The LULC trend analysis from 1991 to 2001 showed that in Satame (midland) and Legabora (highland), respectively, cropland increased at annual increasing rates of 514.2 and 729.9 ha/year at the expense of other land cover classes (Table 6). Between 2001 and 2011, the Satame LULC class also experienced a − 7.4 ha/year annual decline, but it steadily increased in Legabora (Fig. 5). Growing demand for more land for eucalyptus plantations and open-up settlements may be the cause of the decline in cropland. In the Legabora watershed, cropland has been the most prevalent LULC type over the course of the study, accounting for 23.9% in 1976 and 57.8% in 2011. Similar to this, the area covered by cropland in the Satame has steadily increased, rising from 24.8% in 1976 to 60.0% in 2011. In the final period (2022), the situation had changed, and it occupied 74.2% of the Satame and 73.9% of the Legabora watersheds, respectively, while other LULC types accounted for the remaining shares.

Table 6

Rate of LULC dynamics from 1976 to 2022 in Satame and Legabora
WSD	LULC	1976–1991		1991–2001		2001–2011		2011–2022		1976–2022
WSD	LULC	AC (ha)	RC (ha/year)	AC (ha)	RC (ha/year)	AC (ha)	RC (ha/year)	AC (ha)	RC (ha/year)	AC (ha)	RC (ha/year)
Satame watershed	ST	8.73	0.58	–1.62	–0.16	10.2	1.02	238.1	23.8	255.3	5.67
	WB	0	0.46	5.5	0.55	–5.5	–0.55	0	0	0	0.46
	BL	118.7	7.91	–168.5	–16.9	22.7	2.27	–10.3	–1.03	–37.4	–0.83
	SL	–1788	–119.2	–1854.9	–185.5	–821.16	–82.1	–1326.4	–132.6	–5790.5	–128.7
	WL	394.0	26.3	–81.27	–8.13	–409.23	–40.9	–125.1	–12.51	–221.6	–4.9
	CL	2935.4	195.7	5141.5	514.2	–73.5	–7.4	3245.3	324.53	11248.7	249.97
	GL	–1298.5	–86.6	–5086.7	–508.7	–300.2	–30.02	386.6	38.66	–6298.9	–139.9
	FL	–370.95	–24.73	2045.97	204.6	1576.8	157.7	–2408.1	–240.81	843.7	18.8
Legabora watershed	ST	0	0	0.99	0.10	61.9	6.2	200.95	20.10	263.8	5.86
	WB	0	0	0	0	0	0	0	0	0	0
	BL	902.6	60.17	–1143.45	–114.4	44.8	4.48	311.2	31.12	115.2	2.56
	SL	–2441.9	–162.8	–2496.5	–249.7	585.8	58.6	–2148.6	–214.7	–6501.2	–144.5
	WL	0	0	230.76	23.08	–168.8	–16.9	–18.54	–1.85	43.47	0.97
	CL	2684.4	178.96	7298.8	729.9	546.3	54.6	4998.96	499.90	15528.3	345.1
	GL	–1775.6	–118.4	–5229.3	–522.9	–2364.7	–236.5	–1799.9	–179.9	–11169	–248.2
	FL	630.5	42.03	1338.8	133.9	1338.8	133.9	–1544.1	–154.4	1719.8	38.2

According to the LULC classification, cropland was determined to be the most common land cover type across all evaluation periods in the Satame and Legabora watersheds. This demonstrated that in order to increase crop production, croplands were expanded and shrubs and grassland were planted instead. The need for food crops and more accessible land for settlements may also be contributing factors to this growth. According to data gathered through key informant interviews and discussions with FGDs, population pressure was what caused the expansion of croplands as well as the decrease in the amount of grassland and forestland cover. These results support earlier research by Shiferaw and Singh (2011), Siraj et al. (2018), and Hailu et al. (2020), who, respectively, reported a 65% augmentation in cropland.

Table 6 (Inserting near here)

Figure 5 (Inserting near here)

State of wetland LULC between 1976 and 2022

(Table 5) demonstrated that certain LULC classes in the Satame and Legabora watersheds experienced spatiotemporal changes. In the Satame watershed, wetland accounted for total land areas of 3%, 2.7%, 0.9%, and 0.3% in 1991, 2001, 2011, and 2022. Wetlands covered a total land area of 686.7 ha in 1991, 605.4 ha in 2001, and 192.2 ha in 2011 (Fig. 4). However, this coverage declined to 71.1 ha (0.31%) in 2022. This demonstrated that the wetland area maintained declining trends over the spatiotemporal study periods. In summary, the LULC trend analysis revealed that wetlands diminished by (–221.6 ha) between 1976 and 2022, with a percentage change of − 75.7 and an annual decreasing rate of (–4.9 ha/year) in Satame, whereas in the similar study period, wetland area demonstrated increasing trends by 43.5 ha with a percentage change of 100 and an annual increasing rate of (0.97 ha/year) in Legabora watershed, respectively (Table 6). The outcomes of Belayneh et al. (2018), who found a significant decline and complete drying up of wetlands in Ethiopia, are in direct opposition to the study results in the Legabora watershed. Likewise, Hailu et al. (2020) discovered that the area of the wetland decreased steadily from 1973 to 2019, with an average shrinking rate of 172.6 ha/year.

State of bare land LULC from 1976 to 2022

For the two study watersheds of Legabora and Satame, we created classified LULC maps as depicted in Figs. 6 and 7. However, the LULC analysis showed that the area covered by bare land in the Satame (midland) watershed steadily decreased between 1976 (52.2 ha) and 2022 (14.9 ha) over the course of the study span. But in general, the LULC findings showed that bare land decreased by − 37.4 ha in the final comparison period of the Satame watershed (1976–2022), with a percentage change of − 71.6%, and an annual decreasing rate of − 0.83 ha/year has been observed over the last 45 years. This may have been linked to the use of appropriate land management techniques and practicing water and soil conservation activities across sloppy areas, which in turn reduced soil erosion (e.g., wind and water). However, the results of the current study are consistent with a study done in the Wallecha watershed from 1984 to 2000, which showed a trend of consecutively decreasing bare land from 4.8 to 3.2% Barana et al. (2016) and Hailu et al. (2020) also reported that the area of bare land decreased from 7% in 1973 to 1.3% in 2019.

Figure 6 (Inserting near here)

Figure 7 (Inserting near here)

Status of forestland LULC from 1976 to 2022

Forest land covered 2.4%, 8.7%, and 12.9% of the Legabora watershed in 1976, 2001, and 2011, respectively. During the fourth and fifth study periods, the forest cover in this watershed showed an increasing and decreasing trend. In addition, there was also an increase in the forest cover class of the Satame from its level of 3.9% in 1991 to 12.9% in 2001 and 19.8% in 2011 (Table 5). This increase may be related to activities like soil and water conservation carried out as part of a sustainable natural resource management programme, as stated by Barana et al. (2016). When compared to its first period (1976), it also displayed declining trends during the second study period, with a net increase in forestland cover. However, in general, the LULC findings confirmed that forest land was reduced by − 370.9 ha (Fig. 8) in the initial comparison period (1976–1991) of the Satame watershed, with a percentage change of − 29.4% and an annual decreasing rate of − 24.7 ha/year. In the Legabora watershed, it also decreased between 2011 and 2022 by − 1544.1 ha, with a percentage loss of − 38.7% (Table 7) and − 154.4 ha/year annual decreasing rate. The following factors contributed to the loss of the forest and were mentioned by informants for the current study: The expansion of cropland in 1991 is related to the decline in the trend of forest cover. In addition, KIIs and FGDs explained that the military government was replaced in 1991/92, creating an open access situation that left forestlands unenforced and illegally converted to other uses (Demissie et al., 2017). Young people without farms and farmers with little farmland cut down trees for firewood and charcoal, which negatively affects the existing changes to forestland.

Table 7

LULC change in (ha) and percent for five investigation periods in Satame and Legabora
WSD	LULC	1976–1991		1991–2001		2001–2011		2011–2022		1976–2022
WSD	LULC	(ha)	(%)	(ha)	(%)	(ha)	(%)	(ha)	(%)	(ha)	(%)
Satame watershed	ST	+ 8.73	+ 873	–1.6	–18.6	+ 10.2	+ 143.04	+ 238.1	+ 1377.6	+ 255.3	+ 0
	WB	+ 0	+ 0	+ 5.5	+ 0.8	–5.5	–0.83	+ 0	+ 0	+ 0	+ 0
	BL	+ 118.71	+ 227.4	–168.5	–98.6	+ 22.7	+ 933.3	–10.26	–40.86	–37.4	–71.6
	SL	–1788	–23.9	–1854.9	–32.5	–821.2	–21.29	–1326.4	–43.72	–5790.5	–77.2
	WL	+ 394.02	+ 134.6	–81.3	–11.8	–409.2	–67.59	–125.1	–63.8	–221.6	–75.7
	CL	+ 2935.4	+ 51.9	+ 5141.5	+ 59.8	–73.5	–0.54	+ 3245.3	+ 23.8	+ 11248.7	+ 198.9
	GL	–1298.5	–17.6	–5086.7	–83.7	–300.2	–30.21	+ 386.6	+ 55.7	–6298.9	–85.4
	FL	–370.9	–29.4	+ 2045.97	+ 229.6	+ 1576.8	+ 53.69	–2408.1	–53.4	+ 843.7	+ 66.9
Legabora watershed	ST	+ 0	+ 0	+ 0.99	+ 0	+ 61.85	+ 6247.5	+ 200.95	+ 319.8	+ 263.8	+ 100
	WB	+ 0	0	+ 0	+ 0	+ 0	+ 0	+ 0	+ 0	+ 0	+ 0
	BL	+ 902.62	+ 328.51	–1143.45	–97.1	+ 44.8	+ 132.2	+ 311.2	+ 395.1	+ 115.2	+ 29.54
	SL	–2441.88	–30.56	–2496.51	–44.9	+ 585.8	+ 19.2	–2148.6	–59.1	–6501.2	-436.37
	WL	+ 0	+ 0	+ 230.8	+ 0	–168.8	–73.1	–18.5	–29.9	+ 43.5	+ 100
	CL	+ 2684.4	+ 30.02	+ 7298.8	+ 72.1	+ 546.3	+ 3.14	+ 4998.9	+ 27.8	+ 15528.3	+ 67.6
	GL	–1775.6	–12.45	–5229.3	–41.8	–2364.7	–32.6	–1799.9	–36.8	–11169.5	–360.63
	FL	+ 630.5	+ 86.3	+ 1338.8	+ 98.3	+ 1338.8	+ 49.6	–1544	–38.7	+ 1719.8	+ 70.18

Additionally, the 1990s civil war and the extraction of local building materials and fuel wood contributed to the degradation of forestland. Moreover, the expansion of cropland and the 1985–1986 famine caused a decline in forest cover, leading to families extracting resources from their surroundings. Furthermore, the LULC analysis showed that between 1976 and 2022, in the Satame and Legabora watersheds, respectively, the forest cover of the study landscape increased at an annual rate of 18.8 and 38.2 ha/year (Fig. 5). Similar to the current study, Andualem et al. (2018) reported that the forest LULC increased by 1.46% between 2007 and 2018. In a similar vein, Gebrelibanos and Assen (2013), verified that the amount of forest cover increased steadily between 1964 and 2006. Unlike the current study, WoldeYohannes et al. (2018) found that between 1985 and 2010, the amount of forest land decreased by 1.71 ha/year. Likewise, Hassen and Assen (2017), reported that the area covered by forests decreased by 23.1% between 2004 and 2014. According to Agidew and Singh (2017), the amount of forest land in northeastern Ethiopia has decreased. As a result, the findings showed that between 1976 and 2022, the forest land in both study watersheds increased at rates of 18.8 and 32.2 ha/year, respectively. FGDs discussants noted that the eucalyptus tree plantation has contributed to the increase in forestland cover due to its importance for building materials and fuel wood, as well as the income generation of local communities.

State of Grassland LULC between 1976 and 2022

In Satame, grassland covered a total area of 7379.1 ha (32.4%) in 1976, 6080.7 ha (26.7%) in 1991, 993.9 ha (4.4%) in 2001, and 693.7 ha (3.1%) in 2011, respectively (Figs. 3 and 4), while between 1976 and 1991, grassland was the dominant LULC, making up 45.9% and 40.2% of the total land area in the Legabora watershed, respectively. According to the current LULC change analysis, grasslands in Satame showed a significant decline by (–1298.5 ha) with a percentage change of − 17.6% (Table 7; Fig. 8), and an annual decreasing rate of − 86.6 ha/year between 1976 and 1991, and by (–6298.9 ha) with an annual shrinking rate of − 139.9 ha/year in the final comparison period, respectively. This LULC class showed a shrinking trend of (–1775.6 ha) with a percentage loss of (–12.5%) and an annual decreasing rate of − 118.4 ha/year in Lagabora’s first comparison period, whereas in the second and third comparison periods, it displayed a similar trend of − 5229.3 and − 2364.7 ha with an annual decreasing rate of − 552.9 and − 236.5 ha/year, respectively (Table 6). Satellite data, which shows a continuous decrease over the study period and is linked to recurrent fires that happen during the dry season and the expansion of farming activities, is supported by FGDs and key informant interviews.

Furthermore, according to responses from the local community gathered through interviews and focus groups, a rise in the demand for agricultural production on already-existing land and the construction of new homes in the study landscape may have contributed to the loss of grassland. The FGDs discussants in the interview session agreed that growing populations have increased demand for food, building materials, and timber, which has caused agricultural land to spread into mountainous and hilly areas, which in turn has resulted in the obliteration of shrub and bush land ecosystems. As a result, the current study revealed that settlement and the conversion of grassland to croplands are frequent occurrences in the study landscape. Abate (2011), reported that grasslands declined by 73 ha/year between 1972 and 1985. Agidew and Singh (2017), stated that in northeastern Ethiopia, between 1973 and 2015, there was a decline in grassland. Bekele et al. (2018), reported that the amount of grassland in the Awash River Basin progressively decreased from 1985 to 2011.

Figure 8 (Inserting near here)

State of shrub LULC from 1976 to 2022

When compared to the other LULC classes in the Satame watershed, which made up 3855.2 ha (16.9%) and 3034.1 ha (13.3%) between the third and fourth periods, shrub land had the second and third highest percentage cover, respectively, as depicted in Figs. 6 and 7. On the other hand, shrublands steadily decreased in the third comparison period, with a percentage change of − 21.3% and an annual shrinking rate of − 82.1 ha/year. According to the current LULC change investigation, shrublands in the Legabora watershed significantly decreased by − 2441.9 ha with a percentage change of − 30.6% (Fig. 9) and an annual decreasing rate of − 162.8 ha/year between 1976 and 1991; however, it increased by 585.8 ha with a percentage change of 19.2% and an annual increasing rate of 58.6 ha/year, respectively, in the third comparison period (2001–2011). Hence, the results indicated that over the past 45 years, the shrublands in the Satame and Legabora watersheds declined at rates of − 128.7 and − 144.5 ha/year, respectively. This may be caused by the fact that they serve as the local communities' primary sources of building materials and firewood, as well as by the expansion of cropland and settlement areas at the expense of shrublands. In the discussion session, FGDs discussants confirmed that charcoal sales contribute to the destruction of shrublands, bushlands, and forests in rural poor communities as a means of generating additional revenue. This result is consistent with a number of earlier studies conducted in different areas of the country. For example, according to Angessa et al. (2019), the Lake Wanchi watershed's shrub land has vanished at a rate of 1.3% per year. According to Agidew and Singh (2017), between 1973 and 2015, the percentage of shrublands decreased from 28.4–24.6%. Amare (2015), came up with comparable results that described the decline of bush land cover. Mikias (2015), claimed that between 1973 and 2010, shrub land increased at a rate of 22.08 ha/year, which is in contrast to the findings of the present study.

State of settlement LULC from 1976 to 2022

In the initial comparison period for Satame as well as for Legabora between 1976 and 1991, there was almost no settlement cover. In the Satame watershed, the percentage of land that falls under this class over the course of the various study periods was 8.7 ha (0.04%) in 1991, 17.3 ha (0.08%) in 2011, and quickly increased to 255.3 ha (1.1%) of the total land area in 2022. In the Legabora watershed, the settlement's size increased from 0% in 1976 and 1991 to 0.2% in 2011 and then augmented to 0.9% in 2022 (Table 5). Based on the current investigation of the LULC change, settlement land indicated a slight increase of 8.7 ha with a percentage change of (873%) and an annual increasing rate of 0.58 ha/year in the initial comparison period, while in the third period (2001–2011), it was augmented by 10.2 ha with a percentage of 143% (Fig. 9) and an annual increasing rate of 1.02 ha/year, respectively, in the Satame watershed, similarly, between 2001 and 2011, Legabora showed an increasing trend of 61.9 ha with a percentage change of 6247.5% and an annual increasing rate of 6.2 ha/year. However, in the fourth and fifth comparison periods, the trend intensified by 200.9 and 263.8 ha with percentage changes of 319.8% (Fig. 9) and 100% and annual increasing rates of 20.1 and 5.9 ha/year, respectively. In this study, we found that between 1976 and 2022, settlement land in the Satame and Legabora watersheds steadily increased at rates of 5.7 and 5.9 ha/year, respectively. The increase in settlement might be attributed to residential growth and the occupancy of public lands by settlers. Additionally, the expansion of settlements onto other lands may have been influenced by population growth. According to Miheretu and Yimer (2017), in the Gelana sub-watershed, settlement areas grew between 1964 and 2014.

Table 7 (Inserting near here)

Figure 9 (Inserting near here)

LULC change matrix in the Satame and Legabora watersheds for five investigation periods

To ascertain a loss and gain in the land size of the main LULC classes, the LULC conversion matrices were examined. Table 8 provides a summary of the findings from the analysis between 1976 and 2022. All of the land areas in the study watersheds have experienced significant changes (both gains and losses), though the magnitude of each LULC class change has varied. The results showed that between 1976 and 2022 there were significant LULC dynamics in which (52 ha) of bare cover, (1094) of forestland, (5656) of cropland, (7498) of shrub lands, (7179) of grasslands, water bodies (1099), and (293 ha) of wetlands were transformed one to another in Satame, respectively, whereas (262) of bare cover, (630) water bodies, (14220) grassland, and (7946) of shrublands were transformed one to another in Legabora, respectively. In Satame, the shrublands, which covered 6693 ha in 1976, decreased to 1698 ha in 2022, whereas Legabora, which covered 7263 ha in 1976, decreased to 1463 ha in 2022. It was taken from one land use category and given to another, then lost to another land use type. As a result, Satame and Legabora had unchanged coverage of 805 and 683 ha, respectively. It gained 321, 399, 98, and 31 ha from cropland, grassland, forestland, and wetland in Satame, respectively. Similar to this, its original coverage was changed to cropland, grassland, forestland, and settlement with an area of 5386, 172, 903, and 71 ha, respectively (Table 8). The majority of the 89.3% altered shrub cover (71.8%) was transformed into cropland, followed by the conversion of forestland and grassland, each at 12% and 2.3%, respectively. In the Satame watershed, the forest cover has been replaced by cropland (634 ha), shrublands (98 ha), and grassland (31ha) over the past 45 years. In Satame, grasslands of 351, shrublands of 321, and forestlands of 170 ha were transformed into cropland, whereas in Legabora watersheds, grassland, bare land, shrublands, and forestlands were transformed into cropland by 379, 238, 67, and 49 ha, respectively.

Table 8

illustrates the overall gain and losses in LULC classes in between 1976 and 2022
	LULC	WSD	LULC change from the initial year 1976 (ha)
	LULC	WSD	ST	BL	WB	CL	WL	GL	SL	FL	Row	Loss
LULC to final yr. 2022 (ha) LULC to final yr. 2022 (ha)	ST	Satame Watershed	0	0	0	111	0	63	71	14	260	260
	BL		0	0	1	8	0	4	1	0	14	14
	WB		0	0	542	18	3	64	47	14	688	146
	CL		0	44	40	4627	160	5491	5386	634	16742	12115
	WL		0	1	1	14	4	25	20	3	67	63
	GL		0	5	36	351	69	40	172	31	1063	1023
	SL		0	1	43	321	31	399	805	98	1698	893
	FL		0	1	76	170	22	619	903	288	2079	1791
	Column T		0	52	1099	5656	293	7179	7498	1094	0
	Gain		0	52	557	1029	289	7139	6693	806	–	–
	NC		–260	38	411	–11086	226	6116	5800	–985	–	–
	NP		0	0	1	–2.39	56.50	152.90	7.2	–3.42	–	–
	ST	Legabora watershed	0	4	0	4	0	103	137	8	256	256
	BL		0	1	9	238	0	131	6	0	385	384
	WB		0	0	300	0	0	0	0	0	300	0
	CL		0	195	216	6555	0	10436	5296	327	23025	16470
	WL		0	0	2	12	0	23	10	0	47	47
	GL		0	42	69	379	0	1993	628	9	3119	1126
	SL		0	6	17	67	0	652	683	37	1463	780
	FL		0	14	17	49	0	882	1186	311	2460	2149
	Column T		0	262	630	7304	0	14220	7946	692	0
	Gain		0	261	330	749	0	12227	7263	381	–	–
	NC		–256	–123	330	–15721	–47	11101	6483	–1768	–	–
	NP		–256	–124	1.1	–2.39	0	5.57	5800	–5.68	–	–

The conversion matrix analysis showed that there were also significant gains to the forest (806 and 382) from other LULC classes, but the gains from crop cover (634 and 327 ha) and shrublands (98 and 37 ha), respectively, were highest in Satame and Legabora. Table 8 shows the total increase and decrease in LULC classes between 1976 and 2022. In Satame, the analysis of LULC change over the final comparison period (1976–2022) revealed a rise in the area covered by the settlement, cropland, and forest cover, while it revealed a decline in the area covered by bare land, shrub cover, wetland, and grass cover (Table 6). For settlement, cropland, and forest cover during the same time period, the annual rate change (increment) was 5.7, 249.9, and 18.8 ha/year, respectively. The Legabora watershed showed a trend towards an increasing tendency for settlement, bare land, wetland, cropland, and forest cover between 1976 and 2022, but a trend towards a decreasing tendency for shrublands and grassland cover. In Satame, cropland had been gained from other LULC categories; the largest gain came from grassland, followed by shrub, forest, and wetland, with extents of 5491, 5386, 634, and 160 ha, respectively (Table 8). Between 1976 and 2022, crop cover in the Legabora watershed expanded by 10436, 5296, 327, 216, and 195 ha, respectively, from other LULC classes, which include grassland, shrublands, forestland, wetland, and bare cover.

Table 8 (Inserting near here)

Note

Those figures in bold are the sum of diagonals and represent overall persistence (i.e., the landscape that did not change). The NP symbolizes the net change to persistence ratio (i.e., the net change versus the diagonals for each class). Where net change, loss, gain, and Net persistence is calculated using the equations given below.

$$\text{N}\text{e}\text{t} \text{P}\text{e}\text{r}\text{s}\text{i}\text{s}\text{t}\text{e}\text{n}\text{c}\text{e}= \frac{\text{N}\text{e}\text{t} \text{c}\text{h}\text{a}\text{n}\text{g}\text{e} }{\text{D}\text{i}\text{a}\text{g}\text{o}\text{n}\text{a}\text{l}\text{s} \text{o}\text{f} \text{e}\text{a}\text{c}\text{h} \text{c}\text{l}\text{a}\text{s}\text{s}}$$

…………………………………...………………………….

$$Net change=Gain-Loss$$

…………………………………………………………………………….

$$Loss=Row total-Diagonals of each class \left(unchanged\right)$$

………………………………………

$$Gain=Column total-Diagonals of each class\left(unchanged\right)$$

……………………………………

Between 1976 and 2022, in the Legabora watershed, settlement experienced a larger negative net change to persistence ratio, while shrublands, grassland, and wetland showed positive ratios; moreover, forestland, bare land, and cropland had negative ratios as well, representing the dominant trends in the study area. The net change-to persistence ratio was closer to zero for the other land use and land cover types, indicating that they were more likely to persist rather than decrease or increase. These findings are summarized in Table 8. Across the entire landscape, a total of 6,306 ha (i.e., the sum of diagonal elements) did not change during the period covered by the study. The largest net increase in the Satame watershed occurred in the grassland cover class, which was mainly converted from cropland and forestland. Similarly, in Legabora, grassland saw the highest net increase, primarily from cropland and forestland. According to the results, between 1976 and 2022, the net change to persistence ratio was greater for forest (negative), grasslands (positive), cropland (negative), wetland (positive), and shrub land (positive) in the Satame watershed, respectively. The study also found that a total of 9843 ha of land remained unchanged throughout the entire Legabora watershed during the same period, as shown in Table 8.

LULC change and their drivers in Satame and Legabora watershed

We gathered data from KIIs and focus groups about the various factors that influence LULC changes as well as local residents' opinions of the LULC changes adopted by references (Mariye et al., 2022b). According to the summary results of the KIIs, field observations, preference ratings, and FGDs, LULC changes in the Satame and Legabora watersheds were found to be primarily caused by proximate and underlying drivers (Munthali et al., 2019). The informants identified a total of nine factors as the primary influences of LULC changes in the Satame and Legabora watersheds of south-central Ethiopia (Fig. 10). On the other hand, there were variations in each of the elements that the local community saw as being responsible for LULC changes (Angessa et al., 2019). Here, the primary proximate (direct) drivers of LULC change in the study site were identified as infrastructure development, settlement expansion, charcoal production, and agricultural expansions. However, the underlying (indirect) drivers for the observed LULC change in the study landscape were viewed as poverty, land-related policy, population growth, and community conflicts (Fig. 10). The majority of FGD participants identified population growth, agricultural expansion, settlement, charcoal production, and forest fire as the main proximate (direct) drivers of deforestation in the Satame and Legabora watersheds (Angessa et al., 2019). In order to expand cropland, the discussants and KIIs also identified fire as a crucial primary tool for cleaning forests (Betru et al., 2019). According to the LULC change analysis, crop land has grown significantly in the study area over the past 45 years (1976–2022). The major ranking order of the observed LULC change drivers are summarized in Fig. 10. The top five drivers of LULC changes in the Legabora watershed were perceived to be poverty, infrastructure development, agricultural expansion, and settlement expansion, with population growth coming in first and agricultural expansions second, respectively (Munthali et al., 2019). However, in the Satame watershed, agricultural expansion and population growth were ranked first and second, respectively, in terms of the main LULC drivers, followed by settlement expansion, wood extraction, and poverty (Fig. 10). Similar findings came up in KIIs and FGDs, where it was determined that population growth, settlement, and agricultural expansion were the main factors contributing to the LULC change in the study site. Key informants and FGDs agreed that early marriage, high birth rates, and low mortality were the main causes of the Satame and Legabora watersheds' rapid population growth.

Figure 10 (Inserting near hear)

Most local communities overwhelmingly believed that population growth accelerated during the study span. To corroborate local community population growth observations, the population data between 1981 and 2022 was estimated (Munthali et al., 2019). The populations of the Satame and Legabora watersheds were increased from 96,348 to 265,643 and from 82,143 to 295,643 (Fig. 11), respectively (CSA, 2007; KTARDO, 2019). Indeed, the population growth of the Satame and Legabora watersheds has been augmented by 275.7% and 359.8% between 1981 and 2022 (Munthali et al., 2019). The population disclosed an escalating tendency and became more uniform at the Legabora watershed during the entire study span (1981–2022). The growth was trustworthy and positively linked with cropland expansion between similar study spans (Fig. 11). Compared to the previous watershed, the population growth rate at the Satame watershed was more or less uneven between the similar study times. The population growth rate exhibited declining trends from 2001 to 2007, but after 2010, when cropland did not significantly change, the growth trend changed and revealed increasing trends (Fig. 11). Prior research also identified population growth as one of the main forces behind LULC changes (Bewket, 2002). Some research studies conducted in Ethiopia found population growth as major driver of LULC changes adopted by references Poverty is one of the significant observed factors that is causing LULC changes in the Satame and Legabora watersheds. Many watershed societies are unable to purchase agricultural inputs due to their extreme poverty and lack of financial resources. Because of their extreme poverty, watershed communities in south-central Ethiopia are also forced to clear forest land for additional farmland or to sustain their lives, as adopted by references (Munthali et al., 2019).

In the absence of well-organized economic policies, unsustainable use of natural resources such as land, water, and forests leads to severe environmental difficulties, including biodiversity loss, soil wearing away, deforestation, and forest cover degradation (Munthali et al., 2019). The findings of this study are consistent with other studies conducted in Africa that report high poverty as an underlying factor for LULC alterations (Haller et al., 2008; Kindu et al., 2015). Regarding land tenure policy, the military government of Ethiopia announced the nationalization of all rural land, abolishing private and public property rights, and granting all land use rights (Meshesha et al., 2014). During the FGDs, the elders explained that the land reform of 1975 changed the land system by nationalizing all rural land, turning "cultivators into landlords". This situation provides fertile ground for the expansion of cropland to other LULC types in the watershed (Angessa et al., 2019). This policy makes the state the sole owner of property and land-related resources, making it impossible for the state to monitor and enforce regulations. Key informants and FGDs discussants provided information that most land policies and strategies failed because they were not properly implemented and integrated (Fig. 10). Due to poor land management, forested areas have been converted to settlements and agricultural land and are severely degraded (Meshesha et al., 2014). Studies conducted by Bewket and Abebe (2013), Degife et al. (2018), documented the issue of land tenure as an important indirect driver of LULC changes in diverse highland watersheds in Ethiopia.

The focus group discussants and key informants remarked that conversion of various land use categories to cropland is a relatively easy process. For instance, in Ethiopia, farmers can enjoy more legal rights to land by converting forests to arable land since the law states that forests are state-owned. Discussions with local elders and natural resource management professionals revealed that several impoverished households in the study watershed rely heavily on the sale of firewood and charcoal for their survival tactics (Degife et al., 2018; Kindu et al., 2015). This dependence on woody plants is also seen in other rural areas of Ethiopia, where certain plants are cultivated for both domestic use and income generation. Similar observations have been made in other regions of the country, as documented in several studies (Degife et al., 2018; Kindu et al., 2015; Mezgebu & Workineh, 2017).

Figure 11 (Inserting near hear)

Consequences of LULC change perceived by local community

According to the discussants in the FGDs, the changes in land use and land cover (LULC) had various consequences, such as deforestation (94%, 82%), soil erosion and land degradation (92%, 76%), shortage of livestock fodder (90%, 90%), loss of biodiversity (84%, 96%), the decline in production (76%, 80%), decline in water resources (72%, 94%), and local climate change (58%, 62%) in the Legabora and Satame watersheds, respectively (Fig. 12). The current demographic changes and the continued variation in LULC, coupled with uncertain climate conditions, significantly affect livelihoods and put the farmer’s production system under pressure. Our filed observations, FGDs, and KIIs also marked that some native trees, which were used for house construction and are highly valuable medicinal plants, are in danger, like Prunes Africana, Podocarpus falcatus, and Cordia Africana. The study demonstrates that bushlands and shrublands have been extensively degraded and replaced by settlements and croplands over the past four and a half decades. This implies that there is a possibility that both flora and fauna may be at risk, but we were unable to estimate their extent. A decrease in plant biodiversity contributes to deterioration in the sustainability of the environment and the loss of the genetic resources of plants. As a result, economic and medicinal values will be lost (Bewket, 2002). A significant majority of survey respondents in Legabora and Satame, specifically (84%, 96%), reported a decline in plant biodiversity, which included the loss of valuable woody species and medicinal plants, which was consistent with a study done by Belete et al. (2021).

Furthermore, 94% of FGDs discussants indicated that the study area experienced a shift in household energy use from fuel wood to crop residue, cow dung, and eucalyptus plantations due to deforestation (Fig. 12). Prior to 25 years ago, 94 percent of households used natural forests and shrublands as a source of energy, but today only a small number of them have access to these areas to gather wood for cooking and heating. Female participants in the focus group discussions explained that in the past, when the women collected fuel wood from the natural forest, they had opportunities for social interaction, informal communication with their friends, and discussion of social issues. Since they no longer have this means of communication, 94% of respondents now rate deforestation as a disadvantage (Fig. 12). In the interview sessions, 72% of survey respondents indicated that the nearby water sources have dried up, and climate change is blamed for the drying of small brooks and springs. Women are forced to travel long distances to fetch water for household consumption (Fig. 12). To get to water sources for drinking, male farmers in the present day must transport their livestock over long distances. The major LULC changes observed at the Satame and Legabora watersheds were the remarkable expansion of cropland at the expense of existing natural resources, such as grassland and shrubs (Table 8). In the Satame and Legabora watersheds, a significant amount of grassland and shrublands has been converted to cropland (5491, 538 ha of grassland and 10436, 5296 ha of shrublands converted to cropland between 1976 and 2022) (Table 8). This will have a negative impact on the current free grazing system, and, ultimately, the community will have to reduce its livestock population, which will have a detrimental effect on household income and the consumption of livestock products.

Figure 12 (Inserting near here)

According to data from field research and KIIs, the public grassland decreased as agriculture was expanded on communal grazing lands. As a result, a small plot of land is frequently used to graze an excessive number of livestock, which may contribute to environmental degradation. A similar effect was reported by Desalegn et al. (2014), in the highlands of Ethiopia, where they noted that the lack of livestock fodder posed a significant problem. In the study area, 92% of the survey respondents stated that the local area was exposed to high soil erosion and degradation (Fig. 12). On damaged soils, plant growth is poor, which intensifies the issue of eroding soil. The naturally occurring vegetation that covered the study watershed has been severely reduced, and it has been replaced by settlements and farmland. Consequently, the area of land that is more susceptible to soil erosion has grown over time, but this study was unable to quantify it. FGDs and key informants stated that soil wearing away and the associated soil fertility problem were significant consequences for the decline of agricultural production, mainly in the upper part of the Satame and Legabora watersheds respectively (Mariye et al., 2022a). FGDs and KIIs also marked that the area was primarily comprised of forest, grasslands, and that many different wild animals, such as tigers, elephants, warthogs, and buffalo. However, it is now difficult to find animals like buffalo because the forests have been cleared and the rate of deforestation has increased. During the FGDs, 76% of respondents explained that in the past, farmers had produced adequate yields. However, they are currently losing some of their farms as the land becomes less fertile and they do not get enough yields. Moreover, 76% of survey respondents said that the agricultural sector is less profitable because of the area's high exposure to soil erosion and land degradation (Fig. 12).

An investigation of this study has to quantify the dynamics and their drivers within the Satame and Legabora sub-watersheds. The study shows that substantial spatiotemporal LULC changes have occurred in the Satame (midland) and Legabora (highlands) watersheds over the past four and a half decades. Cropland increased significantly in Satame and Legabora between 1991 and 2001, largely at the expense of grassland and shrublands. The Legabora watershed showed a trend towards an increasing tendency for settlement, bare land, wetland, cropland, and forest cover between 1976 and 2022, but a decreasing tendency for shrublands and grassland cover. This study found that grassland and shrublands were the predominant LULC classes during the previous decade but are now preceded by croplands. From 1976 to 2022, shrub land, wetlands, grasslands, and bare land declined annually at a rate of –128.7, –4.9, –139.9, and –0.83 ha/year in Satame, while grassland and shrub cover decreased annually by –248.2, –144.5 ha/year in Legabora. The local communities perceived agricultural expansion, population growth, settlement expansion, and poverty as the main drivers of LULC changes, while deforestation, land degradation, low production, and loss of biodiversity were the major consequences. About 68.3% of the land area of Legabora and 72.3% of the Satame watershed experienced significant LULC changes during the study span. The drivers and consequences highlighted in this research can prove to be a crucial tool for land use planning and future LULC modifications in the study area and beyond. These findings can be used to develop effective land management plans, policies, and strategies for decision-making in similar areas. To address the swift rates of land use and land cover (LULC) alterations in nearby and far-off downstream systems, as well as other watersheds with comparable socio-ecological contexts, it is imperative to implement comprehensive watershed management tactics. To this end, the following suggestions are proposed: (1) Foster the preservation and reestablishment of natural vegetation and areas that are environmentally sensitive, with the aim of benefiting both upstream and downstream communities within the watershed. (2) Develop methods to manage the rapid growth of the population in the research landscape. (3) Adopting effective farming practices, recovering the degraded lands, and conserving biodiversity and natural resources to address undesirable LULC alterations in the watershed. (4) Future studies should examine how LULC modifications affect rural livelihoods and local climate in the study watershed. (5) Land use planning and landscape conservation approaches should be based on scientific knowledge.

Acknowledgment

We stretch out our appreciation to the KTz district Agricultural and Rural Development Office, farmers, and development agents particularly from Doyogena, Angecha, Damiboya, Kacha Bira and Kedida Gamella district who gave us pertinent data on the LULC management scheme of the study area and its adjacent landscape. We also recognize the Ethiopian Meteorological Services and Ethiopian Space Science and Geospatial Information Institute for providing us data on precipitation, satellite information, and historical land use map, respectively.

Author’s contribution: Conceptualization: Melesse Maryo, Data curation: Mehari Mariye, Formal analysis: Mehari Mariye, Investigation: Mehari Mariye, Methodology: Mehari Mariye and Gedion Hailu, Software: Mehari Mariye and Belayneh Yigez, Supervision: Melesse Maryo and Li Jianhua, Validation: Melesse Maryo and Li Jianhua, Writing original draft: Mehari Mariye, Writing review & editing: Mehari Mariye, Melesse Maryo, Li Jianhua, Gedion Hailu, and Belayneh Yigez. All authors reviewed the results and approved the final version of the manuscript. The author declares responsibility for conceiving and designing the study, collecting data, analyzing and interpreting the results, and drafting the manuscript.

Availability of data and materials

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials. The four Landsat images were downloaded from U.S Geological Survey (USGS) Center for Earth Resources Observation and Science (EROS). Digital Elevation Map (DEM), river and road data were collected from Ministry of Water and Energy and space science and geospatial institute of Ethiopia. In addition, primary data were also obtained through extensive field works and in-depth focus group discussions with agricultural development agents and local communities. Moreover, upon a reasonable request, the corresponding author provides access to the datasets produced and/or used in the current study.

- Conflict of Interest Statement: In this statement, we declare that there are no any potential or existing conflicts of interest that might have influenced the study, results, or interpretations detailed in our paper.

- Ethics Statement: In this statement we hereby state that this work has not been published or accepted for publication, and is not under consideration for publication in any journal or book. Each author will be held jointly and individually responsible for the content of the manuscript since they all contributed substantially to its development. All authors have given their consent for it to be submitted for publication. I also confirmed that there are no other parties currently claiming copyright ownership over the work and data in the manuscript.

Finding: This research has been supported by Tongji University, China [Yangtze River Water Environment project]

Abate, S. (2011). Evaluating the land use and land cover dynamics in borena woreda of south wollo highlands, ethiopia. Journal of Sustainable Development in Africa, 13(1), 87–107.
Agidew, A.-m. A., & Singh, K. N. (2017). The implications of land use and land cover changes for rural household food insecurity in the Northeastern highlands of Ethiopia: the case of the Teleyayen sub-watershed. Agriculture & Food Security, 6(1). https://doi.org/10.1186/s40066-017-0134-4
Amare, S. (2015). Land use/cover change at infraz watershed, northwestern Ethiopia. Journal of Landscape Ecology, 8(1).
Anchan, S. S., AShetty, t., Bhat, G., H, & Chadaga, M. (2018). Land use and land cover change detection through spatial approach: A case study of Mangaluru Taluk, Karnataka. Journal of Geomatics, 12(2).
Andualem, T. G., Belay, G., & Guadie, A. (2018). Land Use Change Detection Using Remote Sensing Technology Journal of Earth Science & Climatic Change, 9(10), 1000496. https://doi.org/10.4172/2157-7617.1000496
Angessa, A. T., Lemma, B., & Yeshitel, K. (2019). Land-use and land-cover dynamics and their drivers in the central highlands of Ethiopia with special reference to the Lake Wanchi watershed. GeoJournal. https://doi.org/10.1007/s10708-019-10130-1
Ariti, A. T., van Vliet, J., & Verburg, P. H. (2015). Land-use and land-cover changes in the Central Rift Valley of Ethiopia: Assessment of perception and adaptation of stakeholders. Applied Geography, 65, 28–37. https://doi.org/10.1016/j.apgeog.2015.10.002
Asokan, A., & Anitha, J. (2019). Change detection techniques for remote sensing applications: a survey. Earth Science Informatics, 12(2), 143–160. https://doi.org/10.1007/s12145-019-00380-5
Ayalew, A. D., Wagner, P. D., Sahlu, D., & Fohrer, N. (2022). Land use change and climate dynamics in the Rift Valley Lake Basin, Ethiopia. Environ Monit Assess, 194(11). https://doi.org/10.1007/s10661-022-10393-1
Barana, B., Senbetie, T., & Aklilu, B. (2016). Land use/Land Cover Dynamics and its Implication on Sustainable Land Management in Wallecha Watershed, Southern Ethiopia Land use/Land Cover Dynamics and its Implication on Sustain. Global Journal of Science Frontier Research, 16(4), 49–63.
Bekele, D., Alamirew, T., Kebede, A., Zeleke, G., & Melesse, A. M. (2018). Land use and land cover dynamics in the Keleta watershed, Awash River basin, Ethiopia. Environmental Hazards, 18(3), 246–265. https://doi.org/10.1080/17477891.2018.1561407
Belay, T., & Mengistu, D. A. (2019). Land use and land cover dynamics and drivers in the Muga watershed, Upper Blue Nile basin, Ethiopia. Remote Sensing Applications: Society and Environment, 15. https://doi.org/10.1016/j.rsase.2019.100249
Belayneh, B., & Eyasu, E. (2021). Land Use/Land Cover Change and Its Driving Forces in Shenkolla Watershed, South Central Ethiopia. The Scientific World Journal, 1. https://doi.org/10.1155/2021/9470918
Belayneh, Y., Ru, G., Guadie, A., Teffera, Z. L., & Tsega, M., J. For. Res. (2018). Forest cover change and its driving forces in FagitaLekoma District, Ethiopia. Journal forest resources
–16.
Belete, F., Maryo, M., & Teka, A. (2021). Land use/land cover dynamics and perception of the local communities in Bita district, south western Ethiopia. International Journal of River Basin Management, 1–12. https://doi.org/10.1080/15715124.2021.1938092
Berihun, M. L., Tsunekawa, A., Haregeweyn, N., Meshesha, D. T., Adgo, E., Tsubo, M.,.. . Yibeltal, M. (2019). Exploring land use/land cover changes, drivers and their implications in contrasting agro-ecological environments of Ethiopia. Land Use Policy, 87, 104052. https://doi.org/10.1016/j.landusepol.2019.104052
Betru, T., Tolera, M., Sahle, K., & Kassa, H. (2019). Trends and drivers of land use/land cover change in Western Ethiopia. Applied Geography, 104, 83–93. https://doi.org/10.1016/j.apgeog.2019.02.007
Bewket, W. (2002). Land Cover Dynamics Since the 1950s in Chemoga Watershed, Blue Nile Basin, Ethiopia. Mountain Research and Development, 22(3), 263–269. https://doi.org/10.1659/02764741(2002)022[0263:lcdsti]2.0.co;2
Bewket, W., & Abebe, S. (2013). Land-use and land-cover change and its environmental implications in a tropical highland watershed, Ethiopia. International Journal of Environmental Studies, 70(1), 126–139. https://doi.org/10.1080/00207233.2012.755765
Braimoh, A. K. (2006). Random and systematic land-cover transitions in northern Ghana. Agriculture, Ecosystems & Environment, 113(1–4), 254–263. https://doi.org/10.1016/j.agee.2005.10.019
Congalton, R. G. (1991). A Review of Assessing the Accuracy of Classifications of Remotely Sensed Data. REMOTE SENS. ENVIRON., 37, 35–46.
Congalton, R. G. (2001). Accuracy assessment and validation of remotely sensed and other spatial information. International Journal of Wildland Fire, 10(4). https://doi.org/10.1071/wf01031
Congalton, R. G., & Green, K. (2009). Assessing the accuracy of remotely sensed data: principles and practices (2nd ed.).. Boca Raton: Taylor & Francis.
CSA. (2007). Summary and Statistical Report of the 2007 Population and Housing Census Results, Addis Ababa, Ethiopia. 1-113.
Danano, K. A., Legesse, A., & Likisa, D. (2018). Monitoring Deforestation in South Western Ethiopia Using Geospatial Technologies. Journal of Remote Sensing & GIS, 07(01). https://doi.org/10.4172/2469-4134.1000229
Degife, A. W., Zabel, F., & Mauser, W. (2018). Assessing land use and land cover changes and agricultural farmland expansions in Gambella Region, Ethiopia, using Landsat 5 and Sentinel 2a multispectral data. Heliyon, 4(11), e00919. https://doi.org/10.1016/j.heliyon.2018.e00919
Demissie, F., Yeshitila, K., Kindu, M., & Schneider, T. (2017). Land use/Land cover changes and their causes in Libokemkem District of South Gonder, Ethiopia. Remote Sensing Applications: Society and Environment, 8, 224–230. https://doi.org/10.1016/j.rsase.2017.10.001
Desalegn, T., Cruz, F., Kindu, M., Turrión, M. B., & Gonzalo, J. (2014). Land-use/land-cover (LULC) change and socioeconomic conditions of local community in the central highlands of Ethiopia. International Journal of Sustainable Development & World Ecology, 21(5), 406–413. https://doi.org/10.1080/13504509.2014.961181
Desta, H., & Fetene, A. (2020). Land-use and land-cover change in Lake Ziway watershed of the Ethiopian Central Rift Valley Region and its environmental impacts. Land Use Policy, 96, 104682. https://doi.org/10.1016/j.landusepol.2020.104682
Dwivedi, R. S., Sreenivas, K., & Ramana, K. V. (2007). Cover: Land-use land‐cover change analysis in part of Ethiopia using Landsat Thematic Mapper data. International Journal of Remote Sensing, 26(7), 1285–1287.
Elias, E., Seifu, W., Tesfaye, B., & Girmay, W. (2019). Impact of land use/cover changes on lake ecosystem of Ethiopia central rift valley. Cogent Food & Agriculture, 5, 1–20. https://doi.org/10.1080/23311932.2019.1595876
Fentahun, T., & Gashaw, T. (2014). Evaluation of land use/land cover changes of Bantneka Watershed, Ethiopia. Civil and Environmental Research, 6(7).
Foley, J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R.,.. . Snyder, P. K. (2005). Global Consequences of Land Use. SCIENCE, 309, 570–574.
Garedew, E., Sandewall, M., Söderberg, U., & Campbell, B. M. (2009). Land-Use and Land-Cover Dynamics in the Central Rift Valley of Ethiopia. Environmental Management, 44(4), 683–694. https://doi.org/10.1007/s00267-009-9355-z
Gashaw, T., Tulu, T., & Argaw, M. (2017). Erosion risk assessment for prioritization of conservation measures in Geleda watershed, Blue Nile basin, Ethiopia. Environmental Systems Research, 6(1). https://doi.org/10.1186/s40068-016-0078-x
Gebrelibanos, T., & Assen, M. (2013). Land use/land cover dynamics and their driving forces in the Hirmi watershed and its adjacent agro-ecosystem, highlands of Northern Ethiopia. Journal of Land Use Science, 10(1), 81–94. https://doi.org/10.1080/1747423x.2013.845614
Gebreslassie, H. (2014). Land Use-Land Cover dynamics of Huluka watershed, Central Rift Valley, Ethiopia. International Soil and Water Conservation Research, 2(4), 25–33. https://doi.org/10.1016/s2095-6339(15)30055-1
Hailu, A., Mammo, S., & Kidane, M. (2020). Dynamics of land use, land cover change trend and its drivers in Jimma Geneti District, Western Ethiopia. Land Use Policy, 99, 105011. https://doi.org/10.1016/j.landusepol.2020.105011
Haller, T., Galvin, M., Meroka, P., Alka, J., & Alvarez, A. (2008). Who Gains from Community Conservation? Intended and Unintended Costs and Benefits of Participative Approaches in Peru and Tanzania.. Journal of Environment & Developme, 17(2), 118–144.
Haregeweyn, N., Tesfaye, S., Tsunekawa, A., Tsubo, M., Meshesha, D. T., Adgo, E., & Elias, A. (2015). Dynamics of land use and land cover and its effects on hydrologic responses: case study of the Gilgel Tekeze catchment in the highlands of Northern Ethiopia. Environ Monit Assess, 187(1), 4090. https://doi.org/10.1007/s10661-014-4090-1
Hassan, Z., Shabbir, R., Ahmad, S. S., Malik, A. H., Aziz, N., Butt, A., & Erum, S. (2016). Dynamics of land use and land cover change (LULCC) using geospatial techniques: a case study of Islamabad Pakistan. SpringerPlus, 5(812), 1–12. https://doi.org/10.1186/s40064-016-2414-z
Hassen, E. E., & Assen, M. (2017). Land use/cover dynamics and its drivers in Gelda catchment, Lake Tana watershed, Ethiopia. Environmental Systems Research, 6(1). https://doi.org/10.1186/s40068-017-0081-x
Houghton, R. A., House, J. I., Pongratz, J., Van Der Werf, G. R., DeFries, R. S., Hansen, M. C., & Ramankutty, N. (2012). Carbon emissions from land use and land-cover change.. Biogeosciences, 9,(12), 5125–5142. https://doi.org/doi:10.5194/bg-9-5125-2012
Hurni, H., Wolde-Aregay, B., Chadhokar, P., Goshu, K., Zeleke, G., Danano, D., & Grunder, M. (2016). Soil and Water Conservation in Ethiopia: Guidlines for Development Agents. Centre for Development and Environmen(CDE): Bern Open Publishing (BOP), Bern, Switzerland..
Kidane, Y., Stahlmann, R., & Beierkuhnlein, C. (2012). Vegetation dynamics, and land use and land cover change in the Bale Mountains, Ethiopia. Environ Monit Assess, 184(12), 7473–7489. https://doi.org/10.1007/s10661-011-2514-8
Kindu, M., Schneider, T., Teketay, D., & Knoke, T. (2013). Land Use/Land Cover Change Analysis Using Object-Based Classification Approach in Munessa-Shashemene Landscape of the Ethiopian Highlands. Remote Sensing, 5(5), 2411–2435. https://doi.org/10.3390/rs5052411
Kindu, M., Schneider, T., Teketay, D., & Knoke, T. (2015). Drivers of land use/land cover changes in Munessa-Shashemene landscape of the south-central highlands of Ethiopia. Environ Monit Assess, 187(145), 1–17. https://doi.org/10.1007/s10661-015-4671-7
Kløve, B., Ala-Aho, P., Bertrand, G., Gurdak, J. J., Kupfersberger, H., Kværner, J.,.. . Pulido-Velazquez, M. (2014). Climate change impacts on groundwater and dependent ecosystems. Journal of Hydrology, 518, 250–266. https://doi.org/10.1016/j.jhydrol.2013.06.037
Kothari, C. R. (2004). Research Methodology Methods & Techniques, 2nd edition. New Age International (P) Limited, New Delhi, India.
KTARDO. (2019). Kembata Tembaro District Agriculture and Rural Development Office (KTARDO). Information on SocioEconomic and Geo-Spatial Annual Statistical Abstract. south-central Ethiopia.
Lambin, & Geist. (2006). Land-use and Land-cover Change: Local Processes and Global Impacts. Berlin.Land-use and Land-cover Change: Local Processes and Global Impacts. Berlin.
Lambin, E. F., Geist, H. J., & Lepers, E. (2003). Dynamics of Land-use and land-cover change in tropical regions. Annu. Rev. Environ. Resour, 28, 205–241. https://doi.org/doi: 10.1146/annurev.energy.28.050302.105459
Lillesand, T. M., Kiefer, R. W., & Chipman, J. W. (2008). Remote Sensing and Image Interpretation. 6th Edition, John Wiley & Sons, Hoboken..
Mammo, s., Kebin, Z., & Kidane, m. (2018). Retrospective analysis of land use land cover dynamics using GIs and remote sensing in central highlands of Ethiopia. Journal of Landscape Ecology, 11(2), 0005. https://doi.org/10.2478/jlecol-2018-0005
Mariye, M., Jianhua, L., & Maryo, M. (2022a). Land use and land cover change, and analysis of its drivers in Ojoje watershed, Southern Ethiopia. Heliyon, 8(4), e09267. https://doi.org/10.1016/j.heliyon.2022.e09267
Mariye, M., Jianhua, L., & Maryo, M. (2022b). Land use land cover change analysis and detection of its drivers using geospatial techniques: a case of south-central Ethiopia. All Earth, 34(1), 309–332. https://doi.org/10.1080/27669645.2022.2139023
Maryo, M. (2020). Agroforestry practices and indigenous tree use in enset- based farming systems of Doyogana Woreda, Southern Ethiopia. Unpublished paper Agriculture & Food Security.
Mehari., M., Maryo., M., & Li, J. (2022). The Study of Land Use and Land Cover (LULC) Dynamics and the Perception of Local People in Aykoleba, Northern Ethiopia. Journal of the Indian Society of Remote Sensing, 50(5), 775–789. https://doi.org/10.1007/s12524-021-01462-y
Mengistu, D. A., Waktola, D. K., & Woldetsadik, M. (2012). Detection and analysis of land-use and land-cover changes in the Midwest escarpment of the Ethiopian Rift Valley. Journal of Land Use Science, 7(3), 239–260. https://doi.org/10.1080/1747423x.2011.562556
Meshesha, D. T., Tsunekawa, A., Tsubo, M., Ali, S. A., & Haregeweyn, N. (2014). Land-use change and its socio-environmental impact in Eastern Ethiopia’s highland. Regional Environmental Change, 14(2), 757–768.
Mezgebu, A., & Workineh, G. (2017). Changes and drivers of afro-alpine forest ecosystem: future trajectories and management strategies in Bale eco-region, Ethiopia. Ecological Processes, 6(1). https://doi.org/10.1186/s13717-017-0108-2
Miheretu, B. A., & Yimer, A. A. (2017). Land use land cover changes and their environmental implications in the Gelana sub-watershed of Northern highlands of Ethiopia. Environmental Systems Research, 6(7). https://doi.org/10.1186/s40068-017-0084-7
Mikias, B. M. (2015). Land Use/Land Cover Dynamics in the Central Rift Valley Region of Ethiopia: Case of Arsi Negele District. African Journal of Agricultural Research, 10(5), 434–449. https://doi.org/10.5897/ajar2014.8728
Munthali, M. G., Davis, N., Adeola, A. M., Botai, J. O., Kamwi, J. M., Chisale, H. L. W., & Orimoogunje, O. O. I. (2019). Local Perception of Drivers of Land-Use and Land-Cover Change Dynamics across Dedza District, Central Malawi Region. sustainability, 11(3), 832. https://doi.org/10.3390/su11030832
Muriithi, F. K. (2016). Land use and land cover (LULC) changes in semi-arid sub-watersheds of Laikipia and Athi River basins, Kenya, as influenced by expanding intensive commercial horticulture. Remote Sensing Applications: Society and Environment, 3, 73–88. https://doi.org/10.1016/j.rsase.2016.01.002
Oettera, D. R., Cohenb, W. B., Berterretchea, M., Maierspergera, T. K., & Kennedya, R. E. (2000). Land cover mapping in an agricultural setting using multiseasonal Thematic Mapper data. Remote Sensing of Environment, 76, 139–155.
Okeke, F., & Karnieli, A. (2007). Methods for fuzzy classification and accuracy assessment of historical aerial photographs for vegetation change analyses. Part I: Algorithm development. International Journal of Remote Sensing, 27(1), 153–176. https://doi.org/10.1080/01431160500166540
Othow, O., Obang, Legesse Gebre, S., & Obsi Gemeda, D. (2017). Analyzing the Rate of Land Use and Land Cover Change and Determining the Causes of Forest Cover Change in Gog District, Gambella Regional State, Ethiopia. Journal of Remote Sensing & GIS, 06(04). https://doi.org/10.4172/2469-4134.1000219
Phalan, B., Onial, M., Balmford, A., & Green, R. E. (2011). Reconciling Food Production and Biodiversity Conservation: Land Sharing and Land Sparing Compared. Science 333, 1289–1291. https://doi.org/10.1126/science.1208742
Pontius, R. G., Shusas, E., & McEachern, M. (2004). Detecting important categorical land changes while accounting for persistence. Agriculture, Ecosystems & Environment, 101(2–3), 251–268. https://doi.org/10.1016/j.agee.2003.09.008
Ramankutty, N., & Foley, J. A. (1999). Estimating historical changes in global land cover: Croplands from 1700 to 1992. Global Biogeochemical Cycles, 13(4), 997–1027. https://doi.org/10.1029/1999gb900046
Shiferaw, A., & Singh, K. L. (2011). Evaluating The Land Use And Land Cover Dynamics In Borena Woreda South Wollo Highlands, Ethiopia. Ethiopian Journal of Business and Economics, 2(1), 69–104.
Siraj, M., Zhang, K., & Moges, K. (2018). Retrospective analysis of land use land cover dynamics using gis and remote sensing in central highlands of ethiopia. Journal of Landscape Ecology, 11(2).
Tefera, M. M. (2011). LAND-USE/LAND-COVER DYNAMICS IN NONNO DISTRICT, CENTRAL ETHIOPIA. Journal of Sustainable Development in Africa, 13(1), 123–141.
TekleL, K., & Hedlund, L. (2000). Land Cover Changes Between 1958 and 1986 in Kalu District, Southern Wello, Ethiopia. Mountain Research and Development, 20(1), 42–51.
Thakkar, A. K., Desai, V. R., Patel, A., & Potdar, M. B. (2017). Post-classification corrections in improving the classification of Land Use/Land Cover of arid region using RS and GIS: The case of Arjuni watershed, Gujarat, India The Egyptian Journal of Remote Sensing and Space Sciences, 20, 79–89. https://doi.org/10.1016/j.ejrs.2016.11.006
Tilahun, A. (2015). Accuracy Assessment of Land Use Land Cover Classification using Google Earth. American Journal of Environmental Protection, 4(4), 193. https://doi.org/10.11648/j.ajep.20150404.14
USGS. (2019). Landsat collection level 1 product definition and department of the Interior U.S. Geological Survey Sioux Falls, South Dakota, USA,. 1–26.
Wang, Y., Hu, Y., Niu, X., Yan, H., & Zhen, L. (2023). Myanmar’s Land Cover Change and Its Driving Factors during 2000–2020. International Journal of Environmental Research and Public Health, 20(3), 2409. https://doi.org/10.3390/ijerph20032409
Warner, K., Hamza, M., Oliver-Smith, A., Renaud, F., & Julca, A. (2009). Climate change, environmental degradation and migration. Natural Hazards, 55(3), 689–715. https://doi.org/10.1007/s11069-009-9419-7
Wolancho, K. W. (2015). Evaluating watershed management activities of campaign work in Southern nations, nationalities and peoples’ regional state of Ethiopia. 4, 6, 2–13.
WoldeYohannes, A., Cotter, M., Kelboro, G., & Dessalegn, W. (2018). Land Use and Land Cover Changes and Their Effects on the Landscape of Abaya-Chamo Basin, Southern Ethiopia. Land, 7(1). https://doi.org/10.3390/land7010002
Wubie, M. A., Assen, M., & Nicolau, M. (2016). Patterns, causes and consequences of land use/cover dynamics in the Gumara watershed of lake Tana basin, Northwestern Ethiopia. Environmental Systems Research, 5, 1–12.
Yohannes, H., Soromessa, T., Argaw, M., & Dewan, A. (2020). Changes in landscape composition and configuration in the Beressa watershed, Blue Nile basin of Ethiopian Highlands: historical and future exploration. Heliyon, 6(9), e04859. https://doi.org/10.1016/j.heliyon.2020.e04859

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Investigation of land use and land cover dynamics, drivers, and their consequences in south-central Ethiopia, with special emphasis on the contrasting agro-ecological settings of the Satame and Legabora watersheds

Status:

Version 1

Abstract

Figures

Introduction

Methods and Materials

Description of the study areas

Data sources and analysis methods

Socio-economic data

Result and Discussion

Accuracy assessment for image classification

Spatio-temporal distributions of LULC change within the Satame and Legabora watersheds

Trends and dynamics of LULC change in Satame and Legabora from 1976 to 2022

State of cropland LULC from 1976 to 2022

State of wetland LULC between 1976 and 2022

State of bare land LULC from 1976 to 2022

Status of forestland LULC from 1976 to 2022

State of Grassland LULC between 1976 and 2022

State of shrub LULC from 1976 to 2022

State of settlement LULC from 1976 to 2022

LULC change matrix in the Satame and Legabora watersheds for five investigation periods

LULC change and their drivers in Satame and Legabora watershed

Consequences of LULC change perceived by local community

Conclusion and recommendations

Statements and Declarations

References

Additional Declarations

Status:

Version 1