Research Paper |
Corresponding author: Ranmi Elsa Denise Ayeko ( ayekodenise@gmail.com ) Academic editor: Reginald Tang Guuroh
© 2023 Ranmi Elsa Denise Ayeko, Sêwanoudé Scholastique Mireille Toyi, Achille Ephrem Assogbadjo, Brice Augustin Sinsin.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Ayeko RED, Toyi SSM, Assogbadjo AE, Sinsin BA (2023) Dynamics of inselberg landscapes and their adjacent areas in the Sudano-Guinean zone of Benin through remote sensing analysis. Vegetation Classification and Survey 4: 189-202. https://doi.org/10.3897/VCS.89746
|
Aims: Land cover change in inselbergs and adjacent areas was studied from 2003 to 2018 in a region facing anthropogenic pressures to assess dynamics and preserve rare endemic species. Study area: Inselbergs and their adjacent areas in the Sudano-Guinean zone of Benin are included in this study. Methods: Land cover classes of inselbergs and adjacent areas were obtained through supervised classification of Sentinel-2 (2018) and Spot 5 (2003) satellite images. A Chi-square test was used to compare protected and unprotected LULC classes of inselbergs, with 10 m spatial resolution. Results: The results showed that forest and woodland decreased respectively from 8.55% to 3.05% and from 17.63% to 4.79% between 2003 and 2018 while tree and shrub savanna, and grassland increased respectively from 6.52% to 9.49% and from 7.60% to 16.69%. Field and fallow increased from 5.57% in 2003 to 26.12% in 2018 and tree plantation from 6.05% to 13.47%. The analysis of spatial comparisons using the chi-square test showed that the presence of inselbergs in a protected area has no significant effect on their land use. Conclusions: Natural vegetation in inselbergs and adjacent areas is being converted into human-made landscapes by farmers. An urgent conservation plan is needed, including awareness campaigns, tree planting, and sustainable forest management.
Taxonomic reference:
Abbreviations: DEM = Digital Elevation Model; GCP = Ground Control Point; LULC = Land Use/Land Cover; ROI = Region of Interest; SRTM = Satellite imagery data, Shuttle radar topography mission.
anthropogenic pressure, dynamic trend, inselberg, land use/land cover, protected and unprotected inselberg, Sentinel-2, Spot 5
Terrestrial ecosystems are essential for human well-being and global survival because they provide a variety of ecosystem services, including food production, air and water purification, climate regulation, crop pollination and biodiversity preservation (
In 1900, the German geologist Bomhardt introduced the term ‘inselberg’ to describe granitic or gneissic rocky outcrops that rise above the peneplain of tropical and subtropical regions (
However, over the past two decades, there has been increasing interest in inselberg ecosystems worldwide (
To effectively address the issue of inselberg destruction, it is crucial for each country to collect and share information on the status of inselbergs within their own territory. Without reliable data, it is difficult to develop targeted conservation strategies or to monitor progress in protecting these unique habitats. Therefore, it is incumbent upon each country to take responsibility for assessing the condition of their inselbergs. Unfortunately, in Benin, the inselbergs are mainly located in the Sudanian and Sudano-Guinean zones (
Monitoring the change of land use over time around inselbergs is, therefore, a very important way to effectively assess the trends of the dynamics. To this end, the diachronic analysis of land use that allows the showing of the spatial distribution of land use changes is the best way to achieve this.
However, the presence of protected areas such as the classified forests of Dassa-Zounmé and Savalou that contain parts of the inselbergs, offer an opportunity to preserve the inselbergs from human disturbance. Indeed, protected areas are created so that they can play an essential role in protecting representative samples of living organisms, remarkable geological phenomena or particular landscapes in terrestrial and marine environments from disappearing. Evidence from remote sensing suggests that protected areas slow down the rate of change from ‘natural’ to ‘human-modified’ land cover (
The aim of this study is to analyze the land use/land cover change of inselbergs and their adjacent areas located in the Sudano-Guinean zone of Benin between 2003 and 2018, and to assess the effectiveness of protected areas in preserving inselbergs. In addition, the study seeks to test the hypothesis that natural vegetation has been replaced by anthropogenic landscape features and that inselbergs located in protected areas are under less pressure compared to those outside protected areas.
The inselbergs and their adjacent areas within the department of “Collines”, one of the twelve administrative subdivisions located in the central part of Benin, were considered for this research. The Collines department is located between 7°27’ and 8°46’ North latitude and between 1°39 ‘and 2°44’ East longitude. It covers an area of 13,931 km² or more than 12% of the national territory (
Map of study area. The upper-left map displays the country, with the Collines department highlighted in green to indicate its location. The intermediate map focuses on the Collines department and highlights in white the three municipalities that were studied within this department. The Study Area map provides a closer look at these three municipalities, illustrating their inselbergs, protected areas, and GCPs. GCPs = Ground Control Points.
This work used three datasets: satellite imagery data, Shuttle radar topography mission (SRTM), digital elevation model (DEM), and ground control points (GCPs).
To monitor LULC changes in inselbergs, four Sentinel-2 MSI (multi-spectral instrument, Level-1C) satellite images with 10 m spatial resolution from 4 Jan 2018 covering the study area, were used. Indeed, studies conducted by
Shuttle radar topography mission (SRTM) digital elevation model (DEM) with a resolution of 1 arc-second (approximately 30 m) was used to extract the elevation and slope bands.
A dataset of 110 GCPs was constructed to train the classification algorithms applied to the Sentinel-2 data. Each GCP contains latitude, longitude, and the corresponding observed LULC type (Suppl. material
The digital interpretation of satellite images used the supervised classification method with Envi4.7 software (Exelis Visual Information Solutions, Boulder, Colorado). Three bands (Near-infrared (NIR), Red, Green) of multispectral Sentinel-2B and Spot 5 satellite images with a spatial resolution of 10 m were used for the land-use mapping. After choosing our 10 m resolution band, layer stacking was done by combining three bands into one multispectral image. Then ROIs (Regions of Interest) were created by selecting portions of the images graphically or by thresholding. The regions can be irregularly-shaped and were used to extract statistics for classification (
Following the work of
The SRTM image for the communes of Dassa, Savè, and Savalou was analyzed using ArcGIS 10.4 software. The image was divided into two classes based on elevation. The first class took into account the elevations between 16 and 180 m a.s.l. and the second class, the elevations between 180 and 550 m a.s.l. This was done because the lowest elevation of the inselbergs (a type of rocky hill) was found to be 180 meters during field verification.
The slope of the inselbergs is connected to them by a concave zone whose slopes can exceed 10%, according to studies conducted by
Using these two characteristics (elevation and slope), the inselbergs and their adjacent area were identified and then digitalized. The shapefile of the inselbergs including their adjacent areas was created, where the adjacent areas refer to the immediate surroundings of the inselbergs, which are essentially the plains according to the definition of inselbergs (
Mapping restitution. The images previously classified using Envi 4.7 software were sent to ArcGIS 10.4. These images had been clipped with the shapefile of inselbergs previously delimited to highlight the different classes of land use of inselbergs. The cartographic restitution was made with ArcGIS 10.4. The same land use classes were observed in both years (2003 and 2018) (
The anthropization of inselbergs was quantified using two indices: The dominance and the fragmentation index.
Dominance index (Dj) indicates the proportion of area taken up by the largest patch of class j (amax,j) in the total area aj (
The fragmentation index (Fj) measures the aggregation of pixels into classes and is considered as a measure of image complexity (
nj represents the total number of patches for class j; mj is usually in a raster file the number of pixels (
Inselbergs were classified as protected if their geographic coordinates fell within the boundaries of protected areas in the study area (Figure
The overall accuracy is 84.734% for 2003 and 83.599% for 2018 and the Kappa coefficients for the year 2003 and year 2018 maps were 0.823 and 0.784 respectively. The LULC Classifications results for 2003 and 2018 are illustrated in Table
Land cover classes | Years | |||
---|---|---|---|---|
2003 | 2018 | |||
km² | % | km² | % | |
Forest | 464.58 | 3.69 | 187.46 | 1.49 |
Wooded savanna | 2230.21 | 17.73 | 319.70 | 2.54 |
Tree and shrub savanna | 365.03 | 2.90 | 677.45 | 5.38 |
Grassland | 663.58 | 5.27 | 2225.20 | 17.69 |
Tree plantation | 957.48 | 7.61 | 1358.01 | 10.79 |
Field and fallow | 652.20 | 5.18 | 4719.72 | 37.51 |
Water surface | 7.12 | 0.06 | 14.52 | 0.12 |
Bare rock | 6349.60 | 50.47 | 2795.59 | 22.22 |
Settlement | 892.04 | 7.09 | 284.18 | 2.26 |
Total | 12581.84 | 100.00 | 12581.84 | 100.00 |
The land-use changes of the inselbergs are illustrated by the transition matrix (Table
Inselberg land-use transition matrix in % between 2003 and 2018. Br = Bare rock; W = Water surface; Se = Settlement; Ff = Field and fallow; Tp = Tree plantation; Gr = Grassland; Ts = Tree and shrub savanna; Ws = Wooded savanna; Fo = Forest.
2003/ 2018 | Br | W | Se | Ff | Tp | Gr | Ts | Ws | Fo | Total 2003 |
---|---|---|---|---|---|---|---|---|---|---|
Br | 11.99 | 0.21 | 0.75 | 10.29 | 4.37 | 8.61 | 4.45 | 1.46 | 0.89 | 43.01 |
W | 0.04 | 0.00 | 0.00 | 0.04 | 0.01 | 0.04 | 0.01 | 0.00 | 0.00 | 0.14 |
Se | 1.70 | 0.01 | 0.25 | 1.16 | 0.15 | 1.33 | 0.27 | 0.03 | 0.03 | 4.93 |
Ff | 1.24 | 0.03 | 0.07 | 1.80 | 1.12 | 0.57 | 0.19 | 0.08 | 0.48 | 5.57 |
Tp | 1.51 | 0.02 | 0.55 | 1.49 | 0.69 | 1.26 | 0.16 | 0.02 | 0.35 | 6.05 |
Gr | 1.95 | 0.04 | 0.09 | 2.00 | 0.77 | 1.63 | 0.84 | 0.24 | 0.02 | 7.60 |
Ts | 1.31 | 0.02 | 0.02 | 1.75 | 0.90 | 1.02 | 0.96 | 0.46 | 0.07 | 6.52 |
Ws | 2.92 | 0.21 | 0.02 | 4.83 | 3.48 | 1.71 | 1.93 | 1.78 | 0.76 | 17.63 |
Fo | 1.34 | 0.09 | 0.02 | 2.75 | 1.99 | 0.52 | 0.69 | 0.70 | 0.46 | 8.55 |
Total 2018 | 24.01 | 0.62 | 1.76 | 26.12 | 13.47 | 16.69 | 9.49 | 4.79 | 3.05 | 100.00 |
Table
Anthropization indices for 2003 and 2018 inselberg land use classes. Br = Bare rock; W = Water surface; Se = Settlement; Ff = Field and fallow; Tp = Tree plantation; Gr = Grassland; Ts = Tree and shrub savanna; Ws = Wooded savanna; Fo = Forest.
LULC | Br | W | Se | Ff | Tp | Gr | Ts | Ws | Fo |
---|---|---|---|---|---|---|---|---|---|
(Year 2003) | |||||||||
Dj 2003(%) | 4.76 | 0.01 | 3.71 | 0.22 | 1.59 | 0.22 | 0.04 | 5.41 | 0.04 |
Fj2003 | 0.02 | 0.16 | 0.04 | 0.06 | 0.97 | 0.05 | 0.15 | 0.03 | 0.18 |
(Year 2018) | |||||||||
Dj 2018(%) | 3.52 | 0.01 | 0.29 | 5.50 | 1.23 | 3.71 | 0.55 | 0.22 | 0.07 |
Fj2018 | 0.12 | 0.58 | 0.09 | 0.06 | 0.11 | 0.11 | 0.20 | 0.19 | 0.23 |
The percentage of different LULC on protected and unprotected inselbergs in 2003 and 2018 can be seen in Figure
The analysis of spatial comparisons using the chi-square test showed however, that there was no statistically significant difference in inselberg LULC between protected and unprotected areas in both 2003 and 2018 (χ2=12.637, df=8, p=0.125; χ2=4.204, df=8, p=0.8383).
Land use land cover proportion of protected and unprotected inselbergs in 2003 and 2018. 2003 UnpIns = 2003 unprotected inselberg (Inselbergs outside protected areas), 2003 PrIns = 2003 protected inselberg (Inselbergs inside protected areas), 20018 UnpIns = 2018 unprotected inselberg (Inselbergs outside protected areas), 2018 PrIns = 2018 protected inselberg (Inselbergs inside protected areas), Br = Bare rock; W = Water surface; Se = Settlement; Ff = Field and fallow; Tp = Tree plantation; Gr = Grassland; Ts = Tree and shrub savanna; Ws = Wooded savanna; Fo = Forest.
Open access geospatial data have already been used for the analysis of land use and land cover changes (LULC) of inselbergs (
The land use classes observed on the inselbergs and their adjacent areas in the Sudano-Guinean zone of Benin are the same as those observed in the Bale Mountain Eco-Region of Ethiopia (
Aside from human activities, the global temperature increase caused by climate change has a significant impact on the vegetation of inselbergs. According to
In addition, climate change can also increase the frequency and intensity of forest fires on inselbergs, which can have serious consequences on its vegetation. Indeed, every year, all the vegetation on inselbergs is consumed by dry season vegetation fires (
In addition to the decline in natural vegetation land cover classes, inselbergs and their adjacent areas have also been subject to increases in anthropogenic land cover classes, which are reflected in an increase in the class of bare rock to grassland and fields and fallow land. Indeed, the water that flows over the inselbergs is a source of rock alteration that favors colonization by grassy vegetation (
To appreciate the level of the anthropization of the inselbergs, the anthropization indices were calculated. They show an increase in the fragmentation index for natural land use classes. Fragmentation is recognized as the first consequence of the landscape transformation process (
In 2003, natural vegetation was equally present on both protected and unprotected inselbergs. However, settlements were more common on protected inselbergs, while tree plantations were more prevalent on unprotected inselbergs. By 2018, there had been an expansion of fields and fallow land on both protected and unprotected inselbergs, with a decrease in settlements and bare rocks, and an increase in grassland. This is consistent with previous research on LULC change in protected areas in Benin, which found that protected status alone may not be enough to prevent land use changes. For example, a study by
Interestingly, on protected inselbergs there was a significant increase in tree plantations, which may suggest efforts by conservation authorities to promote reforestation in protected areas. The finding that tree and shrub savanna grew better on protected inselbergs is in line with studies that have shown that protected areas can have positive effects on biodiversity and ecosystem functioning (
It is important to note, however, that the results of the chi-square test revealed no statistically significant differences in inselberg LULC between protected and unprotected areas in both 2003 and 2018. While the changes in LULC on protected and unprotected inselbergs suggest that protection measures may have some impact on land use, the lack of statistical significance suggests that other factors are also influencing land use patterns. These results are consistent with previous research that has shown that a range of social, economic, and environmental factors can influence land use decisions and patterns (
The analysis of land-use changes of inselbergs and their adjacent area was carried out in the Collines department. Sentinel 2 and Spot 5 satellite images of 10 m resolution were used for this purpose. The most observed trends are conversions from natural LULC classes to anthropogenic LULC, particularly with a significant increase of fields and fallows. In addition, while protected areas may have some impact on inselberg land use, other factors are also important in shaping these patterns. Further research is needed to better understand the complex social, economic, and environmental factors that influence land use decisions and patterns on protected and unprotected inselbergs. Also, protected areas in Benin may be facing increasing pressure from land use change and development. It is therefore urgent to develop a conservation and restoration plan for inselbergs of Benin for better conservation of their biological diversity through the following activities:
The involvement of the local population in the restoration and conservation of the inselbergs is crucial for ensuring the long-term sustainability of the project. By working together with the local population, we can restore and conserve the inselbergs and ensure its benefits for future generations.
The data that support the findings of this study are available from https://scihub.copernicus.eu/ for Sentinel 2 images. Spot 5 images are available from the “Observation spatiale des forêts tropicales“ (OSFACO) project (
R.E.D.A., S.S.M.T. and A.E.A. planned the research, R.E.D.A. conducted the field work, performed the images analyses and led the writing, while all authors critically revised the manuscript.
We would like to thank Dr. Thierry Agbanou and Mr. Chiméi Ahouangan for all the advices and help provided in the image processing, and to the administration of the Forest Inspectorate of the Collines department for its availability and advice. We also express gratitude to the villages chiefs of the localities we visited, whose consent facilitated the realization of the present work.