Land Use and Land Cover Dynamics in the Eseka Alluvial Gold Mining District, Centre Region, Cameroon

Shu Vejiline Lum-Ndob^1*, Fonge Beatrice Ambo¹, Ambe Godlove Neba¹, Ateh Kevin Ijunghi², Enerst Tata², Cheo Emmanuel Suh^3,4

¹Department of Plant Science, Faculty of Science, University of Buea, Buea, Cameroon.
²Department of Mining and Mineral Engineering, National Higher Polytechnic Institute, The University of Bamenda, Bambili, Cameroon.
³Economic Geology Unit, Department of Geology, Faculty of Science, University of Buea, Buea, Cameroon.
⁴Department of Geology, Mining and Environmental Science, The University of Bamenda, Bambili, Cameroon.

DOI: 10.4236/jgis.2024.164018                

Abstract

Local populations in Cameroon thrive on forest resources and the flow of ecosystem services they provide are pivotal in sustaining national economy, improving people’s lives, safeguarding biodiversity, and mitigating the impacts of environmental changes. The exploitation of these resources invariably leads to deforestation and forest degradation. This study was designed to evaluate land use land cover change (LULCC) in the Eseka alluvial gold mining district with the aid of Landsat images. In the investigation of forest cover change, four Landsat satellite images for (1990, 2002, 2015 and 2022) were used. Ground-truthing also helped to identify the activities carried out by the local population and to determine agents, drivers and pressures of land use and land cover change. Four main land cover classes namely: forest, agricultural land, settlement/mining camps and water bodies were selected. Between 1990 and 2022, the proportion of forest decreased from 98% to 34% while those of agricultural land and settlement/mining camps increased from 2% to 60% and 0.54% to 6% respectively. Analysis showed ongoing deforestation with forest cover loss of ~98,263 ha in 32 years giving a cover change percentage of 63.94%. Kappa coefficient for the study period ranged from 0.92 to 0.99. Forest cover loss could be attributed to farming activities, wood extraction and alluvial gold mining activities. Economic motives notably the need to increase household income from a frequent demand for farm and wood products in neighbouring towns and the quest for gold were the main drivers of these activities. Hence, this study assesses the impact of human activities from the mining sector on the forest ecosystem in a bid to inform mitigation policies.

Keywords

Land Use and Land Cover Changes, Biodiversity, Alluvial Gold Mining, Drivers, Landsat Images, Ground-Truthing, Mitigation

1. Introduction

Natural resources are vital for human survival and land uniquely forms the basis for all terrestrial ecosystem services [1]. Tropical rainforests based on their characteristic plant cover, are important components of terrestrial ecosystems due to their rich biodiversity and they potentially provide a variety of ecosystem services that serve as the bedrock for human wellbeing [2]. These services range from the supply of wood as well as Non Timber Forest Products (NTFPs), climate regulation, and the release of atmospheric oxygen, the conservation of soil and water resources to the provision of habitats for other living organisms [3]. Forest resources and their provision of numerous ecosystem services play an immeasurable role in supporting both national and regional economies [4], improving livelihoods, protecting biodiversity, and mitigating the impacts of a changing climate [5]. However, the degradation of this priceless resource in the form of land use/land cover (LULC) changes over time has become a severe environmental problem [6]. Deforestation reverses this natural order, causing losses in biodiversity, changing climate conditions, intensifying soil erosion and eminent desertification [7] [8]. Globally, LULC changes have been recorded as the most important cause of the distortion of ecosystem services and Cameroon is not an exception [9] [10]. The extractive industry plays a vital role in this deforestation process since mining encompasses clearing of large expands of land cover. In the tropics, the introduction of mining is also known to bring about an influx of population and major land cover conversions from forest to agricultural land through shifting cultivation [11]. The imbalances caused by these distortions have the potential of mounting enormous pressure on land resources, thus increasing their susceptibility to further change over time.

Cameroon’s forest cover represents the second largest tropical rainforest in the Congo Basin after that of the Democratic Republic of Congo. The forest spans a land area of ~22.5 million hectares, corresponding to about 48% of its national territory [12] [13]. The natural forest has reportedly dwindled over time especially resulting from rapid population growth [14], with a population size that has more than doubled in the last three decades and is projected to increase even further by the year 2030 [15]. Forest cover loss of ~4400 ha was recorded between 1990 and 2010 alone [16]. The average annual rates of gross deforestation in Cameroon were estimated at 0.10% for the period 1990 to 2000 and 0.17% between 2000 and 2005 [17], indicating a drastic increase in deforestation rate during this time span. By the year 2010, the average rate of gross deforestation in the country was in the order of 0.20% [18], ranking Cameroon second in the Congo basin, after the Democratic Republic of Congo (0.22%), with the highest rate of deforestation at the time [18] [19].

The contributions made by extractive industries in general and alluvial gold mining even at artisanal levels in particular to these changes within the country cannot be undermined. Gold is alluvial with evidence of its primary accumulation in different quartz veins [20] and its extraction thereof begins with vegetation clearing which brings about other environmental changes on the landscape. Therefore, to better plan and sustainably manage these natural forest resources, an in-depth knowledge on the land use and land cover is required, as well as the establishment of a change detection mechanism. This is an important tool in the assessment of the environment and its evolution over time.

Eseka is one of the ancient towns of Cameroon and hosts an important terminal of the Cameroon railway line (CAMRAIL) that has been at the centre of the transportation of both goods and humans in and out of the area. The 91.78 km stretch of road from Boumyebel through Eseka down to Lulodolf in the South Region of Cameroon that was paved in the 1980s and got rehabilitated recently in 2010, increased Eseka’s accessibility to neighbouring towns including the capital city, Yaoundé. This has further eased the production and transportation of food and cash crops to bigger neighbouring markets.

Data generated by the National Institute of Statistics Cameroon, reveal that Eseka has experienced population fluctuations over the years due to its socio/political history before independence and more recently from 23,242 in 2005 to 22,221 in 2014. The main economic activities include agriculture, food crop production, small holder schemes for cocoa, oil palm and plantation agriculture (SOCAPALM), extraction of other NTFPs and hunting of bush [21], in a bit to generate income and create employment opportunities for the local population. The area has also seen the exploitation of mineral resources like iron ore and in recent times, gold at artisanal levels [22].

LULC changes play a significant role in spatio-temporal environmental stability especially because it is linked to local, regional, and global climate conditions, biodiversity stability, clean water, agriculture, and food security [23] [24]. Critical thinking and research in the domain of land change science is therefore inevitable in understanding global environmental change and sustainability. It enables public agencies and private organizations understand and make comprehensive plans for future intervention and for the designation of effective land management policies and decision making [25] [26]. It is also crucial for understanding the processes involved in environmental change and the problem that must be addressed for a sustainable improvement of living conditions and standards [27]. Interventions on the dynamics of LULC changes however need to be timely for an in depth understanding of the relationships and interactions between human and natural phenomena. This will go a long way to enhance management of natural resource bases which serve as a main source of livelihood for people living in rural areas [28]. This calls for global attention for continuous evaluation of the LULC changes, and remote sensing presents itself as a fundamental technological tool to assist in the spatial and temporal planning of land use types and vegetation cover evolution [29] [30]. This study therefore investigated LULC in an area recently inundated by artisanal gold mining activities. It quantified LULC in the Eseka alluvial gold mining district and attempted to highlight the constant challenge of balancing earth’s resource extraction and environmental protection.

2. Materials and Methods

2.1. Description of the Study Area

Eseka is the administrative headquarter of the Nyong Sub Division in the “Nyong et Kelle” Division of the Centre Region of Cameroon. It is located between latitudes 2˚24' - 4˚39'North and Longitudes 9˚59' - 11˚31'East [31]. It is covered by low land rainforest and generally characterised by a rugged to undulating topography with steep hills and topographic highs of approximately 600 m above sea level (Figure 1). It comprises dissected plains, isolated and complex hills and mountains that are prone to intense erosion [32].

Figure 1. Location of the study area (Eseka).

The soil types in the area are mainly lateritic soils, which are typically unsaturated and yellowish brown, sandy, clay at the surface with reddish clayey material at depth [33]. These soils are acidic and characterised by low nutrient content with a relatively low cation exchange capacity. They have very low nitrogen contents due to the rapid degradation of the organic material, and are generally deeply weathered and deficient in phosphorus [34]. The soils become hydromorphic, particularly in the wetland areas. The drainage of the study area is dendritic and dominated by the River Nyong and its tributaries which carries its waters into the Atlantic Ocean at Kribi [35]. Areas of watersheds affected by these rivers are not very wide but drain rapidly because of the associated steep slopes and heavy rainfall.

Eseka has a tropical monsoon climate that is hot and humid all year round. The area experiences a bi-modal pattern of rainfall with four seasons each year: two rainy, and two dry seasons. The long dry season spans from December to March, the short rainy season from March to June, the short dry season in July and August, and the long rainy season from September to December [36]. Rainfall averages 1500 - 2000 mm per year and precipitation sometimes occurs even during the dry seasons. The highest amounts of rainfall are measured in the months of September and October [37]. The average annual temperature is 25.6˚C and annual average relative humidity ranges between 83% and 90% [38]. The study area represents part of the dense and humid rainforest of Cameroon, where available data shows that its flora is very diverse and represents over 60% of the country’s biodiversity [39].

The vegetation cover is mainly made up of large primary forest and secondary shrubs with grasses. Some of the exploitable species in these forests include Triplochyton scleroxylon K.Schum. (Ayous), Mansonia altissima (A. Chev.) A. Chev. (Bete), Entandrophragma cylindricum (Sprague) Sprague (Sapelli), Terminalia superba Engl. and Diels (Frake), Milicia excelsa (welw) (Iroko), Guibourtia tessmannii (Harms) (Bubinga), and Nauclea gilletii (Bilinga). The forest also plays host to a number of Non Timber Forest Products (NTFPs) like Garcinia kola Heckel (Bitter cola), Gnetum africanum (Welv.) (okok or eru), Irvingia gabonensis (Aubry-Lecomte ex. O’Rorke) Baill. (Bush mango) and Elaeis guineensis Jacq. (Oil palm) [40]. The secondary forest mostly occurs near the villages and comprises mosaic cultures and secondary forests of different ages and plays host to a variety of fauna ranging from micro fauna to giant reptiles like snakes and mammals like bats, monkeys, gorillas and chimpanzees. The existing fallows in the locality are mainly colonized by grasses like Chromolaena odorata (L) R. M King & H. Robinson and trees like Ceiba pentandra (L.) Gaertn. (Boma), Musanga cercropioides R. br. (parassolier), Cecropia cercropioides (L.), Cecropia peltata (L.) and Eupatorium sp.

2.2. Data Collection

This process involved image acquisition, processing and interpretation. The principal data source for land use land cover classification and change analysis over the study area were remotely sensed data through a series of Landsat ETM+ imagery. This was based on the Landsat ETM+ satellite images for 1990, 2002, 2015 and 2022 respectively. Images were obtained from the United States Geological Survey-Global Visualization Viewer’s (USGS) Earth Data Acquisition Website via http://glovis.usgs.gov where an account was first created and images downloaded. In order to reduce cloud cover, images from the dry season were selected for this study. Pre-processing of the images included atmospheric and geometric corrections, suitable band selection, layer stacking and image enhancements which were applied before performing the classification of the remote sensing data [41]. To combine time series image data sets for change detection at pixel level for the data sets were registered to their corresponding Landsat image based on automated image-to-image registration approaches using a set of ground control points (GCPs). Unsupervised image classification was carried out for the Landsat images where the algorithm categorized all the pixels in the images into land-cover classes or themes based on their reflectance values. Thus the algorithm assigned the image’s pixel to a defined number of classes based on their value in different channels (i.e. no training areas were used). Spatial enhancement technique (that is, focal analysis) was applied on the Landsat ETM+ images in order to improve the image quality. The images were geo-referenced (Universal Transverse Mercator-UTM, WGS84) from the data set provider. Unsupervised classification was conducted to determine spectral differences in land cover classes without creation of any particular training parcels for field verification and eventual use for a supervised image classification.

The supervised image classification via maximum likelihood algorithm was used to classify all the Landsat EMT+ scenes. In classifying the images into various themes, the supervised approach to classification was adopted using the ENVI 4.1 software. The pixel categorization process was supervised by specifying to the computer algorithm numerical descriptors of the various land cover types present in the scene. This was to ascertain the type and spatial extent of various land-use/land-cover types. To do this, representative sample sites of known land-cover types, called training areas or training sites were used to compile a numerical interpretation key. The key was then used to describe the spectral attributes for each feature type of interest. Each pixel in the data was then compared numerically to each category in the interpretation key and labelled with the name of the category it resembled closely. All the LULC type classes (vegetation, hydrology, settlement etc.) were carefully studied for the different images to detect any differences between them. These were compared over the time span of the study to elucidate the rate of environmental change that might have occurred. The reference data for these years were collected through visual interpretation of the raw data of the Landsat ETM+ images of the respective years.

Accuracy assessment is a critical component for maps generated from any remote sensing data and therefore error matrixes to assess the accuracies and errors of the classification is imperative. Computing the error matrix is in the most common way to present the accuracy of the classification results. Overall accuracy, user’s and producer’s accuracies, and the Kappa index (the ratio of the number of well-ranked pixels to the total of the pixels surveyed) were calculated for the data and statistics derived from the error matrices. The Kappa statistics incorporated the off diagonal elements of the error matrices and represented agreement obtained after removing the proportion of agreement that could be expected to occur by chance.

2.3. Data Analyses

The Landsat image of each study year was independently classified with supervised classification technique using ground truth data on existing LULC. Maximum likelihood classifier (MLC) algorithm was applied to classify the LULC types in ERDAS Imagine 2014 software. The total area covered by each land use was then calculated using ENVI 4.1 and values obtained used to produce tables representing the different land use classes in Microsoft excel version 16. Post classification change detection method was used to compare independently produced classified images. This was achieved via an extraction of statistics for time series analysis and trends of change of the different identified land uses and land cover classes. Finally, vectors derived from digital processing in ENVI 4.1 were imported into Arc Map 10.2 software for the extraction of the layers used for digitization, generation of databases, and production of the maps.

3. Results

Image classification of Landsat scenes of 1990, 2002, 2015 and 2022 are presented in four main land cover classes namely; forest, settlement/mining camps, agricultural land and water bodies. The classified Landsat EMT+ of the 1990 scene (Figure 2(a)) revealed that, of the total surface area (153,678 ha), 97.78% was covered by forest, 2.0% by agricultural land, 0.54% settlement and water bodies 0.01%. The 2002 image showed that forest cover was 82%, agricultural land 12%, settlement 6% and water bodies 0.23% (Figure 2(b)). Landsat image for 2015 showed that the forest cover stood at 74%, agricultural land 22% and settlement 4%. The category for water bodies was not detected by the satellite for this period. (Figure 2(c)) and finally, the 2022 image revealed that the forest cover was 34%, agricultural land 60%, settlement 34% and water bodies 0.19% (Figure 2(d)).

The colour changes on the classified Landsat EMT+ maps represent the different land cover categories in the study (Figures 3(a)-(d)). The region of interest (ROI) had a total surface area of 153,678 ha and its distribution into the different land cover class categories for the four study periods is represented below (Table 1).

(a) (b)

(c) (d)

Figure 2. (a) Land cover status for 1990; (b) Land cover status for 2002; (c) Land cover status for 2015; (d) Land cover status for 2022.

(a) (b)

(c) (d)

Figure 3. Land use and land cover change categories from 1990 to 2022 as revealed by satellite imagery.

Table 1. Land Cover categories (surface area and percentages) of the Eseka Alluvial Gold Mining Area from 1990 to 2022.

Class type	1990		2002		2015		2022		1990-2022
	Area (ha)	%	Area (ha)	%	Area (ha)	%	Area (ha)	%	AC (ha/yr)	%CC
Forest	150263.0	97.78	125,635	81.75	114,624	74.59	52,000	33.84	−3070	−63.94
Settlement/ mining camps	823.0	0.54	9728	6.33	5757	3.75	8836	5.75	2811.5	5.21
Agricultural Land	2582.0	1.68	17,965	11.69	33,297	21.67	92,550	60.22	250.406	58.54
Water bodies	10.0	0.01	350	0.23	/	/	292	0.19	8.813	0.18
Total	153,678	100	153,678	100	153,678	100	153,678	100	0.719	−0.01

The computed error matrices for the classified images (Table 2) portrayed an overall classified confusion of less than 10% with overall accuracy varying from 94.26% to 99.54% and degree of representation of ground truths (Kappa index) ranging between 0.90% and 0.99% for the different Landsat images.

Table 2. Accuracy, classified confusion matrix and Kappa coefficient for 1990, 2002, 2015 and 2022 Landsat images.

	1990		2002		2015		2022
Land use Classes	Prod. Acc. (%)	User. Acc. (%)	Prod. Acc. (%)	User. Acc. (%)	Prod. Acc. (%)	User. Acc. (%)	Prod. Acc. (%)	User. Acc. (%)
Forest Area	99.36	93.03	100	98.44	99.62	99.88	99.44	90.34
Agricultural land	95.26	98.53	100	100	96.82	100	86.69	98.97
Settlement/mining camps	97.58	96.80	100	100	100	89.74	95.59	99.54
Water bodies	40.00	95.83	90.63	100	/	/	85.88	100
Overall Accuracy	94.97		99.28		99.54		94.26
Kappa Coefficient	0.92		0.99		0.96		0.91

4. Discussion

Major LULCC have occurred in all biomes, with significant replacement of forest and natural vegetation by pastures and agriculture the world over. Research findings in the Amazon basin indicate that LULCC have imposed important local and remote influences on the temperature, precipitation and cloudiness that have affected plants and ecosystems of South America as a whole [42]. These have induced changes in the composition and structure of the natural vegetation, leading to significant local warming and also circulation changes. The response of the natural vegetation to the LULCC-induced changes in climatic conditions in this rainforest biome is particularly strong during the dry season, even though hydrological cycle over major parts of the areas have intensified during the wet season [43]. Trends similar to these have been reported in the Congo River Basin [44]. Documented that the Kasai catchment of the Congo River Basin is experiencing poor land use planning that is causing sediment deposition thus impacting the stream flow [45]. Equally reported that the rate of deforestation among countries in the Congo Basin increased drastically in recent decades with population expansion and agricultural activities serving as the main drivers. The trends in our study corroborate these findings.

The results demonstrated four major LULC types and revealed that the Eseka alluvial gold mining district has undergone remarkable changes in its LULC over time. Overall, vegetation experienced a regression in land cover, which was particularly evident in the consistent reduction in the forest cover area from 150,263 ha in 1990 to 125,635 ha in 2002 down to 114,624 ha in 2015 and finally to 52,000 ha in 2022, amounting to a net loss of 98,263 ha over the 32 years span of the study. Summarily, there were more net losses in the forest category compared to other land use types, recording a percentage cover change of −63.94%. The study period equally saw net gains in the categories of water volume (0.18%), settlement/mining camps (5.21%) and agricultural land uses (58.54%) (Table 1). Analysis of the rates of deforestation for the study period revealed that between 1990-2002, the rate stood at 1.34% per year. From 2002-2015 the rate of deforestation dropped to 0.55% per year and the rose to 5.82% between the years 2015-2022. The global rate of deforestation in the area over the study period stood at 1.99% per year.

The steady decrease in forest cover area that amounted to −98,263 ha during this period could be attributed to drivers of ecosystem change which could either be direct or indirect. While direct drivers exert tangible impact on the land surface, indirect drivers represent the underlying causes of the direct drivers brought about by direct anthropogenic drivers. The driving forces of LULCC are often a mix between anthropogenic (social, political, economic, demographic, technological, and cultural) and biophysical factors with direct or indirect impacts which may necessitate mitigation on curative actions in response. The main driver of forest cover change in the study area was established to be socio-economic. Land cover sequences like conversion from forest to agricultural land, and to a little extent, settlement/mining camps were principally responsible for the changes observed. The increasing rates of deforestation in the area could be attributed to increasing agricultural practices driven by population booms that equally brought about an increase in settlement area (Figure 2) and hence a rise in household income. The rehabilitation of the road network across the study area also granted easy access to agricultural investors who came in to farm from out of town and transportation of their produce was much guaranteed. The establishment of the SOCAPALM plantations in Eseka during the study period saw the rehabilitation of old farms and the creation of new ones at the cost of primary forest area. Other researchers in the country, Africa and around the world have documented similar findings [18]. Outlined similar trends in the Ajei community forest area in Ngie, North West Cameroon as a result of anthropogenic activities into the forest area [46]. Equally reported declining forest cover in Tubah Sub division, Cameroon between 1986 and 2017 for similar reasons. Other recent studies around Cameroon [47] [48] also showed that even classified forests are facing many anthropogenic pressures that lead to deforestation and degradation.

The commencement of artisanal and small scale gold mining activities that boomed in the centre and southern parts of Cameroon in the early 2000s was also a contributing factor to the changes observed in the landscape as farmers were forced to leave their old farms and sought new ones further into the forest. The influx of gold merchants from within the country and Africa as a whole during peak seasons got the sector buzzling, increased cost of living and further increased the rate of deforestation for the exploitation of this valuable resource. Findings by [49] who reported that the proportion of mining-related deforestation across the Peruvian Amazon increased outside the mining corridor due to the expansion of a few major mining sites in protected area buffer zones and certain indigenous communities. This could have been accounted for by the millions of poor people involved. Moreover, reforms aimed at formalizing ASM pay more attention on measures to improve equity and protect workers but say little about how such measures will protect forests or wetlands.

The lower rate of deforestation (0.55%) for the second study period (2002-2015) was probably due to a decline in human activities within the forest zone. SOCAPALM for example after its establishment faced stiff resistance from the local population of Eseka in attempting to replant pieces of land on which palms were already very old. The locals argued that, only the palms were privatized by the Government to the company and not the land. The company also suffered from high level stealing of fresh fruits bunches by the locals who often encroached into the plantation. This mishap coupled with the old age of the palms drastically increased the cost of production for the company which could not break even and consequently led to a series of layoffs of some company workers at Eseka. Similar trends have been reported by [50], in the rainforest of the Littoral Region of Cameroon. Rural exodus to neighbouring roadside communities like Boumnyebel and to capital city Yaoundé, in the quest for better living standards could also have been a contributing factor. The rates of deforestation (1.34%) observed in the study area between 1990-2002 was lower than that of 1.83% reported by [18] in Ajie community forest, North West Cameroon and even far higher than that of 0.65% reported by [51] for 1990-2000 period in Eastern Region of Cameroon and also greater than that of 0.13% found by [52] for the Congo basin for the same period which doubled to 0.26% between 2000-2010. This rate drastically increased to over 5% during the last phase of the study period. This could be attributed to the commencement of ASGM in the later part of 2015, the intensification of illegal logging activities and agroforestry that is ongoing in the area. An indication that ASGM activities together with other anthropogenic activities like, logging and farming had a great toll on the vegetation cover of the area. Similar results have equally been reported around the country.

Infrastructural development and poor governance also have a significant role to play in the land cover dynamics of the study area. The creation of road networks within the forested area to facilitate movement of goods and persons in and out of the forest patches can further intensify the state of this already fragmented forest. In as much as there are land use policies in the country, the management still has challenges especially in its implementation phase. Difficulties in initiating and scaling up community participation especially on rational harvesting and exploitation of wood products, and the intensive use of the land still pose problems. Similar findings of such direct and underlying drivers of deforestation have been reported by [53] [54] [55] in other parts of Africa and beyond.

The overall confusion in the classification of the downloaded maps was less than 10% and Kappa index values ranged between 0.91 and 0.99 for the entire study period. This is an indication that the classifications performed are reliable according to the scale of [56].

5. Conclusion

The conversion of land to other uses is increasingly threatening the forest cover and therefore the biodiversity of the forest area of Eseka in the Nyong and Kelle Division in the Center Region of Cameroon. From the study, statistics show that the natural forest cover reduced from 64.30% period to 48.67%, indicating a high rate of deforestation during the study period. Even though the rate of deforestation (1.99%) recorded in this study is lower than those reported in other forests across the country, it is equally a very significant loss (15.63% loss in dense forest cover) in 32 years. The above loss (impact) in forest cover is a result of pressure on the ecosystem brought about by actions led by farmers, wood extractors, logging and ASGM activities which convert the forest cover to these other land uses. Of all pressures, agriculture and ASGM are at the centre of forest cover losses in the area. Economic gains notably the need to increase household incomes facilitated by frequent demand for agricultural, wood products and gold at peak seasons in markets represent the main drivers of this forest change. However, the complex and diverse interactions among these social-ecological systems make it difficult to identify and quantify the main drivers of LULCC. A holistic approach across disciplines with a focus on direct and indirect influences causing land cover changes is needed for a comprehensive driver analysis which is still rarely conducted in land system science. Furthermore, an assessment of LULCC should include, as much as possible, the impact of the change on natural resources availability and/or disaster risks. Knowledge of relevant driving forces and their impact on gains and losses in ecosystem services contributes to provide consultations for sustainable development by delivering improved decision criteria and policy statements. These findings are expected to contribute to the discussion on the reduction in CO₂ emissions from deforestation and forest degradation (REDD+) and serve as reference for this period.

Acknowledgements

The authors are thankful to the small-scale miners and geologists of CAMINA S.A. for assisting during the ground-truthing phase.

Author Contributions

Project is part of the Ph.D. thesis of SVLN supervised by ABF and CES. SVLN drafted the manuscript while AGN, AKI and ET took part in the satellite image data processing and field works. All authors read, reviewed and approved the manuscript.

Conflicts of Interest

The authors declare that they have no conflict of interest and all data used are available from the first author.

References

[1]	Fang, Z., Ding, T., Chen, J., Xue, S., Zhou, Q., Wang, Y., et al. (2022) Impacts of Land Use/Land Cover Changes on Ecosystem Services in Ecologically Fragile Regions. Science of the Total Environment, 831, Article ID: 154967.[CrossRef] [PubMed]
[2]	Watson, J.E.M., Evans, T., Venter, O., Williams, B., Tulloch, A., Stewart, C., et al. (2018) The Exceptional Value of Intact Forest Ecosystems. Nature Ecology & Evolution, 2, 599-610.[CrossRef] [PubMed]
[3]	Mo, R., Wang, Y., Dong, S., Ma, J. and Mo, Y. (2023) Ecosystem Service Evaluation and Multi-Objective Management of Pinus massoniana Lamb. Plantations in Guangxi, China. Forests, 14, Article 213.[CrossRef]
[4]	Acharya, R.P., Maraseni, T. and Cockfield, G. (2019) Global Trend of Forest Ecosystem Services Valuation—An Analysis of Publications. Ecosystem Services, 39, Article ID: 100979.[CrossRef]
[5]	Imbert, J.B., Blanco, J.A., Candel-Pérez, D., Lo, Y., González de Andrés, E., Yeste, A., et al. (2020) Synergies between Climate Change, Biodiversity, Ecosystem Function and Services, Indirect Drivers of Change and Human Well-Being in Forests. In: Venkatramanan, V., Shah, S. and Prasad, R., Eds., Exploring Synergies and Trade-Offs between Climate Change and the Sustainable Development Goals, Springer, 263-320.[CrossRef]
[6]	Kuma, H.G., Feyessa, F.F. and Demissie, T.A. (2022) Land-use/land-cover Changes and Implications in Southern Ethiopia: Evidence from Remote Sensing and Informants. Heliyon, 8, e09071.[CrossRef] [PubMed]
[7]	Njora, B. and Yilmaz, H. (2022) Analysis of the Effects of Deforestation on the Environment and Agriculture in Kenya. International Journal of Water Management and Diplomacy, 1, 91-110.
[8]	Hu, X., Næss, J.S., Iordan, C.M., Huang, B., Zhao, W. and Cherubini, F. (2021) Recent Global Land Cover Dynamics and Implications for Soil Erosion and Carbon Losses from Deforestation. Anthropocene, 34, Article ID: 100291.[CrossRef]
[9]	Assede, E.S.P., Orou, H., Biaou, S.S.H., Geldenhuys, C.J., Ahononga, F.C. and Chirwa, P.W. (2023) Understanding Drivers of Land Use and Land Cover Change in Africa: A Review. Current Landscape Ecology Reports, 8, 62-72.[CrossRef]
[10]	Shiferaw, H., Alamirew, T., Kassawmar, T. and Zeleke, G. (2021) Evaluating Ecosystems Services Values Due to Land Use Transformation in the Gojeb Watershed, Southwest Ethiopia. Environmental Systems Research, 10, Article No. 22.[CrossRef]
[11]	Kumi, S., Addo-Fordjour, P., Fei-Baffoe, B., Belford, E.J.D. and Ameyaw, Y. (2021) Land Use Land Cover Dynamics and Fragmentation-Induced Changes in Woody Plant Community Structure in a Mining Landscape, Ghana. Trees, Forests and People, 4, Article ID: 100070.[CrossRef]
[12]	De Wasseige, C., Tadoum M., Atyi E., and Doumenge C. (2015) The Forests of the Congo Basin-Forests and Climate Change. Weyrich.
[13]	Kabelong Banoho, L., Zapfack, L., Weladji, R.B., Chimi Djomo, C., Nyako, M.C., Nasang, J.M., et al. (2020) Biodiversity and Carbon Sequestration Potential in Two Types of Tropical Rainforest, Cameroon. Acta Oecologica, 105, Article ID: 103562.[CrossRef]
[14]	Fogwe, Z., Paul Tume, S.J. and Fouda, M. (2019) Eucalyptus Tree Colonization of the Bafut-Ngemba Forest Reserve, North West Region, Cameroon. Environment & Ecosystem Science, 3, 12-16.[CrossRef]
[15]	Oumar, S.B. and Sama, M.C. (2017) The Implications of Poverty on the Cameroon Economic Emergence Vision 2035. Applied Economics and Finance, 4, 111-128.[CrossRef]
[16]	Temgoua, L.F., Meyabeme Elono, A.L., Mfonkwet Njiaghait, Y., Ngouh, A. and Nzuta Kengne, C. (2021) Land Use and Land Cover Dynamics in the Melap Forest Reserve, West Cameroon: Implications for Sustainable Management. Geology, Ecology, and Landscapes, 6, 305-315.[CrossRef]
[17]	Meli Fokeng, R., Gadinga Forje, W., Meli Meli, V. and Nyuyki Bodzemo, B. (2020) Multi-temporal Forest Cover Change Detection in the Metchie-Ngoum Protection Forest Reserve, West Region of Cameroon. The Egyptian Journal of Remote Sensing and Space Science, 23, 113-124.[CrossRef]
[18]	Temgoua, L.F., Ajonina, G. and Woyu, H.B. (2018) Land Use and Land Cover Change Analysis in Ajei Upland Watershed Community Forest, North West Region, Cameroon. Journal of Geoscience and Environment Protection, 6, 83-99.[CrossRef]
[19]	Verhegghen, A., Eva, H., Desclée, B. and Achard, F. (2016) Review and Combination of Recent Remote Sensing Based Products for Forest Cover Change Assessments in Cameroon. International Forestry Review, 18, 14-25.[CrossRef]
[20]	Suh, C.E. and Lehmann, B. (2003) Morphology and Electron-Probe Microanalysis of Residual Gold-Grains at Dimako, Southeast Cameroon. Neues Jahrbuch für Mineralogie—Monatshefte, 2003, 255-275.[CrossRef]
[21]	Nkongho, R.N., Ndjogui, T.E. and Levang, P. (2015) History of Partnership between Agro-Industries and Oil Palm Smallholders in Cameroon. OCL, 22, A301.[CrossRef]
[22]	Sababa, E., Mbesse, C.O., Wandji Mouko, C.N., Ekoa Bessa, A.Z. and Ndjigui, P. (2021) Geochemistry of Stream Sediments from Eséka Area (SW Cameroon): Implications for Surface Process Assessment and Precious Metals (Au, Pd, and Pt) Exploration. Journal of Sedimentary Environments, 7, 43-66.[CrossRef]
[23]	Zhao, H. and Deng, X. (2024) Spatiotemporal Variation in Ecosystem Health Caused by Land Use and Land Cover Changes in Pakistan. Ecosystem Health and Sustainability, 10, Article ID: 0161.[CrossRef]
[24]	Roy, P.S., Ramachandran, R.M., Paul, O., Thakur, P.K., Ravan, S., Behera, M.D., et al. (2022) Anthropogenic Land Use and Land Cover Changes—A Review on Its Environmental Consequences and Climate Change. Journal of the Indian Society of Remote Sensing, 50, 1615-1640.[CrossRef]
[25]	Serrai, S.C. and Djiar, K.A. (2024) Algiers Master Plan, Land Use and Forced Relocation: Monitoring Change with a Spatial Decision Support System. Land Use Policy, 139, Article ID: 107065.[CrossRef]
[26]	Ipbes, W. (2019) Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. Summary for Policy Makers of the Global Assessment Report on Biodiversity and Ecosystem Services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES Secretariat.
[27]	Sanz, M.J., De Vente, J., Chotte, J.L., Bernoux, M., Kust, G., Ruiz I. et al. (2017) Sustainable Land Management Contribution to Successful Land-Based Climate Change Adaptation and Mitigation: A Report of the Science-Policy Interface. United Nations Convention to Combat Desertification (UNCCD).
[28]	Abebe, G., Getachew, D. and Ewunetu, A. (2021) Analysing Land Use/Land Cover Changes and Its Dynamics Using Remote Sensing and GIS in Gubalafito District, Northeastern Ethiopia. SN Applied Sciences, 4, Article No. 30.[CrossRef]
[29]	Pandey, P.C., Koutsias, N., Petropoulos, G.P., Srivastava, P.K. and Ben Dor, E. (2019) Land Use/Land Cover in View of Earth Observation: Data Sources, Input Dimensions, and Classifiers—A Review of the State of the Art. Geocarto International, 36, 957-988.[CrossRef]
[30]	Wang, Y., Sun, Y., Cao, X., Wang, Y., Zhang, W. and Cheng, X. (2023) A Review of Regional and Global Scale Land Use/Land Cover (LULC) Mapping Products Generated from Satellite Remote Sensing. ISPRS Journal of Photogrammetry and Remote Sensing, 206, 311-334.[CrossRef]
[31]	Ndongmo, G.N., Koumetio, F., Ngapgue, F. and Lemoubou, E.L. (2023) Characterization of near Surface Intrusions in South-West Cameroon Zone Using Gravity Data: Mining and Geothermal Implications. Journal of Geoscience and Environment Protection, 11, 268-296.[CrossRef]
[32]	Thierry, A.A., Bernard, N., Armel Zacharie, E.B., Jacques Bertrand, O., Tasin, B.G., Jean-Bosco, O., et al. (2021) Correlations between Morphoneotectonic Parameters and Precambrian Tectonic Structures in the Nyong Group Greenstone Belts: Example of the Kopongo Area (NW of the Congo Craton, Southwestern Cameroon). Journal of African Earth Sciences, 182, Article ID: 104272. [Google Scholar] [CrossRef]
[33]	Engon, T.C., Abane, M.A., Zo’o Zame, P., Ekomane, E., Bekoa, E., Mvogo, K., et al. (2017) Morphological, Physico-Chemical and Geochemical Characterization of Two Weathering Profiles Developed on Limestone from the Mintom Formation in the Tropical Humid Zone of Cameroon. Journal of African Earth Sciences, 131, 198-212.[CrossRef]
[34]	Kagou Dongmo, A., Guedjeo, C.S., Azinwi Tamfuh, P., Wotchoko, P., Chenyi, M.L. and Aziwo B.T. (2018) Geochemical and Geotechnical Characterization of Soils Developed on Volcanic Rocks on the Bamenda Mountain (Cameroon Volcanic Line). International Journal of Advanced Geosciences, 6, 184-194. [Google Scholar] [CrossRef]
[35]	Nsangou, M.N., Owona, S., Owono, F.M., Ateba, C.B., Ondoa, J.M. and Ekodeck, G.E. (2019) Evolution of the Edea-Eseka Geomorphic Units: A ‘Natural Laboratory’ of a Prototype Active Rifting Cross-Cutting the Southern Cameroon (West African) Passive Margin. Geological Journal, 55, 4864-4888.[CrossRef]
[36]	Bang, H.N. (2022) A Concise Appraisal of Cameroon’s Hazard Risk Profile: Multi-Hazard Inventories, Causes, Consequences and Implications for Disaster Management. GeoHazards, 3, 55-87.[CrossRef]
[37]	Sangen, M. (2012) Late Quaternary Palaeoenvironments in Southern Cameroon as Evidenced by Alluvial Sediments from the Tropical Rain Forest and Savanna Domain. Palaeoecology of Africa, 31, 137-236.
[38]	Grelle, M.H., Auguste, O., Philippe, S.A. and Etienne, B.A. (2016) The Impact of Climate Variabilities on the Water Resources of Nyong River (Cameroon) during Few Anterior Decades. International Journal of Geosciences, 7, 459-469.[CrossRef]
[39]	Chimi, D., Zapfack, L. and Djomo, N. (2018) Diversity, Structure and Biomass (above and below) in a Semi-Deciduous Moist Forest of East Region of Cameroon. Journal of Biodiversity and Environment Sciences, 12, 60-72.
[40]	Ingram, V. and Schure, J. (2010) Review of Non Timber Forest Products (NTFPs) in Central Africa. CIFOR.
[41]	Fotso Kamga, G.A., Bitjoka, L., Akram, T., Mengue Mbom, A., Rameez Naqvi, S. and Bouroubi, Y. (2021) Advancements in Satellite Image Classification: Methodologies, Techniques, Approaches and Applications. International Journal of Remote Sensing, 42, 7662-7722.[CrossRef]
[42]	Caballero, C.B., Ruhoff, A. and Biggs, T. (2022) Land Use and Land Cover Changes and Their Impacts on Surface-Atmosphere Interactions in Brazil: A Systematic Review. Science of the Total Environment, 808, Article ID: 152134.[CrossRef] [PubMed]
[43]	Wu, M., Schurgers, G., Ahlström, A., Rummukainen, M., Miller, P.A., Smith, B., et al. (2017) Impacts of Land Use on Climate and Ecosystem Productivity over the Amazon and the South American Continent. Environmental Research Letters, 12, Article ID: 054016.[CrossRef]
[44]	Tshitende, E.K., Onema, J.M.K., Gumindoga, W., Tshimanga, R. (2022) Analysis of the Changes in Landcover, Streamflow, and Sediment Yield in Kasai Catchment of the Congo River Basin. IAHS-AISH Scientific Assembly 2022, Montpellier, 29 May-3 June 2022, IAHS2022-7.
[45]	Aquilas, N.A., Mukong, A.K., Kimengsi, J.N. and Ngangnchi, F.H. (2022) Economic Activities and Deforestation in the Congo Basin: An Environmental Kuznets Curve Framework Analysis. Environmental Challenges, 8, Article ID: 100553.[CrossRef]
[46]	Egbe, A.M., Fonge. A.B. and Tabot, P.T. (2023) Dynamics of Land Use/Cover Changes and Plant Diversity in Tubah Sub-Division, Cameroon. Sustinere: Journal of Environment and Sustainability, 6, 251-270.[CrossRef]
[47]	Djiongo, B.J.E., Desrochers, A., Avana, M.L.T., Khasa, D., Zapfack, L. and Fotsing, É. (2020) Analysis of Spatio-Temporal Dynamics of Land Use in the Bouba Ndjidda National Park and Its Adjacent Zone (North Cameroun). Open Journal of Forestry, 10, 39-57. [Google Scholar] [CrossRef]
[48]	Liliane, M.D., Romain, N.J., Caroline, M.S.M., Abigaelle, T. and Loic, T.T.A. (2022) Quantifying Forest Loss in the Mbalmayo Forest Reserve (Center Region, Cameroon). Journal of Geoscience and Environment Protection, 10, 271-288. [Google Scholar] [CrossRef]
[49]	Álvarez-Berríos, N., L’Roe, J. and Naughton-Treves, L. (2021) Does Formalizing Artisanal Gold Mining Mitigate Environmental Impacts? Deforestation Evidence from the Peruvian Amazon. Environmental Research Letters, 16, Article ID: 064052.[CrossRef]
[50]	Mahmoud, M.I., Campbell, M.J., Sloan, S., Alamgir, M. and Laurance, W.F. (2019) Land-Cover Change Threatens Tropical Forests and Biodiversity in the Littoral Region, Cameroon. Oryx, 54, 882-891.[CrossRef]
[51]	Zekeng, J.C., Sebego, R., Mphinyane, W.N., Mpalo, M., Nayak, D., Fobane, J.L., et al. (2019) Land Use and Land Cover Changes in Doume Communal Forest in Eastern Cameroon: Implications for Conservation and Sustainable Management. Modeling Earth Systems and Environment, 5, 1801-1814.[CrossRef]
[52]	Ernst, C., Mayaux, P., Verhegghen, A., Bodart, C., Christophe, M. and Defourny, P. (2013) National Forest Cover Change in Congo Basin: Deforestation, Reforestation, Degradation and Regeneration for the Years 1990, 2000 and 2005. Global Change Biology, 19, 1173-1187.[CrossRef] [PubMed]
[53]	Ewunetu, A., Simane, B., Teferi, E. and Zaitchik, B.F. (2021) Land Cover Change in the Blue Nile River Headwaters: Farmers’ Perceptions, Pressures, and Satellite-Based Mapping. Land, 10, Article 68.[CrossRef]
[54]	Gondwe, M.F.K., Chirwa, P.W.C., Cho, M.A., Geldenhuys, C.J. and Tshidzumba, R.P. (2022) Key Drivers of LULCC in Co-Managed and Governmentmanaged Forest Reserves in Malawi (1999-2018). International Forestry Review, 24, 1-14. [Google Scholar] [CrossRef]
[55]	Assede, E.S.P., Orou, H., Biaou, S.S.H., Geldenhuys, C.J., Ahononga, F.C. and Chirwa, P.W. (2023) Understanding Drivers of Land Use and Land Cover Change in Africa: A Review. Current Landscape Ecology Reports, 8, 62-72.[CrossRef]
[56]	Landis, J.R. and Koch, G.G. (1977) The Measurement of Observer Agreement for Categorical Data. Biometrics, 33, 159-174.[CrossRef] [PubMed]