Land Use and Land Cover Change Analysis in Ajei Upland Watershed Community Forest, North West Region, Cameroon

This study was carried out in Ajei upland watershed community forest in the North West Region of Cameroon with the aim to analyze spatial and temporal dynamics of land cover and land use from 1988 to 2018 and to identify and characterize the agents, drivers and pressures of this change. The databases used were made up of 3 Lands at satellite images (5 TM of 1988, 7ETM+ of 2003 and 8OLI of 2018). Field survey, interview and focus group discussion methods were used to identify the activities carried out by the local population and to determine agents, drivers and pressures of land use and land cover change. The Ajei community forest is consisted of four main land cover namely; dense vegetation, sparse or degraded vegetation, savanna and bare soil. Between 1988 and 2018, the proportions of dense vegetation and savanna decreased and that of bare soil and sparse vegetation increased. Analysis showed ongoing deforestation with forest cover loss of about 240 ha in 30 years. For the considered time periods 1988-2003 and 2003-2018, deforestation rates were 1.83% per year and 1.49% per year respectively, signifying on average 8 hectares of forest cover undergoing conversion each year. This lost results from the actions of farmers, cattle grazers and wood extractors who through farming, grazing and lumbering activities pressurize and convert the dense forest cover. Economic motives notably the need to increase household income from a frequent demand for farm and wood products in nearby markets represent the drivers of forest cover change. Controlling grazing activities notably in the dry season to check out the use of fires, community sensitization especially among cattle headmen on the importance of the community forest, reforestation activities through natural regeneration or tree planting are needed in the forest. How to cite this paper: 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. https://doi.org/10.4236/gep.2018.69007 Received: July 30, 2018 Accepted: September 4, 2018 Published: September 7, 2018 Copyright © 2018 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/

methods were used to identify the activities carried out by the local population and to determine agents, drivers and pressures of land use and land cover change. The Ajei community forest is consisted of four main land cover namely; dense vegetation, sparse or degraded vegetation, savanna and bare soil. Between 1988 and 2018, the proportions of dense vegetation and savanna decreased and that of bare soil and sparse vegetation increased. Analysis showed ongoing deforestation with forest cover loss of about 240 ha in 30 years. For the considered time periods 1988-2003 and 2003-2018, deforestation rates were 1.83% per year and 1.49% per year respectively, signifying on average 8 hectares of forest cover undergoing conversion each year. This lost results from the actions of farmers, cattle grazers and wood extractors who through farming, grazing and lumbering activities pressurize and convert the dense forest cover. Economic motives notably the need to increase household income from a frequent demand for farm and wood products in nearby markets represent the drivers of forest cover change. Controlling grazing activities notably in the dry season to check out the use of fires, community sensitization especially among cattle headmen on the importance of the community forest, reforestation activities through natural regeneration or tree planting are needed in the forest.

Introduction
Forests are valuable resources providing enormous benefits. Forests regulate climate and fresh water flow, protect and enrich soils, control pests and diseases, maintain biodiversity, safeguard water quality, offer beautiful landscapes and enrich humans spiritually [1]. The cameroonian forest, covers an area of about 22,523,732 hectares, which represents about 48% of the national territory [2]. The importance of the forest is related to its multiple and sometimes conflicting uses and functions at local, national and global levels. From a conservation perspective, the forest constitutes a crucial reservoir of biodiversity, including many endemic species, and its contribution to climate regulation and other environmental services [3]. Regrettably, these socio-economic and ecological functions are under threat from high rates of deforestation [4]. According to [5], global forest cover has drastically decreased from 4128 million ha in 1990 to 3 999 million ha in 2015. In the Congo Basin in general, deforestation is increasing, with the rate of deforestation passing from 0.13% throughout 1990 to 2000 period to 0.26% in 2000 to 2005 period [6]. In Cameroon, [7] and [8] estimate the average annual rate of net deforestation at 0.14% for the period 1990-2000, with an average of gross deforestation in the order of 0.20%. With these figures, Cameroon appears as the second country in the Congo basin, after the Democratic Republic of Congo (0.2%), with the highest deforestation rate [7] [9]. Rapid urbanization, anthropogenic and socioeconomic activities, in local and regional levels, are important components responsible for extensive deforestation [1]. According to [10] and [11], in tropical countries the main drivers of degradation and deforestation are the combination of several factors, including the expansion of peasant farming and the extraction of wood.
The need for a healthy environment has given rise to the global fight against deforestation identified as the main problem affecting climate, biodiversity, ecosystems and the usefulness of the forest as a resource [12]. For many authors as [13], change detection analysis is important for planning and management activities for a wide variety of subjects such as the monitoring of deforestation.
To reduce deforestation, forests mapping and the monitoring of their evolution are very important. Mapping areas of forest cover change is essential for developing locally adapted strategies to better control these dynamics [14]. To carry out such monitoring, remote sensing is a less-expensive method that has proven its effectiveness for assessment of forest cover dynamics and degradation over several decades and at different scales [15] [16] [17] [18] [19]. Land use change detection is often conducted to critically discern trends, drivers and their impacts on the landscape to enhance decision making for of conservation biodi-versity and ecological functions [20].
In a bid to redress threats and impacts of forest cover lost, to promote community participation and alleviate poverty, the cameroonian forestry wildlife and fisheries law enacted in 1994 enshrined the concept of community forest corresponding to a maximum area of 5000 hectares, granting local community's access, to use and manage substantial portion of the non-permanent forest estate [21]. The Ajei community forest in Ngie subdivision of Momo Division, North West Region of Cameroon was attributed on the 14 th of January 2014 [22]. Its valorization stems from the valuable socio-economic and ecological services it offers. This community forest is an important watershed for the village and its neighbors as it supplies water to streams and drinking sources downhill. The riparian landscape of Ajei community forest is one of the few remaining forest patches of the Bamenda highlands and is a source of wood for energy and construction purposes, non-timber forest products (fruits, spices, medicines) for the population residing around the forest. Unfortunately, this forest is threatened by increasing human activities including agriculture, settlement, bush fires, encroachment by grazers [22]. Hence there is a need for a land use management plan. As such, information on land use and land cover status and dynamics of the forest, agents and drivers of this change are greatly needed. In Cameroon, very few studies on land cover and land use change focused on gallery forest and highland. This paper, therefore, initiated a remote sensing-based vegetation baseline assessment that is nonexistent in Ajei upland watershed community forest, as a strategy for informing policy makers involved in the sustainable governance and management of this forest. The objective of this study was to analyze the spatial and temporal dynamics of land use and land cover in theAjei upland watershed community forest and determine agents and drivers of the change. rugged topography and relief of Ajei village and its environs ranging between 1500 m above sea level cumulating at the highest peak of Ndek Akorbiri 2040 m is typical of an area underlain by granitic basement rocks. Soils derived from this parent rock material are generally poor, being rapidly depleted of fertility through cropping and the erosive action of rain [22].

Data Collection and Analysis
Assessment of land cover change was made using Landsat 5TM, Landsat 7ETM+ and Landsat 8OLI satellite images taken in 1988, 2003 and 2018 respectively. Imagery data for the study were downloaded from the United States Geological Survey (USGS) website (http://glovis.usgs.gov/). Images were acquired approx-imately for the same period, at the dry season (January) to ensure that they were clear and cloud free and that the phenological stages of plant cover were not too different between dates.
These different images were processed using ENVI 4.5 software to extract desired information. Landsat images were subjected to image enhancements activities such as pan sharpening, normalized difference vegetation index calculations, layer stacking and color composition activities were carried out to ameliorate the brightness, contrast and visibility of objects, to reduce omission and confusion errors to increase accuracy in spectral differencing of objects during classification. Extraction of the study zone was made using a geo-rectified vector map (UTM 32 WGS 84).
A hybrid image classification method consisting of unsupervised and supervised image classification was executed. Unsupervised classification was conducted to determine spectral differences in land cover classes for the creation of 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 Landsat scenes. Error matrixes to assess the accuracies and errors of the classification were computed. The Kappa index (the ratio of the number of well-ranked pixels to the total of the pixels surveyed) was calculated.
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 were imported into Arc Map 10.2 software for the extraction of the layers to be used, digitization, generation of databases, and production of maps. Figure   2 shows the different phases of image processing.
To identify the agents and drivers and pressures of change, the DPSIR (drivers, pressure, state, Impact and response) frame work approach designed by the European Union alongside field observations and questionnaires were used. The

Forest Status and Dynamics
From the hybrid image classifications of Landsat scenes of 1988, 2003 and 2018, four main land cover classes were identified namely; dense vegetation, sparse or degraded vegetation, savanna and bare soil (Table 1) Figure 4) showed a dominance of sparse vegetation with a surface area of 539 ha representing 31.03% of the forest.
In2018, bare soil was dominant with a surface area of 519 ha representing a 29.84% occupation of the total forest area ( Figure 5). Table 1 (Table 3) revealed an overall confusion of less than 15% with overall accuracy and degree of representation of real world ground truths (Kappa index) of more than 85% and more than 0.8 respectively. With these levels of errors and accuracies the image classifications were found acceptable given the fact that, in all three cases, sources of confusions are between classes almost possessing the same spectral characteristics that is between dense and degraded vegetation and between savanna and bare soils classes. Pests and diseases also constitute natural agents at work in the forest, these agents attack trees, provoke their death and decay.

Drivers of Land Use Change
Sectorial drivers shaping gallery forest landscape in Ajei community forest were identified and grouped into; economic, socio cultural, governance and communication drivers.
-Economic driver This driver's indirect influence on the forest results the need to step up household income from high and frequent demand for agriculture and wood  products. This high and frequent demand comes from an increase in the number of markets and market days set out for the buying and selling of wood and agricultural products within a week around Ajei. These markets include the Ajei mixed market, Andek, Mbengwi bone, Nyen-bo, batibo, Bamenda food markets.
There is also the Tugi cattle market situated within the heart of Ngie Subdivision purposely functioning for the buying and selling of cattle. Many carpentry workshops and buyers of wood also exist within the listed neighborhoods above.
Respondents revealed the persistent arrival of many buyam-sellams from other areas notably those from Bamenda, Douala, Yaounde and Bafoussam who flood these markets and sometimes buy through command before the market days.
Local inhabitants around having to look for end means to satisfy this high fre- where to better sell, in what quantities and prices.

Pressures of Land Use Change
Field observations show that the forest is pressured from every front, which is from downhill to uphill and from uphill to downhill directions and inside the forest by both anthropogenic and natural pressures.

Discussion
The Kappa indexes of 0.82, 0.84 and 0.88, respectively for the three images, prove that the classifications performed are reliable if we refer to the scale of [23]. Results on land cover status and overall forest cover change show an ongoing deforestation process within the community forest. spectively. This lower rate of deforestation for the second period is probably due to a decline in human activities in the community forest following its attribution in 2014 and a better sensitization of population. The rates of deforestation observed in our study area are higher than that of 0.65% reported by [33] for 1990-2000 period in Eastern Region of Cameroon and also greater than that of 0.13% found by [6] for the Congo basin for the same period. In Cameroon, [7] and [8] reported a deforestation rate of 0.2%. The rates found in our study area are higher.
In Ajei community forest, the conversion of the natural densevegetation into sparse vegetation and bare soils area is a consequence of the total sum of actions led by farmers and cattle headers leading farming and grazing activities on the forest. Wood extraction practices are limited and not in advanced state in this community forest. Similar drivers are found in other part of Cameroon by [26] [27] [28], respectively in Santchou, Koupa Matapit and Mount Cameroon caldera. [26] reported a Market-oriented agriculture based on cocoa and coffee as the main driving force of forest cover changein Santchou forest reserve. In Asia, an assessment on forest cover dynamics in the tropical deciduous forest in Kalrayan hills, in India highlighted illegal lumbering, forest fires and shifting cultivation as leading causes to land use and land cover change [34]. These findings do clearly reveal that agricultural expansion generally affects land cover conditions and status and is at the center of most forest lost registered in the world.
Development in areas of infrastructures and poor governance do also have a role to play. In Ajei community forest, a dense network of roads and tracks of about 25 km facilitate movement of goods, persons and cattle, in and out of the forest patches and can further intensifies the state of fragmentation in this forest.
Though there are now well-developed land use policies and practices to protect forest areas, the management board of the community forest still faces difficulties in the management of the forest. Difficulties to initiate and to scale up community participation notably on rational harvesting and exploitation of wood products, rational and intensive use of land still pose problems. These proximate and underlying drivers of deforestation have been reported by [10] for Congo Basin.

Conclusion
The trend of the change observed in Ajei community forest was a conversion of the natural dense forest into sparse vegetation and bare soils. The rate of deforestation is higher than that reported in the Congo basin and in Cameroon. The