Estimating the Extent of Degradation in the Bounfum Forest Reserve, Ghana, Using Historical Remotely Sensed Data and Landscape Fragmentation Indices

Abstract

Land use and land cover changes, especially deforestation and forest degradation and its driving factors, are key factors hindering sustainable forest management. Currently, there is limited knowledge concerning the detection of the extent and interpretation of the spatial and temporal pattern of forest cover dynamics in the Bounfum Forest Reserve, which when available will inform sustainable policies. Using the Landsat TM image of 1986, Landsat ETM+ image of 2002 and Landsat 8 OLI image of 2014, the study identified and quantified the forest cover dynamics in the Bounfum Forest Reserve from 1986 to 2014. The ERDAS maximum likelihood classification algorithm was used to classify the pixels into five major land cover classes namely, bare/built areas, farmlands, closed forest, open forest and shrub/grassland. The Kappa coefficients of 0.83 (1986), 0.72 (2002) and 0.75 (2014) respectively were obtained for the classified images. The findings showed that the closed forests decreased by 3.5% (563.90 ha) per annum whilst the open forests and farm lands increased by 19.5% (385.60 ha) and 2.9% (65.00 ha) per annum within the 28-year period. This implies that the Bounfum forest reserve has been highly degraded over the past 28 years, evident through the trends of its patch densities and the number of patches. Collaborative forest management is required in the management of the forest reserve to conserve the socio-ecological and economic benefits derived from the resource on sustainable basis.

Keywords: Land use and land cover change, Bounfum forest reserve, deforestation, forest degradation, remote sensing, sustainable forest management