EVALUATION OF LAND-USE AND LAND-COVER CHANGES CUM FOREST DEGRADATION IN SHASHA FOREST RESERVE, OSUN STATE, NIGERIA USING REMOTE SENSING

Sustainable forest management requires accurate information on forest covers and periodical changes. Availability and use of spatial data has become very helpful in x-raying structural changes in most tropical rain forest management options due to reliable accuracy of satellite images with readily-available image processing tools. Thus, we assessed forest drains and land-cover changes in Shasha Forest Reserve using Landsat TM, ETM+ and OLI/TC images acquired from USGS. The images were analyzed using ERDAS Imagine and Maximum Likelihood Algorithm in ArcGIS 10.5. Land-use/cover change-dynamics were characterized using Land Change Modeller. Classification accuracies were assessed using Kappa’s and confusion matrix. Normalized Difference Vegetation Index analysis was performed in ArcGIS. The distinguished LULC were forest, farmland, built-up and water bodies/swamp. The forest cover and water bodies/swamp shrank by 638.91 and 1653.48 ha at -23.66 and -61.24 hayr-1, respectively within the 27-year period. Meanwhile, farmlands and built-up areas increased by 1508.40 and 783.99 ha at 55.87 and 29.04 hayr-1, respectively. Results further showed that the reserve was better-vegetated in 1991 (0.39) than 2018 (0.05). The major drivers of forest degradation in the area were subsistence agriculture, illegal timber exploitation and overexploitation of non-timber forest products. Overall classification accuracy was 95.1% with Kappa’s coefficient of 0.9439. User’s accuracy ranged between 90.8 and 96.2%, while Producer’s accuracies were between 90.8 and 98.1%. The result of LULC change simulation showed that if the prevailing trends subsist without regeneration, and if other factors remained unchanged, forest and water bodies/swamp would dip further by -653.74 ha and -318.66 ha by 2050, while farmlands and built-up areas would increase by 622.62 ha and 349.78 ha, respectively, with potential negative consequences on environmental variables