Change Detection in Tree Species Composition and Structure in Meidob Hills, Northern Darfur State, Sudan

Abstract

This study was conduct in the Meidob Hills area of El-malha locality about 170 km north El-fasher, North Darfur State. The main objective of the research was to study and describe the quantitative changes in tree composition, structure and species richness as well as to explore the impacts of anthropogenic and environmental factors. In addition, the study attempted to detect the nature of changes affected in structure and composition of the tree species as a consequence of both increased human exploitation and microclimatic effects during the period (1995- 2010) using remote sensing. The ecological field data for the present work were collected during the period 2010 to 2011. A pilot survey was first conducted to gather enough information which has been used to decide upon the sampling procedure and intensity. Based on the results of the pilot survey, a full inventory was 5 then designed and executed using a stratified sampling procedure. Thirty four 34 sample plots of one ha sample size each (56.43 m radius) were laid down. Subplots of 0.25 ha (28.21 m radius) were laid down in the center of each sample plot for regeneration Measurements were made to record diameter and studies. height of all trees having a diameter ≥ 6 cm within each sample plot. A regeneration count was then conducted inside the subplots. The findings revealed that the total sampled population in the study area amount 922 trees, comprising ten species within seven families giving a stocking density of 27 trees per ha. The regeneration count registered a total number of 13 seedlings per ha for all species. Species wise analysis indicates Acacia tortilis subsp. raddiana was the most ecologically dominant species contributing a high importance value index (IVI) of 92.21 %. The landsat TM image of April 1995 and June 2008 were analyzed using Erdas Imagine software to unveil tree species change detection of the study area. Visual interpretation and digital image processing was applied to process the imagery for determining land cover classes. In addition, post- classification change detection methods were used to detect changes in land cover classes. After performing unsupervised classification on these images, a total of four land cover classes were identified and 6 mapped. These were Hills and basement, sparse trees and shrubs, bare land and grassland. An NDVI calculation was made on these images and vegetated areas were identified. The results of the analysis revealed that from 1995 to 2008 the tree cover has decreased by 3847 ha representing 10 % of the total area. These changes were attributed directly to environmental and anthropogenic factors viz. desertification, drought, illegal cutting of trees for fire wood, building material, agriculture expansion and overgrazing. The field work and remote sensing data were supported by a socioeconomic study through which data collected by using a questionnaire. Information including major indicators for degradation, reasons behind land degradation, useful tree species in the area, and major causes for forest destruction in the area were gathered from local people. Lessons learned from the use of satellite imagery data and remote sensing techniques in the present work proved to be effective, reliable, suitable and valuable tool for detection, prediction and identification of forest cover change in such degraded areas of arid and semi arid regions under risk.