Spatio-temporal evaluation of land cover change and habitat fragmentation using geospatial techniques and landscape metrics. The case of Akagera National Park, Rwanda

Abstract

Several studies on the park environmental monitoring have mentioned that Akagera National Park area and size has been reduced (African Parks, 2016). Accurate and up-to-date land-use and land-cover maps produced using geospatial techniques and landscape metric are proven to be important inputs to biophysical and environmental assessment models required for decisionmaking and resource planning, monitoring and evaluation fragments (Zhang and Johnson, 2017). The aim of the study is to undertake spatio-temporal evaluation of Akagera National Park land cover change and ecosystem fragmentation using geo-spatial techniques and landscape metrics. The first specific objective was to describe and map the LCU types of the Eastern savanna ecosystem and define habitat for specific flora and fauna. With help of maximum likelihood classifier on time series Landsat imagery land cover/use types maps of Akagera National Park during last three decades (1990, 2005, to 2020) were generated and further used to detect the changes of main LCU between 1990 and 2020. The second specific objective was to analyse the impact of land covers/uses changes on park fragmentation during the study period. GIS techniques were also used for landscape metrics calculation using FRAGSTATS software. The results revealed that from 1990 to 2005; grassland/savannah area, water and wetland decreased by -3561.98ha, -442.30ha and -2435.83ha respectively with a rate of change of -237.47ha/year, -29.53ha/year and -162.39ha/year respectively. Whereas, bare land increased to 4582.190 ha with a rate of change of 305.48ha /yr. Shrub/forested land also shows significant changes with an increase of 1858.6 ha with a rate of change of 123.91ha/year. In 2020, 57% of the study area covered by grassland area which was increased to 4829.678038 Ha in 2020 with 321.98ha/year while area occupied by water has increased to 4296.450ha with 286.43ha/year. On the other hand, the land use land cover categories like bare land, wetland and forested land showed decreasing pattern with -455.24ha/year, -128.15ha/year and -25.02ha /year of average rate of changes respectively. Considering the overall study period (30years), there was a remarkable increase in a real extent of grassland/savannah area of land from 60274.2ha (55.8 %) in 1990 to 61541.9ha (57%) in 2020. From landscape analysis at class level, landscape level and patches level, the results indicated that the values of overall patches decreased proportionally from 28509 to 13054. Since changes are dynamic in time and space, the continuous monitoring of land covers/uses changes and their impacts on habit fragmentation are recommended.