Use of remote sensing in landscape-scale vegetation degradation assessment in the semi- arid areas of the Save catchment, Zimbabwe.

Abstract

The deteriorating condition of land in parts of the world is negatively affecting livelihoods, especially, in rural communities of the developing world. Zimbabwe has experienced significant vegetation cover losses, particularly, in low and varied rainfall areas of the Save catchment. The concern that Save catchment is undergoing huge vegetation losses has been largely expressed, with the causes being environmental and anthropogenic. Given the magnitude of the problem, research studies have been undertaken to assess the extent of the problem in the south eastern region of Zimbabwe, which, nevertheless, have been mainly localized. The present study seeks to identify and quantify vegetation degradation at a landscape scale in the Save catchment of Zimbabwe, using remote sensing technologies. To achieve this, two objectives were set. The first objective provided a review of the application of satellite earth observations in assessing vegetation degradation, the causes, as well as associated impacts at different geographical scales. A review of literature has revealed the effectiveness of satellite information in identifying changes in vegetation condition. A second objective sought to establish the extent of vegetation degradation in the Save catchment. Moderate Resolution Imaging Spectroradiometer- Normalised Difference Vegetation Index (MODIS NDVI) datasets were used for mapping NDVI trends over the period 2000-2015. Further analysis involved application of residual trend (RESTREND) method to separate human influences from climatic signal on vegetation degradation. RESTREND results showed an increasing trend in NDVI values in about 33.6% of the Save catchment and a decreasing trend in about 18.3% from 2000 to 2015. The results of the study revealed that about 3,609,955 hectares experienced significant human induced vegetation degradation. Approximately 38.8% of the Save Catchment was significantly degraded (p< 0.05), 3.6%, 12.8%, and 22.4% of which were classified as severely, moderately, and lightly degraded, respectively. Severe degradation was mainly found in the central districts of the Save Catchment, mainly Bikita, Chipinge and northern Chiredzi. The results of this study support earlier reports about ongoing degradation in the catchment. Vegetation changes observed across the landscape revealed different degrees of the impacts of land use activities in altering the terrestrial ecosystems. The study demonstrated the usefulness of the RESTREND method in identifying vegetation loss due to human actions in very low rainfall areas.