Remote sensing of waterhole distribution and its impact on wildlife and livestock interactions across a gradient of land use in a Southern African savanna

Abstract

Arid and semi-arid ecosystems are water limited ecosystems. Thus, understanding the spatial and temporal dynamics in water availability is critical. In this study, we tested whether and to what extent we can use remotely sensed data to estimate the spatial and temporal distribution of surface water based on a semi arid region in the South-East Lowveld of Zimbabwe. Next, we tested the extent of interaction between wildlife and domestic herbivore at waterholes based on four years of data. Results show that remote sensing derived MNDWI can successfully (Kappa= 94%) be used to map surface water in semi arid environments. Results also showed that surface water is significantly predicted from annual rainfall (r2=0.98, p<0.01 for the wet season and for the dry season (r2 =0.96, P < 0.01). Furthermore, results show evidence of interaction between wild and domestic herbivores but mainly at waterholes located at the boundary of the protected areas and in the agricultural areas. Finally, results of this study demonstrate that animal species diversity declines significantly along a gradient from protected areas to agricultural areas. These findings could provide an important avenue for undestanding livestock and wildlife interactions in African savannas.