Abstract

Predictive and GIS modelling for roan antelopes at multi-spatial scales
Track: Parks, Natural Reserves, Fisheries, and Wildlife Management
Authors: Johnstone Kimanzi, Jennifer Wanyingi

GIS modelling provides a powerful tool for understanding species-habitat relationships. This study evaluates the strength and usefulness for conservation purposes of habitat suitability models developed using GIS and information theoretic approach for the endangered roan antelope in Ruma National Park, Kenya. It investigates the relationship between GIS derived habitat features and the presence/absence of roans at the park boundary and homerange levels.

Results showed that vegetation type, surface water and slope were the best predictors of habitat suitability for roans at the boundary level whereas slope, surface water, soil type and distance to poaching hotspots were the most important at the homerange level. The park boundary models had more predictive power and explained more variation in the data than the homerange - based models. The habitat suitability models and maps produced by this study were used in formulating a national roan conservation strategy that is currently being implemented in Kenya.