Habitat suitability modeling of African forest elephant (Loxodonta cyclotis) in Omo Forest Reserve, Ogun State, Nigeria

Abstract

Maximum entropy (MAXENT) algorithm was used to explore how environmental variables influence habitat of African forest elephant in Omo Forest Reserve. The occurrence data and the predictor variables were processed using Geographic Information system software. The survey was conducted for both wet and dry season. Result shows that distance to river, mean diurnal range, distance to farmland, distance to road, and NDVI contributed mostly to predicting the habitat suitability for both dry and rainy seasons. There were several streams that flows within the forest reserve which reflects in elephant movement pattern towards riverine areas. Suitable habitat within the study area were found to be areas with dense vegetation. Elephant frequent movement around agricultural and cultivated areas within the reserve resulted in crop raiding and destruction. The model combined result for the study area suitability shows a suitable area of 332.90km², moderate area of 434.47km² and unsuitable area of 529.65km2. The performance ratings of AUC values (area under receiver operating curves) were 0.779 and 0.781 for dry and rainy seasons respectively. Due to increasing anthropogenic activities, elephants’ activities were greater in the northwestern part of the forest reserve which provides suitable habitat that meets food, water and vegetation cover requirement for the species.