Assessing the distribution pattern of elephant (Loxodonta africana) carcasses in Etosha National Park and its implications for management

Abstract

Mortality is an important component of understanding elephant population dynamics and disease ecology. We use carcass location data from Etosha National Park, Namibia, collected under the CITES Monitoring the Illegal Killing of Elephants programme, to assess the spatial distribution of elephant deaths, as identified through automated surface feature selection, and explore implications for park management. Carcass location data were modelled using a regression spline framework, with targeted flexibility, a spatial term and additional environmental covariates (annual rainfall, distance to water and roads). The novel modelling approach chosen acknowledges the localised and patchy distribution of carcasses and recognises physical barriers (Etosha pan) to substantially reduce the risk of false conclusions about the location of elephant deaths in the park. Our results showed high carcass intensity close to waterholes (< 2.5 km) and roads (< 5 km) and in areas of the park with average rainfall (~450 mm annually). Some high-risk areas were identified, particularly in the north-east of the park, and the mortality risk did not always coincide with elephant distribution. These findings are useful for understanding population dynamics and drivers for the park’s elephant population and park management, particularly for disease surveillance.