How to count elephants in West African savannahs? Synthesis and comparison of main gamecount methods

The comparison of some elephants count methods – aerial sampling count, aerial total count and direct foot count – has been presented through surveys' study cases implemented in Benin, Burkina Faso and Niger. We compare them in terms of effort and cost to survey an area (sampling efficiency), efficiency in data collection (detection efficiency) as well as accuracy and precision. Aerial sampling count returns low sampling and detection efficiency and very low precision and accuracy figures when counting elephants populations below 0.5 elephant·km-2. Precision can partly be improved by stratification. Direct foot count returns the best sampling and detection efficiency but insufficient data number to produce reliable estimates. Therefore some authors often produced no results for elephants. Aerial total count is a costly method in absolute value. It requires important logistic and no measure of error is possible. However it provides relatively good sampling and detection efficiency as well as relatively accurate figures appreciated by wildlife managers if technical requirements are respected. Aerial total count could be recommended to survey small population of West African elephants each 3 to 5 years to minimize yearly count costs.