Monitoring population decline: can transect surveys detect the impact of the Ebola virus on apes?

In 2004 the Ebola virus caused a drastic decline in western gorilla Gorilla gorilla abundance at Lokoue´ Bai, a clearing in Odzala National Park, Republic of Congo. This decline was detected by observations of gorillas visiting the clearing. We confirm that the sympatric chimpanzee Pan troglodytes population was also affected by the Ebola outbreak, and test whether the decline in the
ape population would have been detected with linetransect surveys, the most commonly used wildlife monitoring methodology in Central Africa. We also evaluate the potential of transect surveys for describing the extent and pinpointing the timing of drastic population declines when this information is not known from other evidence. Both nest survey using the marked nest count method and standing stock survey of other signs of ape presence (dung, feeding remains, prints) were able to detect the decline. However, only nests and dung were reliable indices for estimating the magnitude of the decline and accurately pinpointing the timing. It was necessary to pool data across many survey replicates because of small samples sizes. Our results suggest that transects methods are able to detect drastic changes in ape abundance but that large sample sizes are necessary to achieve adequate statistical power. We therefore recommend that those intending to use transect methods as tools for monitoring large forest mammals evaluate in advance how much effort will be necessary to detect meaningful changes in animal abundance.