A group of elephants drinking from waterholes
in Dzanga Bai, Central African Republic

A forest elephant in the Central African rainforest.

What is the Elephant Listening Project?

The Elephant Listening Project (ELP) was founded in 1999 with the primary focus on African forest elephants, a unique species (Roca et al. 2001) which lives in deep equatorial rainforests where sightings are rare and visual censusing is impossible. In the Bioacoustics Research Program of the Laboratory of Ornithology, ELP is creating an acoustic monitoring system which uses elephants’ vocal patterns as indicators of the size and composition of their populations. Elephants make powerful infrasonic calls (below the level of human hearing) which travel long distances, allowing researchers to identify the presence of elephants over large areas without visual sightings. Our current efforts include:

A spectrogram of a series of eight low-frequency forest elephant calls. The line represents the lower threshold of human hearing. Listen to these calls>>.

A forest elephant, up close!

What can we learn from elephant calls?

The rate of elephant vocalizations in acoustic recordings is directly related to elephant abundance (Payne et al. 2003, Thompson et al. in prep-a). This relationship can be used to estimate population size in forested habitats. A recent study in Kakum National Park in Ghana (Thompson et al. in prep-b) found that population estimates based on acoustic monitoring were more precise than those made using genetic sampling of dung ( mark-recapture DNA methods; Eggert et al. 2003) and dung counts (Barnes 2000), with a confidence interval half the size of other methods (Thompson et al. in prep-b). Unlike existing censusing methods, acoustic monitoring provides continuous information on the spatial and temporal patterns of elephant activity in forests. Furthermore, the effects of habitat characteristics such as forest cover, understory vegetation, and other physical barriers is negligible (Wiley and Richards 1978, Garstang 2004). Thus, acoustic monitoring can be used to determine migration patterns, hotspots of elephant activity, and responses to rapid environmental or habitat changes resulting from logging, mining, or human settlements. Recordings from elephant habitats also provide information on human activities, including gunshots, chainsaws, and vehicle noise which helps to identify potential threats to forest elephant populations.

In addition to information on elephant abundance, Mya Thompson, a PhD student with ELP is working to classify elephant calls based on demography (the sex and age class of elephants) and behaviors (such as reproductive condition and level of excitement) by incorporating quantifiable measures of elephant calls into "classification trees", a statistical procedure only recently applied to vocal classification (Van Parijs et al. 2003). Mya is using existing data, including acoustic recordings and focal observations of forest elephants conducted at Dzanga forest clearing in Central African Republic during 2002 and 2002. Her work will aid in determining if population structure and health can be assessed from acoustic recordings in areas where elephants cannot simultaneously be observed.

Why is this research important?

Forest elephants in Asia and Africa are increasingly endangered by poaching for ivory and bushmeat, habitat loss, and the constraints of current methods used to monitor forest elephant populations (Blake & Hedges 2004). Estimates of this newly identified species in Central Africa vary by more than four-fold from 16,450 to 82,563 individuals (via dung counts; Blanc et al. 2003, Blake & Hedges 2004) creating tremendous difficulty in identifying priority areas for conservation efforts. Improvements to monitoring will allow evaluation of current conservation efforts and provide information critical to the future of this species. Poaching for bushmeat and trade in animal parts is not limited to elephants, but is one of the greatest threats to the persistence of tropical wildlife (Robinson et al. 1999). An effective gunshot monitoring system that could identify hotspots of poaching activity could be employed in tropical forests throughout the world, not only in Africa.