What depths will elephants go to for a drink of mineral water?
A party of elephants, diving for minerals in the Mpassa River. © Elephant Listening Project

Elephant Listening Project

June 23, 2014

We’re sorry it’s been so long since we sent out a newsletter, but as you’ll see, we’ve been busy!

Trekking through the Congo

Peter was in the Republic of Congo for a two-month field trip deploying 27 recording units over a 2,500 sq. km area of remote rainforest. The goal was to see if our acoustic method is a more efficient and accurate way of estimating elephant populations than the traditional dung-transect method (a dung-transect was carried out there shortly after our recording units were put up). Peter and a team of about 5-7 Congolese guys went on three treks that each started with a long ride in a pirogue down a river.

They were dropped off at a certain point along the river (different for each trek), from where they started to make their way on foot through incredibly dense understory vegetation in the rainforest putting up recording units as they went. They emerged from the forest after seven or more days (one trip took about 15 days) on one of the rare roads in the area, where they were picked up and taken back to the launching point on the river, so they could have the pleasure of repeating the performance! They had to literally hack a narrow path through the understory with a machete nearly all the way.  The temperature and humidity were high, the bugs were bad, the food was minimal, and the water was surprisingly scarce – sometimes they had to dig for it! They did get to see chimps and elephants, though at rather too close range: the chimps raced through their camp, and the elephants charged them! Peter usually loves field work, but this trip was a bit much even for him!  Luckily, for Peter, one of the Congolese guys agreed to lead a team to pick up the recording units several weeks later. We’re hoping to get the sound data back soon, so then the analysis will begin.

Rumble detector

Back here in the office, we’ve been analyzing data and improving our methodology. We’re testing an automatic detector (designed by our colleagues who are software engineers) that identifies elephant rumbles in our sound files. The results so far are very promising. This detector could reduce the time it takes us to analyze a six-month recording from several weeks to just a few days. Consequently, we’ll quickly be able to get the results about both elephant and poaching (gunshot) activity out to the people in Africa who need it to make effective management decisions.

Diving elephants

For the last school year, we’ve had the pleasure of having two high-school seniors, Catie and Sylvie, from the Tompkins Seneca Tioga New Visions Program working with us. We challenged them to investigate how elephant activity varied seasonally at a place on the Mpassa river in Gabon where we had a recording unit. When the river is low, elephants wade into the river and stand in certain places for quite long periods with their trunks reaching down, apparently drinking from a source on the river bed.

We presume there are springs with mineral-rich water that open out on the river bed, and that the elephants are drinking this mineral-rich water. When the river is high, the elephants swim into the river, then dive down to access the springs. Sylvie and Catie tested the hypothesis that in the wet season fewer elephants would visit this place in the river because only the larger ones would be able to dive down in the deeper river and hold their position in the stronger current long enough to obtain the minerals. Also, even larger females may not go the river in the wet season if they have infants that would not be able to follow them into the river.

Catie and Sylvie used our sound analysis software to count elephant rumbles in recordings made there over several years. Our previous work has shown that the number of rumbles is directly related to the elephants. They found that there were indeed significantly more elephant rumbles –and therefore more elephants - in the dry season compared to the wet. So, thanks to Catie and Sylvie, we now know the answer to one of our questions. As they go off to college, we’re hoping they become one of the next generation of scientists that will continue to build on our knowledge of forest elephants and thus help to ensure the survival of this species.

Training Gabonese

Peter’s now in Gabon, teaching a Gabonese colleague, Toussaint, how to use recording units, including how to use climbing gear to hang them in trees (hopefully Toussaint will enjoy this as  much as a previous student). I’ll be following Peter next week to teach Toussaint how to use our sound analysis software. I hope to also see an elephant!

We’ll try not to let such a time lag before we write again!

Liz Rowland
Research Specialist, Elephant Listening Project

The Elephant Listening Project is dedicated to the study and conservation of elephants, with a focus on the forest elephants of Central Africa. Visit the project's website at

Copyright © 2012 Cornell University, all rights reserved.
Questions or comments? CONTACT US

You are receiving this newsletter because you requested it directly or subscribed on our website. If we are sending this in error please accept our appologies and use one of the links below to unsubscribe.

Our mailing address is:
Cornell Lab of Ornithology
159 Sapsucker Woods Rd
Ithaca, NY 14850

Unsubscribe from the ELP eNewsletter


Unsubscribe from all Cornell Lab eNewsletters

Donate to the Elephant Listening Project to make a difference. Your gift supports the study and protection of endangered forest elephants.