By Nicholas St. Fleur
With their trumpet-like calls, elephants may seem like some of the loudest animals on Earth. But we can't hear most of the sounds they make. The creatures produce low-frequency noises between 1 to 20 Hertz, known as infrasounds, that help them keep in touch over distances as large as 10 kilometers. A new study reveals for the first time how elephants produce these low notes.
Scientists first discovered that elephants made infrasounds in the 1980s. The head female in a herd may produce the noises to guide her group's movements, whereas a male who’s in a mating state called musth might use the calls to thwart competition from other males. Mother elephants even rely on infrasounds to keep tabs on a separated calf, exchanging "I'm here" calls with the wayward offspring in a fashion similar to a game of Marco Polo. These noises, which fall below the hearing range for humans, are often accompanied by strong rumbles with slightly higher frequencies that people can hear. By recording the rumbles and then speeding up the playback, the scientists can increase the frequency of the infrasounds, making them audible.
Researchers have speculated that the noises come from vibrations in the vocal folds of the elephant larynx. This could happen in two ways. In the first, called active muscular contraction (AMC), neural signals cause the muscles in the larynx to contract in a constant rhythm. Cats do this when they purr. The second possibility is known as the myoelastic-aerodynamic (MEAD) method, and it occurs when air flows through the vocal folds causing them to vibrate—this also happens when humans talk. The AMC is like a doorbell or buzzer that requires electricity, or in this case a brain signal, to work, while the MEAD is similar to a woodwind instrument such as the clarinet, which only needs airflow to produce sounds. Neural signals are only needed in this case to control breathing, but not produce the sound.
Until now, field researchers were unable to test which method was behind the infrasound. “You cannot just walk up to an African elephant and stick an endoscope in his mouth and ask him to say 'Ahh,' " says voice scientist and lead author Christian Herbst of the University of Vienna. So his team did the next best thing. They obtained an intact larynx from a female African elephant (Loxodonta africana) that died of natural causes at the Berlin Zoo.
With only the excised larynx, the researchers could not test for the AMC method because it requires neural messages for the muscles to contract. To test the MEAD hypothesis, they mounted the larynx on a tube connected to an air tank that acted as a psuedolung, regulating airflow and pressure. The psuedolung was also connected to a heater and humidifier, which simulated the natural conditions of air in elephant lungs. Using high-speed cameras, the researchers found that the vocal folds vibrated as the air passed over them (see first video). They also recorded the sound from the vibrations and compared it with a database of more than 470 live elephant calls collected over 4 years. This analysis revealed that the excised larynx produced sounds similar to the low-frequency calls of live elephants (see second video).
The results, published in tomorrow’s issue of Science, showed that the MEAD mechanism alone can produce infrasound in elephants and that the AMC mechanism was not necessary. Additionally, the team observed that the vocal folds often vibrated in a nonlinear pattern, which would be expected in the MEAD method. AMC vibrations occur in regular patterns and produce a sound like singing a single note. MEAD vibrations, on the other hand, don’t follow a particular pattern and sound rough like a heavy metal singer screaming. The findings may reveal how a variety of other animals make noise, says Herbst.
“It’s a nice elegant paper that does a really good job of showing that the easiest explanation for how elephants make infrasound is through the way that air passes over the vocal cords in their larynx,” says William Langbauer, a bioacousticist and the director of the Buttonwood Park Zoo in New Bedford, Massachusetts. Mya Thompson, an ecologist with Cornell University’s Elephant Listening Project, agrees. “We know for sure that this myoelastic mechanism is working, and it clears things up for future research on the structure of calls,” she says.
ScienceNOW, the daily online news service of the journal Science