ELEPHANTS have mysterious powers. A thunderstorm in Angola prompts animals in
Namibia to move north in search of water. A large elephant cull leaves another
herd 50 kilometres away tense and agitated. They must be picking up signals
somehow. Hearing or smelling events at such long range is out of the question,
so what鈥檚 going on?
Zoologist Lynette Hart thinks she has the answer. She is convinced that
elephants can talk to each other through the ground. That would make them the
only large terrestrial mammal we know that communicates using seismic signals.
Now Hart and her colleagues are starting to piece together the evidence. But
elephants can be a subtle and elusive bunch鈥攁nd watching them in action
provides many surprises.
The first real clue that the animals might be sensing distant vibrations came
in 1992 when Caitlin O鈥機onnell-Rodwell noticed some elephants in Namibia acting
strangely. 鈥淭hey would freeze, lean forwards, and lift up one foot,鈥� says
O鈥機onnell-Rodwell, who was then a student of Hart鈥檚 at the University of
California in Davis.
Advertisement
She felt the elephants鈥� behaviour was strangely familiar, and realised that
she鈥檇 previously seen insects doing the same thing when they were feeling for
rumblings. 鈥淚f they lift one leg up, they get better coupling to the ground with
the other three legs.鈥� But what might the animals be listening to?
When O鈥機onnell-Rodwell returned to the US, she told Hart about the enigmatic
behaviour she had seen. Intrigued, Hart talked it over with her brother,
geophysicist Byron Arnason. They wondered whether the elephants were using
seismic waves to carry their calls through the ground. 鈥淪he鈥檇 been talking about
the possibility for a long time,鈥� says Arnason. 鈥淚 was involved in oilfield
seismic exploration, looking for oil. Our minds melded together.鈥�
Thanks to Arnason鈥檚 expertise with seismic waves, they knew that some
elephant calls had the necessary characteristics to travel efficiently through
the earth. They have frequencies as low as 20 hertz, and they are loud. So
Arnason temporarily stopped looking for oil, gathered up his monitoring
equipment, and set off with the others to an elephant sanctuary in Texas to
listen in.
To test for vibrations, Arnason placed microphones 10 metres from the
elephants and also further away, at around 45 metres. Directly beneath each
microphone he buried a geophone, which measures ground vibration. Sound travels
at a different speed through earth than through air, so this would show whether
calls were being carried in the ground independently of those in the air. 鈥淲e
had to measure both the acoustic and seismic signals to show that they arrived
at different times,鈥� says Arnason.
The results, published last December in The Journal of the Acoustical Society
of America (vol 108, p 3066), showed exactly that. What鈥檚 more, by measuring the
rate at which they faded, the researchers also worked out that sounds travelled
much farther through the ground than through the air. They estimate that
elephant calls could travel around 16 kilometres through the ground. Vibrations
caused by mock charges鈥攖he sort of crashing around that elephants do when
they are nervous or afraid鈥攎ight travel even farther, up to 32 kilometres.
By contrast, calls can鈥檛 be heard more than about 10 kilometres away.
Why would elephants need these long-range seismic signals? 鈥淭he communication
is for coordination within and between herds of closely related individuals,鈥�
says O鈥機onnell-Rodwell, who is now at Stanford University. 鈥淭he herds move in
the same direction, even though they are not in eye contact.鈥� Elephants roam
over vast areas and seismic communication would extend the range over which they
could stay in touch.
But how might the animals pick up the seismic signals? One possible answer
came from a surprising quarter. When the researchers mentioned their ideas to
Bets Rasmussen, an expert in elephant chemical communication from the Oregon
Graduate Institute of Science and Technology, she had no doubts. 鈥淚 know they
can do this with their trunks,鈥� was her response.
Rasmussen鈥檚 conviction came from her work in India, investigating how
elephants use their highly developed sense of smell. Pheromones that signal
sexual status are particularly important, and elephants detect them using a
specialised organ in the roof of their mouths. Rasmussen wondered whether the
tip of the trunk also had chemical receptors in it. But when she searched the
literature she was amazed to find that nobody had ever studied the trunk tip.
鈥淚n some parts of South-East Asia, the tip is considered sacred,鈥� she says.
鈥淧eople cut it off and keep it as a good-luck charm. So when an elephant dies
that is the first bit to go鈥攖he scientists never get their hands on
颈迟.鈥�
To plug the knowledge gap, Rasmussen enlisted the help of anatomist Bryce
Munger from the University of Tasmania, who had previously studied the anatomy
of monkey snouts. 鈥淗e was about to retire, but I convinced him to do this one
last project,鈥� she says. 鈥淭hen he called me up and said, 鈥榊ou wouldn鈥檛 believe
it!'鈥� The trunk tip was the most sensitive tissue Munger had ever seen.
But it wasn鈥檛 picking up chemical signals. Instead, Munger found specialised
cells called Pacinian corpuscles which pick up vibrations. They comprise
concentric membranes of connective tissue, like the layers of an onion, with the
gaps between filled by a slimy gel. Movements or vibrations deform the layers,
sending a nerve signal to the brain. These corpuscles are found in other mammals
too, in human fingertips, for example, but in the elephant trunk they are
particularly densely packed. 鈥淭hey would definitely make it possible for
elephants to detect low-frequency vibrations in their trunk tip,鈥� says
Rasmussen.
The finding explains stories about elephants laying their trunks on water
pumps, as if to see whether they are switched on or not. But O鈥機onnell-Rodwell
and Hart were convinced this couldn鈥檛 be the whole story. What about that
strange foot-lifting behaviour they had observed? Surely it must hold another
clue.
鈥淓lephants have extraordinary feet,鈥� says Hart. 鈥淭hey stand on their toes.
Their toes rest on this big gooey ball of fatty tissue. It feels like a
waterbed.鈥� That鈥檚 why you never hear an elephant walking, she adds. Hart plans
to dissect elephant feet to see if she can find the same vibration-sensitive
cells that Munger found in the trunk. Vibrations in the ground would make the
dense fatty cushion of the foot undulate, stimulating any Pacinian corpuscles
within.
The idea that elephants are sensing the world through the soles of their feet
has got other researchers in the field talking. 鈥淧eople are really excited,鈥�
says O鈥機onnell-Rodwell. 鈥淚t answers a lot of mysterious questions for them about
what elephants do in the wild.鈥� For example, Iain Douglas-Hamilton, founder of
the Nairobi-based conservation group Save the Elephants, says the notion would
explain one strange habit he has witnessed. When elephants come across a dead
elephant, he says, they touch the body very delicately with their feet, as if
checking whether the animal is alive by looking for vibrations in the body. 鈥淚
get the feeling that they use the foot as a sensory organ,鈥� he says. 鈥淭hey are
like large plate-like scanners.鈥�
O鈥機onnell-Rodwell and Hart say that if elephants are also sending messages
through the ground deliberately, rather than simply sensing information, the
messages are probably quite simple鈥攎ore 鈥淚 am here, where are you?鈥� than
鈥淐ome down to the waterhole, there鈥檚 a party on Friday night鈥�. That鈥檚 because
the signals travelling through the ground are very plain. 鈥淭he seismic signal
has virtually no modification on it at all,鈥� says Arnason. 鈥淪o if they do
communicate seismically, there鈥檚 very little information they can put in. The
only specific characteristics are duration and frequency.鈥�
Arnason points out that it would be impossible for elephants to transmit
complex information through seismic signals, because any details would get
garbled. 鈥淲hen you propagate a pulse through the ground using surface waves, it
becomes very distorted,鈥� he says. And if you analyse elephant calls you find
that the high pitched ones such as trumpeting, which could not travel through
the ground, tend to carry more detailed information than lower pitched ones.
That fits with the idea that elephants intend some calls to travel
seismically.
Even without complex signals, elephants might be able to work out how far
away the caller is by comparing the arrival time of the signal in the air with
that in the ground. They could also tell the direction from which the call came,
depending on which foot felt it first, says Arnason.
The details of how elephants analyse sound signals remains a mystery. What is
clear is that they have plenty of brain power with which to do the complex
information processing involved. In work soon to be published in Animal
Behaviour, Hart compared the brains of different sized mammals, and calculated
how much of the brain would be required simply to run the body. 鈥淭he Asian
elephant has more cortex to spare than any other mammal,鈥� she says. 鈥淭hey have a
tremendous amount of brain power. Maybe they could use it to integrate all this
information coming from their trunk and feet.鈥�
In addition to detecting specific calls, elephants would also be able to use
their seismic sense to glean information about distant events. O鈥機onnell-Rodwell
believes they might be able to tell what mood another group was in, from how
often they called, for example, or by sensing any agitated mock charges. 鈥淚f
they can sense a herd running away from a waterhole, then they might guess there
are lions there,鈥� she says.
Then there are culls. 鈥淭here鈥檚 anecdotal evidence that when you have large
elephant culls, other herds are able to sense it 50 kilometres away,鈥� says
Arnason. That鈥檚 too far for the animals to be hearing or smelling anything, but
they could be picking up the vibrations from the helicopters used by the
hunters, or from stampeding herds.
Another mystery is how elephants seem to sense distant thunderstorms. 鈥淲hen
it rains in Angola, elephants 150 kilometres away in Etosha start to move north
in search of water,鈥� says O鈥機onnell-Rodwell. 鈥淭hey shouldn鈥檛 be able to hear
it,鈥� says Arnason. He鈥檚 now investigating whether it would be possible for
elephants to pick up seismic waves caused by thunder so far away.
Arnason even speculates that sensing thunderstorms could have been the
original reason why elephants evolved the ability to detect vibrations. 鈥淚n
Central Africa, what was the ambient seismic noise 10,000 years ago? It was
thunderstorms. And water would have been very important to the elephants. Maybe
first they were able to sense the vibrations, and then found out they could
generate them too,鈥� he says.
But so far the evidence is circumstantial. To prove their theory, Hart and
her colleagues need to show that the elephants can respond to the vibrations set
up by their calls, something that鈥檚 easier said than done. 鈥淭hey are very subtle
animals,鈥� says Hart. 鈥淭hey don鈥檛 necessarily respond in a constant way.鈥� They
might orient towards the source of the sound, shift their weight, freeze, hold
their ears out, or do nothing at all. 鈥淚f they don鈥檛 respond they may still be
hearing it, just not interested,鈥� says Katy Payne, of the Elephant Listening
Project at Cornell University in New York. Payne says she likes the seismic
theory.
Hart and O鈥機onnell-Rodwell have already carried out one study where they
recorded three common acoustic calls鈥攁 warning call, a greeting call, and
the elephant equivalent of 鈥渓et鈥檚 go!鈥� They played the calls back through
seismic transmitters to eight young, domesticated elephants in Zimbabwe. 鈥淲e
made sure the elephants were only feeling the seismic stimulation,鈥� says Hart.
They compared the elephants鈥� responses to the calls with their responses to
vibrations derived from synthesised tones, rock music and to silence.
The seven males were blas茅. 鈥淎t first we thought they weren鈥檛
responding, and we got disheartened, but the trainers would say they are
responding, they鈥檙e just pretending not to!鈥� says Hart. Certainly, one female
became very agitated when the 鈥渨arning鈥� signal was played, even biting the
ground鈥攁 behaviour only seen in elephants under extreme stress. The
researchers videotaped the elephants鈥� responses and are now using computer
analysis to try to pick out more subtle changes in the elephants鈥� behaviour.
In the meantime, they are planning a further round of tests due to start in
October, in which they will train elephants at a nearby zoo to respond to
certain sound signals in a specific way. 鈥淥nce the elephant senses the thing
that it has been trained to respond to, it touches its trunk to a target,鈥� says
O鈥機onnell-Rodwell. The researchers then plan to play seismic versions of the
calls, to see whether the elephants still respond. They hope the project will
provide definitive evidence for their theory.
But if elephants can pick up seismic signals, vibrations caused by human
activity could now be interfering with or preventing this ability. 鈥淎nywhere
near a town has gigantic amounts of background seismic noise,鈥� says Arnason.
Even out in the desert, his equipment still senses trucks passing up to 30
kilometres away.
Remote areas of Africa and India aren鈥檛 immune either鈥攍ow-tech devices
such as water pumps create vibrations too, and aircraft noise permeates
everywhere. 鈥淭he amount of seismic noise that transcontinental jets produce is
enormous,鈥� says Arnason. 鈥淲hen a jet comes towards you it鈥檚 not such a problem,
but once it passes you, there鈥檚 a large Doppler shift.鈥� This moves the high
frequencies of aeroplane noise down into just the range that elephants use.
Hart fears that sound pollution could be very damaging. 鈥淪eismic vibrations
from human activities may limit elephants鈥� access to information. It may be
disruptive to their social activities, and it may be stressful to them,鈥� she
says. 鈥淭he kind of environment in which animals evolved these sensibilities no
longer exists.鈥�