44 | New Scientist | 2 April 2022
Earth-based receivers. The information
gathered is freely available to all. And
increasing numbers of grassroots animal
taggers are getting involved in ICARUS.
As the project grows, so do Wikelski’s
ambitions. He believes that, as well as
shedding new light on animal behaviour, the
internet of animals can also help us forecast
environmental change, track emerging
diseases, conserve endangered species
and make humans more responsive to the
needs of other creatures.Matthew Ponsford: To drum up interest in
ICARUS, you published a list of mysteries
that it could shed new light on. Which in
particular would you like to solve?
Martin Wikelski: There are hundreds, but one
we want to start with is where do European eels
go to reproduce? Are they really going to the
Sargasso Sea and, if so, where exactly? This is
all a big mystery. Another is what are the causes
of death in European storks? We’re planning
to tag 15,000 storks a year to understand why
about 70 per cent die in their first year. Where
are they? I mean, if you imagine the pile of
10,000 dead storks, that’s a hell of a lot. But
nobody sees them die out in the wild. Also, we
have massive declines in songbirds: in Europe,
their numbers are down by about 600 million,we can tap into. For example, we have some
early indications that animals can sense
upcoming disasters like earthquakes or
volcanic eruptions.How can you put such long-held folk
beliefs to the test?
The key is to work with local people who
understand the animals. It’s almost like a good
dog handler – if you have a good person, they
can search for drugs at customs. Not everybody
can do that. But it works if you ask those
people: ‘What kinds of signs are you looking
for? How can the dog tell you?’ That’s exactly
how we start with disasters. We try to get a
read-out of this black box, this sixth sense of
animals, this interaction of intelligent sensors.
We’re doing this now at Mount Etna in
Sicily, Italy. We have goats running around,
and whenever we see the goats do something
crazy – it happens every few weeks or so – we
predict that something is happening with
the volcano. And in most cases, we are right.
We did an internal evaluation of our
predictions and it was pretty powerful.
We aren’t yet at an early warning system,
but we are getting towards these systems.
And then we can go further and ask what kind
of environmental features the animals could
tune into. That’s so far pretty much unknown.
It isn’t seismic information because that’s easy
to measure, so it must be something else.Could animal networks also help with more
mainstream predictions, such as weather
forecasting?
Yes, many systems are really ripe for animal
predictions. For example, the gannets and the
shorebirds in the western part of Mexico – they
tell you in spring how the harvest of anchovies
and other fish will be in the fall, because they
are already tuning in to the fry production
early in the season. Or boobies in the Indo
Pacific will tell you how strong the next El Niño
will be because – months ahead of time – they
all give up their breeding schedules. They
either abandon the eggs or they don’t even
lay eggs. And then you know it will be a
strong El Niño. We already have those kinds
of long-term predictions, but we have not
brought them together yet on a global
scale. And that goes back to the internet
of animals. That is what’s coming.“ Boobies in the
Indo-Pacific will
tell you how
strong the next
El Niño will be”
in the US, it’s 3 billion birds missing. Where did
they get lost? Nobody knows.What kinds of things have you observed
with ICARUS so far?
Every time we look, we see something
completely unexpected and new. People
only know on average what animals are doing,
but no animal is average. They all do really
interesting and crazy things. So we have
white storks flying over the Sahara five times
a season. We have cuckoos going from
Kamchatka [in Russia’s far east] through
India to Angola and back. We have sooty
terns from the Seychelles going all the way
to eastern India and Sri Lanka. For us, every
day is super exciting.Other groups are watching animals from space.
What does ICARUS do that’s new?
It can track smaller animals, it can track more
animals, and it can track their behaviour
through sensors. We record temperature,
acceleration, humidity, magnetometry,
pressure – the kind of environmental
sensing that the other satellite trackers
don’t have. We can track wing beats in birds,
energy expenditure, foraging behaviour
and see when they are excited or stressed.Why do you call this the “internet of animals”?
This internet is, in principle, the collective of
the animals, because they constantly interact
with each other. It’s really what you could
call swarm intelligence. We are combining
the terrestrial Internet of Things (IoT) with a
space-based IoT. So, if an animal is close enough
to connect to an Earth-based data station,
then it communicates terrestrially. If it’s
away from these areas – if it is killed in a valley
in the rainforest, if it dies over the Sahara,
or whatever – then you need the satellite IoT
to send and receive that information.
The interesting part is that we now know
that collective behaviour is basically responsible
for the “sixth sense” of animals. In the old days,
people saw this as something metaphysical or
crazy. But we know from physics and chemistry
that, where there are interacting parts, new
properties emerge. This is what you have in
animals as well. With intelligent sensors that
interact, you get emergent new properties that
aren’t visible on the lower level. That’s what