Tool use?
« previous post | next post »
From the holiday special issue of New Scientist, a short item "Coconuts make the best hideouts" (in the print version, the more sober "Octopuses use coconut shells as portable shelters" in the on-line version). Andy Coghlan writes (in the on-line version):
Octopuses have been observed carrying coconut shells in what researchers [in Current Biology] claim is the first recorded example of tool use in invertebrates.
… "This octopus behaviour was totally unexpected," says Julian Finn, a marine biologist at Museum Victoria in Melbourne, Australia, who has filmed at least four individual veined octopuses performing the trick off the coast of Indonesia.
People living in Indonesian coastal villages discard coconut shells into the sea after use. When the octopuses come across these on the seabed, they drape their bodies over and around the shells, hollow-side up, leaving their eight arms dangling over the edges.
The octopuses then lift the shells by making their arms rigid, before tiptoeing away in a manoeuvre Finn calls stilt-walking.
When the octopuses feel threatened, they flip the half shells over themselves and hide. Some even use two shells to create a more spacious shelter with an opening through which they can keep a lookout.
… Finn argues that the behaviour qualifies as tool use for a number of reasons. First, the shells are not permanent homes like those occupied by hermit crabs, but are carried around for future use.
It is also a costly behaviour, both in terms of energy use and in potentially making the octopuses more vulnerable to attack.
Others are not so sure:
"The conventional definitions of tool use include the use of non-attached objects to act on other object(s), which may be food," says Alex Kacelnik of the University of Oxford, whose team recently studied whether crows plan their tool use in advance.
"The use of a coconut shell for protection does not fit this definition, but I sympathise with the authors in that this does not imply that the cognitive demands for such an action are lower," Kacelnik says. "The interesting issues are not whether this observation fits a pre-established definition, but what cognitive operations make the behaviour possible."