10 June 2010, by Tamera Jones

Large ocean predators like tuna, sharks and ocean sunfish use two different tactics to search for food depending on whether it's abundant or sparse, scientists have shown for the first time.

Mako shark

A mako shark.

When their prey is patchy, open-ocean predators make long swims followed by lots of smaller ones – called Lévy flights – to find their prey. But when fish are abundant, they use a local 'scattergun' approach to find food.

'Lévy flights are specialised random walks made up of long steps followed by lots of short steps occurring at every spatial level. If these fish were tracked from space or from a boat, you'd see the same pattern,' explains Professor David Sims from the Marine Biological Association, who led the research.

It may seem obvious that large predators have to swim for miles and then make shorter swims to optimise their chances of finding prey. Indeed scientists have hypothesised since the 90s that in unproductive waters, predators would use Lévy flights to locate prey, while in rich waters, they should make smaller, more localised movements.

But finding evidence to back these ideas up has, until now, proved challenging.

Large datasets

'The fact is this is an idea that's been extremely difficult to test. You need very large datasets,' says Sims.

Earlier research that claimed to demonstrate Lévy flights in albatrosses, bumblebees and deer has since been discredited, because the data was problematic or the statistical methods the scientists used weren't accurate enough.

'Part of the difficulty has been in dividing up foraging behaviour and other behaviours like resting, travelling or interacting that are unlikely to have patterns best approximated by Lévy flights. We now have better statistical methods to identify the laws that govern these complex patterns making it easier to distil a pure signal,' says Sims.

The research team – made up of 16 scientists from five countries – describe today in Nature how they attached electronic tracking tags to 55 individual predators from 14 different species of shark, tuna, ocean sunfish and swordfish. This produced a large dataset of more than 12 million movements collected over 5700 days.

Using complex statistical methods, they divided up this data so that they could analyse it in detail.

They found ample evidence for Lévy search patterns in nearly all 14 species. And, as predicted by the Levy flight hypothesis, individuals switch between Lévy and a localised approach to foraging depending on what habitats they're in.

The researchers call these localised movements Brownian searches after the term scientists use to describe the random movement pollen grains make when they're suspended in water and are being bombarded by water molecules.

One blue shark swapped Brownian behaviour in the rich waters of the continental shelf edge off northern Spain for Lévy searching in the comparatively empty waters of the Bay of Biscay.

Another fish, a bigeye tuna in the central eastern Pacific near the Galapagos Islands made big steps followed by small steps to find food, but when it moved to cooler waters full of fish, it switched to Brownian-type movements.

The researchers are keen to test their statistical methods on animals like octopus, cuttlefish and snails next.

'If this is a universal law of searching it should occur everywhere. The next question is, did animals evolve Lévy flights when faced with challenging environmental conditions at some time in prehistory?' says Sims.