91av

JEFF LEIS’s job requires total concentration. He swims through the crystal
waters of the Great Barrier Reef, his eyes glued to a single, tiny fish larva
swimming in front of him. Leis, a fish expert from the Australian Museum in
Sydney, has spent hundreds of hours in the waters around Lizard Island following
his tiny prey. It is not an easy job. Each attempt requires two divers—one
to keep the scarcely visible animal in view, the other to operate a stop watch,
compass and depth gauge. A third person sits above in a boat following their
bubbles and watching for sharks. Sometimes there is another problem: the fish
takes off with such speed that the divers can’t keep up.

A little farther up the coast, Eric Wolanski is also watching fish larvae.
The difference is that his tiny specks of life are artificial, the creation of a
computer program—and there are many thousands of them. Wolanski is an
oceanographer who models currents and fish movements around the Great Barrier
Reef. Crunching data collected over almost two decades, Wolanski’s program has
also revealed some astonishing things about the skills of fry that live around
coral reefs.

Both low-tech and high-tech approaches have led to the same conclusion. Coral
fish fry are not the helpless creatures everyone thought. Although they are
often only millimetres long, these larvae have the speed and endurance to free
themselves from the grip of an ocean current and the ability to detect a reef
from at least a kilometre away. This new discovery will change the way
biologists think about the colonisation of reefs, and the best way to protect
them from overexploitation.

Like most marine animals, reef fish have two distinct stages in their life.
The adults tend to stick close to home, often remaining on the same part of a
reef all their lives. For these fish, the deep water between coral shoals and
islands is an uncrossable void. By comparison, their offspring lead adventurous
lives. Almost all reef fish shed their eggs and sperm directly into the water,
and the resulting embryos are quickly swept away by tides and currents.
Developing larvae spend the next few weeks in open water until they are ready to
settle down and adopt the sedentary habits of their parents.

According to conventional wisdom, these offspring of reef fish are passive
specks in the plankton, hapless scraps at the mercy of currents, tides and
wind-driven movements of the ocean’s surface waters. As a result, only a
tiny proportion of the young produced by reef fish will ever find a home,
fetching up on a distant reef more by chance than design. Many more must be
“lost at sea”, having failed to find a place to settle.

“For most world fisheries, we are told there is a tremendous amount of wasted
fish larvae,” says Wolanski, who works at the Australian Institute of Marine
Science near Townsville in northern Queensland. Even so, there were thought to
be far more larvae around than reefs could support—plenty of spare fish to
replace stocks reduced by overfishing.

But the story that is now unfolding is almost exactly the opposite. Larval
fish are not passive, nor are there enough waiting in the wings to fill the gaps
left by fishermen with hooks, crowbars, dynamite and cyanide.

Most studies of fish larvae have been carried out on species from temperate
waters, such as cod and herring. These are much less well developed than coral
reef species at a given size and swim very badly—in the lab at least.
Unlike herring and cod, the young of reef fish have good reason to travel under
their own steam. When the time comes to settle, the larvae have to find a small
reef in a big ocean, otherwise they won’t be able to complete their life
cycle.

Leis and his fellow fish-followers have timed more than 50 species belonging
to 15 families in the waters of the Great Barrier Reef and off Rangiroa Atoll in
French Polynesia. Their latest findings leave no doubt that these small fry are
champion swimmers.

On average, the young fish could do a pace-setting 20.6 centimetres a second,
equivalent to almost 14 times their own body length each second. An Olympic
swimmer who could match this would finish the 100 metres in 3 seconds. Not
surprisingly, the bigger larvae, such as those of surgeonfish and soldierfish,
swim faster, but taking size into account, some of the smaller ones give more
impressive performances. A small damselfish, for instance, achieved an
astonishing 50 body lengths a second. The all-time champion, a soldierfish
followed off Rangiroa, averaged 56 centimetres a second over 10 minutes, but put
on such a spurt in the final two minutes that it left the divers behind.

At James Cook University in Townsville, PhD student Ilona Stobutzki has
tested both the speed and the staying power of larval reef fish. She released
them into specially designed race tracks, where she could control the speed at
which the water flowed through the tracks. The fish had to swim against the
current to avoid being swept into a mesh fence across the track.

Stobutzki put the larvae of dozens of species through their paces, increasing
the speed of the “current” every five minutes until the fish could no longer
keep up. Even the small-bodied species reached 40 centimetres a second. To test
their endurance, Stobutzki made the fish swim against a current of 13.5
centimetres a second—the average around Lizard Island—and kept them
swimming until they tired. This time, the champion was a surgeonfish, which kept
going for 94 kilometres without a rest.

Taking control

“We were phenomenally surprised by how good they were,” says Stobutzki. “We
had no idea they could do this much. This means that in the final stages before
they settle, they have a huge degree of control over where they go.”

Both Stobutzki and Leis have also found clear indications that before they
settle down, the young fish have a sense of direction and can detect reefs from
at least a kilometre away. Leis captured larvae close to Lizard Island just as
they were about to settle at night—luring them into traps “baited” with a
bright fluorescent light. The next day, he took them back out to sea, a
kilometre from the reef, and freed them one at a time. “The fish chooses where
it’s going and the divers follow it,” says Leis. What he found surprised him.
About 80 per cent of the larvae swam in a very definite direction—not
towards the reef but away from it, regardless of whether the reef was to the
east or the west of them.

This shows that larvae can detect a reef from this distance, says Leis. He
believes the larvae chose to swim away because the experiment was done during
daylight. Most studies suggest that larvae arrive on a reef after sunset, when
the big schools of plankton-eating fish that form the “wall of mouths” around a
reef have retired to their night-time lairs. Once on the reef, the young fish
must find a safe place very quickly—in the deadly tentacles of an anemone
in the case of the clownfish, and among the branching corals for the
black-and-white humbugs. By morning, any small fish that has not found suitable
shelter is likely to be eaten by the hordes of emerging predators.

Stobutzki tried another approach—and her results suggested that Leis
might be right. She collected fish in light traps and then placed them in an
elaborate apparatus close to Lizard Island. When the fish were released from a
central chamber, they could choose to swim either towards the reef or away, only
to be captured in another chamber of the trap. The experiments were done at
night and every time the fish headed reefwards.

At the Australian Institute of Marine Science, Wolanski and fish biologist
Peter Doherty have just reached the same conclusions. But instead of looking at
individual fish in a small patch of water, they analysed the movements of
hundreds of thousands of larval fish in a huge section of the Great Barrier
Reef, trying to reconcile the distribution of fish with the physical movement of
the water.

Currents and tides

Until now, models of the movements of fish have been based on patterns of
movements of the water. “All you needed were modellers to tell you where the
water was going and that would tell you about the fish,” says Leis. “That’s no
longer enough.”

Wolanski has spent the past 18 years collecting data on the patterns of water
movement around some of the three thousand islands and shoals in the Great
Barrier Reef. The central part is “chock-a-block” with reefs, says Wolanski:
“Water must find its way through the reefs like wind blowing through the trees
in a forest.”

With reefs of all sizes and shapes, some separated by shallow water, some by
deep channels, the circulation of water is extremely complex and modelling it is
a Herculean task even for a computer. But with supercomputer and special
software donated by IBM, Wolanski has plotted the way water circulates through
this obstacle course, turning a confusing mass of data into a 3D computer
animation that shows how the water moves around each reef, sometimes swirling
and creating eddies behind the obstruction before breaking free and travelling
on down the chain of reefs.

While Wolanski has been collecting data from current meters and remote
sensing satellites, Doherty has spent the past seven summers trapping larval
fish to build up an equally detailed picture of what they are doing in those
waters. He has discovered that in this part of the Great Barrier Reef the young
of all sorts of reef-dwelling fish—from damselfish and angelfish to
commercial species such as groupers and sweetlips—approached a reef from
upstream. They came in pulses, arriving on the reefs at the time of the new
Moon—the darkest nights of the month. Soon the larvae began to concentrate
in hot spots around the edge of the reefs, especially off the downstream
end.

On the face of it, this might be explained by hydrodynamics. Eddies that form
in the lee of a reef might trap larvae, holding them close to the reef and
providing them with an opportunity to settle. But in the Great Barrier Reef,
these eddies are unstable and don’t last long enough to retain the fish. Other
findings also made Doherty and Wolanski suspect that physics was not the only
force in operation. For instance, if all larvae were passive particles, why were
there no young of open water fish in the traps set near the reefs? And why were
there such large concentrations of reef larvae? It looked as if the bluewater
fish were deliberately keeping away, while the larval reef fish were
deliberately making for the reefs.

With so much data, the only way to find out what was really going on was to
turn to the computer model and simulate the behaviour of fry by creating
artificial fish. After feeding all his oceanographic data into the model,
Wolanski introduced a cloud of larvae moving towards the reefs. In the first
computer run, the larvae were treated as passive particles. They did turn up in
the lee of the reefs, as Doherty’s traps had shown. “The model with passive
larvae did predict that they would be there,” says Doherty. “But only at the
same densities as elsewhere around the reef.”

At the next attempt, Wolanski gave the fish the ability to swim and a sense
of direction. This time, the model produced the same sort of results as the
biologists. “The concentrations are a hundred times those in the original patch
and that’s what is observed from trapping,” says Wolanski.

Making some fish swim faster than others—mimicking the differing
abilities of large and small larvae—produced an even closer match to
Doherty’s observations: larger larvae were spread more evenly whereas smaller
ones invariably collected in the lee of a reef, where currents are weaker.

Clearly, larval fish are strong swimmers, and there is good evidence that
they can sense where a reef is from some way off. The question is how do they
know? What can they sense and from how far away?

“We can eliminate some possible cues,” says Leis. Observations suggest that
they don’t rely on currents or wind-driven waves. The fact that Leis’s
larvae swam away from the reef regardless of their position relative to it rules
out a sense of direction based on the Sun or the Earth’s magnetic field. Vision
can also be ruled out. “Initial orientation starts at distances well beyond the
extent of vision,” says Doherty.

Noisy and smelly

That leaves two possibilities: the fish could be homing in on sound or some
sort of smell. Fish can detect chemicals in the water. Clownfish, for example,
rely on scent to find their protective hosts in the dark. Some species, such as
the humbug damselfish, detect others of their kind by their smell. “But we’re
talking about a scale of a few metres, not the sorts of distances you need to
find a reef,” says Leis. Besides, in the mixed-up waters of the Great Barrier
Reef, any reef smells will be so widespread that they will be of no help in
pinpointing a particular reef. So that leaves sound.

Almost nothing is known about the auditory powers of larval fish, but reefs
are very noisy places. The crashing of surf on coral can be phenomenal. Inside
the outer barrier reef, however, the wind whips up surface waves that generate
enough noise to drown out the coral breakers. “But lots of reef animals make
noises,” says Leis. “Sea urchins make scraping noises as they graze over the
coral and so do parrot fish.” Fish can also generate their own music by grinding
their jaws or “drumming” on their swimbladders with certain muscles.

The best bet is probably the “nocturnal chorus”—a piscine equivalent of
the dawn chorus. Navy scientists have detected it in waters throughout the
tropics. “They think it originates from fish, although they’ve tried to catch
them and failed,” says Leis. “It has peaks of intensity in summer, at night and
at the time of the new Moon. All three of these are peak times for recruitment
of larvae to reefs.”

The crucial question now is at what stage in their development do larval fish
begin to take control of their fate? This depends largely on how quickly their
swimming abilities develop. Leis has found that small damselfish can swim at up
to half their eventual top speed by the midpoint in their larval
development.

“If we can look at the earliest stages, we might find that even they are not
entirely passive,” says Stobutzki. “They might be able to move vertically in the
water—perhaps into water that is not moving so fast.”

The astonishing skills of young fish have far-reaching implications, both for
biologists and for those responsible for protecting reefs. The findings destroy
the notion that larvae are inevitably swept away from the place they are born to
settle somewhere far downstream. Some young fish may resist the movement of
water and remain close enough to their home reef to make their way back when the
time comes to settle. In the case of isolated oceanic reefs, this may be crucial
because the alternative is the loss of most of each year’s young. It may not be
so important for fish on the Great Barrier Reef, where there are plenty more
reefs nearby. There, the larvae may use their skills to ensure that they don’t
travel too far, or end up outside the barrier altogether. They may travel only a
few tens of kilometres rather than the hundreds of kilometres people had
assumed—and there may be definite “highways” linking particular reefs.
“What influence larvae have on where they end up will be important for those who
manage reefs. They will have to think about things very differently,” says
Leis.

With most of the world’s coral reefs seriously threatened by development and
overexploitation, many reef ecologists believe that the only way to preserve
their fish populations is to create a network of refuges, reefs or parts of
reefs that are out of bounds to fishermen.

The concept is simple. If you allow some fish to reach adulthood, they will
produce huge numbers of offspring which will be carried on the currents to
restock any overexploited reef—perhaps hundreds of kilometres away. But
the discoveries in Australia suggest that the strategy is badly flawed. First,
it looks as if a much larger proportion of each year’s young make it to a reef
and so have a good chance of settling. That destroys the myth that there is a
huge surplus of spawn to fill in the gaps on overfished reefs.

Added to this, if refuges are to work, biologists need to know if a reef
replenishes itself, or where there is traffic between reefs, they need to know
how particular reefs are linked. “At the one end of the spectrum, you would have
to manage each reef independently. At the other you would have to protect great
chunks of reef systems,” says Doherty.

Leis has another suggestion. “If we can find out exactly what it is the fish
homes in on, we might be able to use that to guide larvae towards overfished or
damaged reefs.” But discovering the musical preferences of tiny fishes is likely
to be as difficult as chasing them in the dark.