Biology Recent studies have shown that fish are adaptable. They learn new tasks quickly, and have the ability to discriminate among colours, shapes and sizes, writes Sean B Carroll
I have long suspected that fish are smarter than we give them credit for. As a child, I had an aquarium with several pet goldfish. They certainly knew it was feeding time when my hand appeared over their tank, and they excitedly awaited their delicious fish flakes.
They also exhibited a darker, disturbing behaviour. A safe life with abundant food was not fulfilling. From time to time, either sheer ennui or the long gray Toledo winter got to one of the fish and it ended its torment with a leap to my bedroom floor. Maybe my anthropomorphizing is a bit over the top. But, really, just how smart are fish? Can they learn?
A 10-gallon tank with a plastic sunken pirate ship is certainly not the most stimulating habitat. But in the colourful and diverse world of coral reefs, fish must be able to recognise not only food, but also to discriminate friends from foes, and mates from rivals, and to take the best action. Studies have revealed that reef fish are surprisingly adaptable. Freshly caught wild fish learn new tasks and discriminate among colours, patterns and shapes, including those they have never encountered. These studies suggest that learning and interpreting new stimuli play important roles in the lives of reef fish.
Responding to stimuli
To test the ability of fish to learn to discriminate shapes, a research team led by Ulrike E Siebeck at the University of Queensland in Brisbane, Australia, trained damselfish to feed from a tube to which they attached a variety of visual stimuli. The latter included a three-dimensional latex disc, a two-dimensional blue disc painted on a plastic board, or black circles or propeller patterns on white boards. The fish were rewarded with food when they repeatedly tapped the stimulus, not the tube, with their snout or mouth.
The fish rapidly learned this task. The researchers then presented the fish with the original stimulus and one alternative distracting shape, bars versus discs, squares versus discs, or circles versus propellers, and the fish had to nose the shape they had been trained to tap in order to receive a reward. The fish tapped the correct shape about 70 per cent of the time in the first trial; this improved to 80 per cent to 90 per cent in subsequent trials.
The fish also learned when the food reward was delayed and delivered far from the stimulus. They exhibited anticipatory behaviour, in that they would tap the image and then swim quickly to the other end of their tank in anticipation of their food reward. This is like Pavlov’s dogs who learned to anticipate food at the sound of a bell.
In another set of experiments, Siebeck trained damselfish on different colour stimuli. She selected blue and yellow because they are highly contrasting colours that are found on many reef fish. After the fish learned to repeatedly tap coloured latex targets to gain a food reward, they were presented with a choice between the training target and the alternative colour target. The fish were even better at colour discrimination, tapping the correct target more than 90 per cent of the time.
Perhaps it is less surprising that the fish learned to discriminate colours. After all, they live in a colourful environment. But the question of why reef fish are so colourful has challenged biologists for a very long time. It seems obvious that bright colour patterns would be effective communication signals in the shallow, well-lit water around coral reefs. But in that fish-eat-fish world, bright colours would also make fish conspicuous to predators. So how are these advantages and disadvantages balanced?
It turns out that some brightly coloured fish make a living by providing a valuable service to what may otherwise be their predators. They clean them. In fact, cleaner fish like the cleaner wrasse form an important part of coral reef communities. They establish small territories as “cleaning stations,” which are visited by all sorts of “client” fish that have their parasites removed. The cleaners’ work ethic is astounding. Alexandra Grutter of the University of Queensland found that individual cleaner wrasse inspected 2,300 fish and consumed up to 1,200 parasites a day, which amounted to about seven per cent of their body weight. Furthermore, Grutter found that fish on reefs without cleaner fish had about five times the number of parasites compared with fish on reefs with cleaners.
It would seem, then, that it would benefit potential clients to visit cleaning stations, and for carnivorous clients not to eat their cleaners. How do clients recognise cleaners? Certain body colours, particularly blue and yellow, signal cleaning behaviour to potential clients.
To study the role of colour in the cleaner-client relationship, another research team from the University of Queensland, including Grutter, Karen Cheney, Simon Blomberg and N Justin Marshall, looked at the distribution of body colours among cleaner and noncleaner fish from the same families. They found cleaner fish were more likely to have a blue or yellow coloration.
They also showed that these colours were the most contrasting ones on coral backgrounds to clients like barracuda or surgeonfish and that the contrast was enhanced against black backgrounds. In fact, all species they examined that make their living from cleaning also had a contrast-enhancing black lateral body stripe adjacent to these colours, whereas none of the 31 noncleaner species were so marked.
To test whether potential clients paid attention to these colours, researchers painted models with various colours in which they omitted the blue pattern or replaced it with red, or altered the pattern, orientation and width of body stripes. They then placed these models around reefs fringing Lizard Island, at the northern end of the Great Barrier Reef, and observed the frequency with which client fish visited the models. They found that the model that represented the natural blue-streak cleaner wrasse pattern was visited more often than any other model colour scheme.
In a similar study performed off Sulawesi, Indonesia, the length of the model’s black body stripe also affected the frequency of client visits. On coral reefs, it pays to advertise, even when potential enemies abound.