Sea bream have a fine ''ear'' for music, being particularly fond of softly played piano melodies.
But unhappily, this refined musical taste is going to lead a lot of fish straight to the dinner tables of Japan.
Experiments are under way in southern Japan using music for ''systemized fish management.'' It will be something like a cattle ranch. Artificially bred fish will be raised for the first few months until it becomes a reflex action for them to associate certain sounds with food.
When they are full grown, the lilting melody will sound one last time. The fish will congregate - only to find they are the food.
It may not sound very sporting, but it is considered essential by both government and private industry if the fish-loving Japanese are to continue to enjoy this dietary staple without problems.
The Japanese are concerned about the many coastal nations imposing 200-mile exclusive offshore economic zones and about last week's global ban on commerical whaling. These concerns provide an impetus for developing a more rational system of fishing close to home.
The Tokyo government has designated 17 areas throughout the country where fishing by music could eventually be developed. Current experiments are concentrated on Saki Bay, on the island of Kyushu, already a major fish-breeding area.
Engineering consultant Tomoo Nishino explains that most breeding projects today rely on nets to keep the fish in one place. But some types of fish don't respond well to the enclosed environment. They don't grow as big as they should and their flesh isn't so tasty.
Yet, left to their own devices, most fish would disappear into the ocean depths. Nishino is convinced that ''herding'' them to music is the answer.
The Saki Bay project calls for a network of underwater structures. A central tower will be manned by experts using their eyes and ultrasonic acoustic waves to measure the size and numbers of fish in the farm area.
Three radio towers will emit the chosen sound for 30 minutes twice a day. Fifteen minutes after the performance ends, food will be released automatically to what tests have shown should be a large number of fish docilely queueing up for dinner.
The experiments so far have concentrated on sea bream. For the first year of their life they tend to congregate within half a mile of the shore. For the following two years they move out to a range of 1 1/2 miles, before suddenly heading out to sea.
Through constant monitoring by the central control tower, fishing boats would be directed to the spot to catch a good proportion of the fully grown fish shortly before they perform this disappearing act.
In the first tests, divers went out with receivers to check on the efficacy of the sound emissions, which were found to travel over half a mile from the towers.
The musically trained fish had been marked for identification, and it was found they tended to bring along numerous ''strangers'' when they headed for the feeding towers.
Various tests established the sea bream's predilection for piano music - a rather low sound in the 200 to 300 hertz range, played at between 50 and 60 decibels.
By comparison, squid, another Japanese favorite, respond to sounds of 600 to 700 hertz at 15 to 30 decibels.
Do the sea bream have any favorite piece of music?
An amused Mitsuaki Kobiki, another member of the design team, regretfully has to destroy the reporter's hopes for a good headline. ''It doesn't have to be a melody actually. That's just our romanticism at work. A single tone or sound could do equally as well. So we haven't yet established whether the fish like Chopin or Mozart, for example.''
Tomoo Nishino foresees the full-scale system in use within a few years - as well as being exported. He explains the advantages it offers: ''Japanese tastes are changing. People today want to eat better quality fish, rather than, say, sardine and tuna, which are still abundant. So the idea is to systematically raise the type of fish people want to eat. This will also benefit the local fishing industry, which can catch all the fish we need right close to our own shores.''
He says experiments so far have shown that about 30 percent of all the fish hatched end up being caught.