Last week, scientists issued their latest, grim assessment of the world's coral reefs. But as Steve Connor reports from Florida, extraordinary new ocean 'reseeding' techniques mean there may still be time to halt - or even reverse - the destruction of mother nature's marine marvels.

    Coral reefs are often described as the tropical rainforests of the oceans. But marine biologists sometimes use another analogy: that of the canary in the coalmine. These birds were used by miners as an early warning for lethal gas; corals, too, are extraordinarily sensitive to environmental change. For Nancy Knowlton, a scientist at the Smithsonian Natural History Museum, it's an apt description: "If that's the analogy, then the canary has passed out on the floor of the cage. Coral reefs are potentially immortal. They only have to die if we make them."

    And that's just what we seem to be doing. In the 25 years that Knowlton has been studying the reefs, she has witnessed all the signs of their terminal decline. They are being degraded at a rate of 2 per cent a year. About a fifth of the world's stock has already gone, and nearly half of the remainder is in danger of disappearing within the next 20 years. And like so many other experts in her field, Knowlton is worried: a lethal combination of pollution, predators, disease, rising sea temperatures, over-fishing and the acidification of the sea have put our coral reefs on the critical list.

    Its plight is bad news for all of us, but will horrify anyone who has put on flippers, mask and snorkel to experience its magnificence first-hand. Snorkelling over a reef for the first time, as I did last week off the coast of Florida, is like floating over a brilliantly coloured Garden of Eden landscaped by some maritime Capability Brown. Corals of all shapes and sizes grow in the dappled sunlight. Vast, bulbous species covered with beautifully etched crenulations look like the intricate folds on the surface of a human brain. Others resemble petrified trees, their branches sticking up like fingers, or flat pancakes woven with intricate lacework. Waving sea fans drift back and forth with the gentle pulse of the waves, a hypnotic motion that sends you into a trance-like state of awe.

    And then there are the fish - lots of fish. Nothing quite prepares you for the variety of sizes, colours and shapes swimming in and out of the coral latticework. There are iridescent blue ones with fins like a teddy boy's quiff. There are green ones with metallic scales, each a slightly different hue from the next, like the scaly armour of a Scythian warrior. A much larger fish with camouflaged skin and a big, ugly head spies me with his swivelling eyes and tries to hide, ostrich-like, behind a skinny staghorn coral; a huge ray, five or six feet long, glides effortlessly past, trailing a menacingly spiked tail in its wake.

    This scene is repeated everywhere on earth where tropical reefs form - which is just about anywhere on the vast, watery belt around the equator. The biggest of them all is the Great Barrier Reef of Australia, which is some 2,300km long. It took more than 10,000 years to get to where it is today, growing at a rate of a centimetre or two each year.

    Coral reefs are the product of tiny animals called polyps, which secrete an exterior skeleton of aragonite, a mineral made of calcium carbonate. Each generation of polyp grows on the dead skeletons of its ancestors, but because they clone themselves they have achieved a kind of immortality. Some grow slower than the rate at which the continents move - fingernails grow faster. But eventually they form massive structures such as the Great Barrier Reef, which is more properly a collection of 3,000 separate reefs and 900 coral islands, divided by narrow channels. They are the only biological structures that can be seen from space.

    With so much coral in the world, it is hard to see how they could be at risk of destruction. But they are. Last week, the latest warning came from scientists who warned that an estimated third of all reef-building coral species were in imminent danger of extinction.

    The threat had been steadily growing, and has now become a full-blown crisis. The scientists, working on behalf of the International Union for the Conservation of Nature, found that 231 reef-building coral species, out of a total of 704 that were capable of being studied, can be classified as either vulnerable, endangered or critically endangered, as defined by the IUCN's Red List of threatened species.

    Equally alarming, however, was what happened when they extrapolated back to the situation before 1998, the year when a large-scale "bleaching" episode occurred worldwide. When they back-calculated using the same methodology, they could only find 13 species before 1998 that would have fallen into the same three categories. In other words, the situation now is something like 18 times worse than it was just 10 years ago.

    The bleaching episode of 1998 is perhaps the single biggest milestone in the timeline of coral decline. As the name implies, the vividly coloured organisms turned a ghostly white, as a result of a king of marital row between the tiny polyp - the fleshy part of the stony coral - and its brightly decorated partner, a microscopic algae that lives inside the polyp.

    Corals may look superficially like plants, but in fact each polyp feeds on passing food parcels floating in the marine plankton. Although technically an animal, the polyp forms a mutually beneficial relationship with photosynthetic algae called zooanthellae. It is a give-and-take marriage of convenience. The polyp offers shelter and possibly other nutritional benefits to the algae, while the zooanthellae provides the polyp with energy-packed carbohydrates left over from its daytime job of converting sunlight into sugar.

    But, like many marriages, the relationship is fragile. When the corals are stressed out by something in the environment - in 1998, the trigger was excessively high sea temperatures exacerbated by an El Niño event and global warming - the polyp evicts the algae, and turns white in the process. Often the polyp survives, but it is almost always seriously weakened by the rift. As a result, the coral is prone to disease when some other stress comes along - a one-two punch that so often proves fatal for the reef.

    Scientists predict that, as sea temperatures rise, mass bleaching will begin to occur on an annual basis within the next 30 or 50 years. But already there are signs that it is increasing in both frequency and severity on a more localised scale. Douglas Fenner, of the Marine and Wildlife Resources in American Samoa, says that bleaching is now a regular event on his small patch of the Pacific Ocean. In some of the warmer-water coral pools he studies, bleaching has occurred every month or so for the past five or six years. "It's the thin end of the wedge, and it's not good," he says.

    But bleaching is just a symptom of the many deep-seated problems. Rising sea temperatures and pollution caused by run-off from the land - along with sewage and sedimentation - are just some of the factors that can either trigger a bleaching episode or kill the coral outright. Rising sea levels are another problem, because some corals are just not able to grow as fast as the sea level is rising.

    "Rising temperatures could trigger mass coral bleaching once every couple of years," says Rich Aronson, of the University of South Alabama. "And corals are going to have a pretty hard time coping with rising sea levels if they are also suffering from chronic bleaching."

    A still more sinister threat, however, is increasing ocean acidity, which is also caused by rising levels of carbon dioxide. Normally, the sea is slightly alkaline, but when carbon dioxide dissolves in it, the water becomes slightly more acidic. This makes it more difficult for the polyp to build its aragonite skeleton. At its worse, an acidic ocean will actually dissolve the coral reef, in the same way that tooth enamel dissolves in a glass of Coca-Cola.

    When coral scientists from around the world met last week at the International Coral Reef Symposium in Florida, ocean acidification was high on the agenda as one of the most worrying unknowns. Scientist after scientist warned that acidification was one environmental insult that really could wipe out all reef-building corals. If the oceans do become as acidic as is predicted later this century, coral - and many of the species that rely on it - could go the way of the dinosaurs. As Joan Kleypas, of the US National Centre for Atmospheric Research, says: "Ocean acidification is the evil twin of climate change. It is the osteoporosis of the oceans."

    With so much doom and gloom, scientists are left wondering what else they can do to alert the wider world to the plight of the corals - some even talk about "compassion fatigue" and "paralytic nihilism". Is there any hope left within the community of coral researchers?

    The answer, surprisingly, is yes. Corals may be on the verge of extinction, but scientists believe there is still a window of opportunity left open and they point to the other common analogy used to describe coral reefs - tropical rainforests.

    Coral reefs are often described as the rainforests of the ocean because of the diversity of life that both support. And, like the rainforests, coral reefs are under threat. But until recently, few have thought it possible to regenerate a coral reef in the same way it is possible to regenerate a tropical rainforest. Now, however, many scientists believe it is feasible to talk about a "reforestation" programme for reefs to prevent, or at least slow down, the damage.

    These researchers believe this is more than a forlorn hope, despite the widespread pessimism about the future for coral reefs in a warmer world. Indeed, if one message emerged from the International Coral Reef Symposium, it was that no matter how dire the threat to corals has become, there is still time to save them - and coral regeneration could provide a critical stop-gap that could allow at least some corals to live through the climatic rigours of the 21st century.

    Rebuilding or regenerating a coral reef, even if the environment is right for it, is no simple matter, however. The first problem has to do with coral sex, a mysterious business that occurs just once a year within an hour or so of sunset, and (usually) precisely five days after a full moon. For some unknown reason, many coral species release all their sperm and eggs in one huge ejaculate, communicating their mass excitement to their partners in the coral equivalent of pillow talk.

    The resulting microscopic larvae then drift away, with only some of them returning to the reef where they settle down - literally - on the rock to form baby corals. One scientist said that trying to think of a way of managing this process to help with reef regeneration is like trying to manage a herd of deer that produce millions of fawns, which then drift off into the wind like dandelion seeds.

    Fortunately, there is another way to regenerate reefs, which is based on vegetative reproduction - cloning. Scientists can take fragments of coral and grow them on a movable substrate in an underwater nursery. Sometimes the fragments are taken from "corals of opportunity" formed, for instance, from the rubble left behind when a boat smashes into a reef.

    On other occasions, scientists can take tiny samples of corals from existing reefs, grow them in a nursery and replant them back in the wild on a damaged or degraded reef.

    "We can remove fragments as small as between one and 10 polyps, which is equivalent to the bite of a parrot fish. It does no harm," says Yael Horoszowski, of the National Institute of Oceanography in Israel, who has successfully regenerated reefs in the Red Sea using the "coral gardening" concept.

    "The method involves generating and farming large stocks of new coral colonies in an in situ floating nursery prior to their installation into degraded reefs," Dr Horoszowski explains. "These nurseries liberated coral babies at the site. We intended to restore the coral reef but we ended up restoring all the creatures at the coral site."

    Scientists working in such far-flung sites as Biscayne Bay off the south-east coast of Florida - which is in sight of a nuclear power plant and a landfill site known as "Mount Trashmore" - and the Komodo National Park in Indonesia, where fishermen have taken to using home-made bombs to increase their catch, are now actively engaged with different coral gardening techniques in the hope of regenerating their reefs.

    But no one is under any illusion that this is the answer to the coral crisis. Any attempt at regenerating reefs with hi-tech methods of cloning corals, or low-tech methods of re-arranging the local rocks, are doomed to fail if carbon dioxide levels continue to rise.

    Rising sea levels, sea temperatures and ocean acidity will not be stopped otherwise. "The thinking behind this is that it's a stop-gap measure. We've got to do other things to address CO2," says Helen Fox of the WWF, who works on reef regeneration in the Komodo National Park.

    Steve Palumbi, a coral scientist at Stanford University in California, says that nothing will stop the demise of the reefs unless the world addresses its addiction to fossil fuel.

    "Planting coral has been successful, but it's really expensive to do, so you can only do it over a given area. The other alternative is to allow them to do it for themselves," Palumbi says. "I doubt that we can replant all the coral reefs in the world and get them back in a century. It's a really important concept, but you just can't rely on that."

    Joan Kleypas probably speaks for the scientific consensus when she says that reef regeneration is just one important - albeit expensive - component in the conservation toolbox. But she says there is no way it will work unless carbon dioxide levels are stabilised at some point this century. "We need to think of things that can buy time so that when we do emerge from the climate crisis we'll still have some coral reefs left," Kleypas says.

    But would life really be so different if there were no coral reefs left in the world? Aside from the aesthetic loss of one of the most beautiful habitats on earth, corals are a vital source of food and provide a livelihood for a surprising number of the world's inhabitants - somewhere between 200 million and 500 million people.

    Rich Aronson points out that if all the coral reefs collapsed, then there would be broader implications for those of us who have no contact with reefs other than perhaps an occasional diving holiday.

    "Coral reefs are an example of the type of ecosystem that, if they go, will produce social consequences that will ripple back to us all," Aronson says. "If they are lost, it will make the world a more difficult place to live."