The ornithological community has been divided by a proposal for a new definition of species which could double the number of bird species in the world to 20,000.

Some specialists are predicting that a two-tier system could result in certain academics using the new defintions while some ornithologists, bird-watchers and conservationists use the old one.

The controversy has arisen because of criticism of the biological definition of species, which says that they are groups that are incapable of breeding with members of another species. The critics say that this definition is arbitrary and some point out that about 10 per cent of bird "species" can in fact mate successfully with other species.

An example is the case of the North American ruddy duck, a controversial cull of which is just beginning in Britain. The duck escaped from wildfowl reserves in England and has reached southern Spain where it is threatening the white-headed duck. But it is threatening the white-headed duck by successfully breeding with it, which raises questions about the whether the two are different species if the biological definition is used.

³ÉÈËVRÊÓƵ

Instead critics say a phylogenetic definition should be used. This is based on the idea that any difference between two groups of birds could mean that they are different species. Only by looking at a bird's DNA and tracing its evolutionary tree can those with a common ancestry be put together to form a species.

German researchers, led by Andreas Helbig of the Vogelwarte bird observatory, are expected to announce soon that the chiffchaff, which breeds and summers in the United Kingdom, is in fact three and probably four distinct species.

³ÉÈËVRÊÓƵ

David Parkin, an avian geneticist at Nottingham University, who heads the British Ornithologists' Union's definitive list of birds said: "We will probably recommend that the British list accepts the revision. " Proponents of the new definitions, which caused upset at a recent British meeting on avian taxonomy, say that techniques, such as DNA sequencing would be more objective. "The biological species concept is beginning to outlive its usefulness," said Dr Parkin.

Graham Martin, professor of ornithology at Birmingham University, said that bird-watchers are ambivalent. If one species becomes two, there is suddenly another bird to tick off on the must-see list, which is good.

But if distinguishing between species becomes impossible to the naked eye then spotting new species will be impossible for the non-scientist. The proposal has also caused problems with conservationists. Nigel Collar, research fellow at Birdlife International, says that doubling the number of bird species would cause a dramatic increase in the number defined as threatened. This would dilute attempts at protection and confuse politicians.

"I do foresee trouble ahead," Dr Collar said.