Bird flu may be spread by using chicken dung as feed in fish farms, a practice now routine in Asia, the world's leading bird conservation organisation believes.
Fertilising fish ponds with poultry faeces, which can dramatically improve fish growth, may in fact set up major new reservoirs of avian influenza infection if the chickens providing the manure are infected themselves, according to BirdLife International, the Cambridge-based umbrella body for bird protection groups in more than 100 countries.
The suggestion, which has echoes of the BSE problem in Britain, in which cattle were infected by organic feed, is an explosive one, on an international scale.
It puts a serious question mark over a technique firmly backed by the UN's Food and Agriculture Organisation (FAO) as a primary means of providing protein for mushrooming populations in the developing countries - suggesting that millions of people, instead of being helped by it, might ultimately die because of it, in a global pandemic.
Known as integrated livestock-fish farming, the technique involves transferring the wastes from raising pigs, ducks or chickens directly to fish farms, with chicken or duck sheds sometimes sited directly over the fish ponds themselves.
At the right dosage, the nutrients in the manure give an enormous boost to the growth of plankton in the ponds, which are the main food of fish such as carp and tilapia.
This is taking place on an enormous scale. According to an FAO report from two years ago, it is now the main basis for aquaculture in China and neighbouring countries.
"Livestock wastes purposely used in ponds, or draining into them, support the production of most cultured fish in Asia," the report said.
"Much of the vast increase in China's recent inland aquaculture production is linked to organic fertilisation, provided by the equally dramatic growth of poultry and pig production."
BirdLife International is now calling for an investigation into the possibility that these thousands of manure-fed ponds across Asia may be the means by which the new, potentially deadly strain of avian influenza, H5N1, is being spread.
BirdLife points out that outbreaks of H5N1 have occurred this year at locations in China, Romania and Croatia where there are fish farms.
The Chinese outbreak, which involved a big die-off of 6000 wild birds, mainly bar-headed geese, took place in May this year at Qinghai Lake, a location where the FAO helped establish an integrated livestock-fish farm in the early 1990s, BirdLife said.
"This outbreak helped lead to the widespread media speculation about wild birds spreading H5N1," said BirdLife's Richard Thomas.
"We pointed out that bar-headed geese migrate from India, where H5N1 has never occurred, and migrate early, so they must have contracted the disease locally, at Qinghai."
BirdLife does not think that wild birds are vectors - carriers - of H5N1, and believes that the widespread speculation during the autumn that migratory birds would spread bird flu from Asia far and wide into Europe was entirely misplaced.
"Millions of birds have now migrated, and it hasn't happened," said BirdLife's Director and Chief Executive, Dr Michael Rands.
"Wild birds are victims, not vectors, of avian influenza. We believe that when wild birds get it, they're very susceptible, and die before they can move very far. Some fall over straight away."
Although no mention was been made of the possible links between manure-fed ponds and influenza in the recent alarm over bird flu, it has been raised before, and the FAO, although actively promoting the technique, is well aware of the threat.
A FAO report from 2003, "Integrated Livestock-Fish Farming Systems," noted: "Recently, livestock and fish have been implicated in the irregular occurrence of influenza pandemics; the global impacts on public health of promoting livestock and fish integration are huge if these claims are substantiated."
In fact, the FAO has been aware that some scientists think there is a risk for very much longer.
The references to the 2003 report include one to a paper published in Nature, the international scientific journal, as long ago as 1988.
This paper, by Christoph Scholtissek from the University of Giessen in Germany and Ernest Naylor from the University of Bangor in North Wales was headed "Fish Farming and Influenza Pandemics", said that bringing together fish farms with farm livestock "may well be the creation of a considerable human health hazard."
However, the FAO has continued to promote integrated livestock-fish farming actively throughout the ensuing period.
Dr Rands said: "There appears to be some circumstantial evidence that one of the ways in which the disease might be spreading is through the use of chicken faeces for feeding fish.
"We are not aware that people have confirmed that that is a likely cause - but people have suggested it. If it's a possibility, and if it presents a serious human health risk, it certainly ought to be researched, and if it has been researched, the research should be out there for us all to see."
He went on: "Wild birds are often being blamed for the spread of avian influenza, but as far as we can tell there is no clear evidence, in fact no evidence.
"The science that there is suggests that wild birds are not to blame for the spread. If you look at migration routes and spread of the disease, the correlation is really quite poor. Millions of birds migrated this autumn, and if they really were the means of spreading bird flu, it should be all over the place in Africa and southern Asia - and it's not happened."