Much has been written in recent months about the role of wild birds in spreading the Highly Pathogenic Avian Influenza (HPAI) H5N1 virus. But there is a distinct lack of evidence to support these assertions.

Tens of thousands of apparently healthy wild birds have been tested for HPAI H5N1 over the last decade. According to the UN's Food and Agriculture Organization (FAO) in November 2005: "To date extensive testing of clinically normal migratory birds in the infected countries has not produced any positive results for H5N1."

Does the westward spread of HPAI H5N1 from the origins of the current outbreak in China implicate wild birds? "No species migrates from Qinghai, China, west to Eastern Europe," says Dr Richard Thomas, BirdLife International's Communications Manager. "When plotted, the pattern of outbreaks follows major road and rail routes, not flyways. And the absence of outbreaks in Africa, South and South-East Asia and Australasia this autumn is hard to explain, if wild birds are the primary carriers."

Are other transmission methods plausible? Movement of infected poultry and poultry products is a likely cause of spread. South Korea and Japan are two countries to have suffered outbreaks of H5N1 in poultry and wild birds following importation of infected duck meat. Both countries stamped the virus out by culling infected poultry around disease areas, and imposed strict controls on poultry and poultry meat imports.

"Neither country has suffered a recurrence of the virus despite the influx each autumn of hundreds of thousands of wild migrant birds," Richard Thomas points out.

US experience in the 1980s with controlling an outbreak of H5N2-a strain of avian influenza non-infectious to people-in domestic poultry is relevant. A report found: "AI virus transmission was possible and occurred through movements of live and dead birds, contaminated equipment and vehicles, contaminated eggs, feed, water, insect vectors, and human vectors. In fact, any fomite that had contact with contaminated manure was capable of transmitting the virus. It was remarkably easy to isolate AI virus anywhere or from any inanimate object associated with an infected poultry flock."

Outbreaks at fish farms in Qinghai, Romania and Croatia may also be significant. The use of poultry manure as feed in fish-ponds appears to be widespread in east and south-east Asia and across Kazakhstan, southern Russia and west to Ukraine, Moldova and several eastern and central European countries. A joint FAO, WHO (World Health Organisation) and OIE (World organisation for animal health) conference in July 2005 reported: "In farming of poultry, high risk production practices include the farming of multiple species of animals, including poultry and waterfowl, within one farm unit; the keeping of chickens over fish ponds; the use of untreated chicken faeces as fertilizer or livestock feed..." The current advice on the FAO website warns: "the feeding of poultry manure/poultry litter should be banned in countries affected by or at risk from avian influenza, even if correctly composted, ensiled or dried with heat treatment." Manufacture and movement of fertiliser and feed in which the virus remains infective could be a very effective way of spreading H5N1 over long distances.