The strain of bird flu responsible for the deaths of tens of millions of chickens and 54 people in east Asia over the past two years is now circulating in long-distance migratory birds, potentially opening a way for the deadly virus to reach India, Australia and Europe.

That is the conclusion of two research teams whose findings were rushed into print by the rival journals Science and Nature yesterday.

Spread of the virus beyond its current home in China and neighboring countries could cause billions of dollars in losses to poultry farmers around the world. It could also give influenza A/H5N1 -- the virus's formal name -- further opportunity to adapt to human as well as avian hosts, a development that theoretically could lead to a global flu epidemic.

Until now, the H5N1 virus has chiefly attacked chickens and ducks in farms and markets. It also killed a small number of birds in two Hong Kong nature parks in late 2002, and since then has been found sporadically in hawks, herons and swans. Those birds presumably acquired it from direct contact with poultry.

Now, however, it appears the virus is being transmitted among wild birds that have had no known contact with domesticated birds.

"It has been difficult to tell whether the true migrating birds had been infected by this terrible virus. This leaves no doubt in my mind," said Robert G. Webster, a flu virologist at St. Jude Children's Research Hospital in Memphis who helped analyze virus samples collected during a recent die-off of birds at a huge saltwater lake in western China.

Since the first reports emerged on April 30, between 1,000 and 6,000 birds have died on the shores and islands of remote Qinghai Lake. The species most affected is the bar-headed goose, a large bird whose migration over the Himalayas to Burma, India and Pakistan starts in about a month. Illness and death were also recorded in brown-headed gulls, black-headed gulls and great cormorants.

There is a web of migratory flyways around the globe. The ones taken by the species congregating at Qinghai Lake intersect with others that lead to Europe. That theoretically provides a way for the H5N1 virus to reach that continent.

H5N1 influenza virus was first detected in southern China in 1996. In 1997, it caused a major outbreak in Hong Kong, which led to the death of 1.5 million poultry and six people.

The virus most recently emerged in South Korea in late 2003. Since then, it has led to the death of 100 million to 200 million chickens in China and Southeast Asia. It has also infected 108 people (most of them in Vietnam), of whom 54 -- exactly half -- have died. Most human victims had direct contact with dead or dying chickens, but in a few cases it appears the virus was acquired directly from an infected person.

While person-to-person spread of H5N1 influenza is rare and occurs with difficulty, the more the virus circulates the greater its chance of acquiring genetic changes that permit easy human transmission.

If that occurs, the virus would have "pandemic potential"; it could travel quickly and infect much of the world's population, which has no immunity to it.

There is no guarantee H5N1's presence in migratory birds will lead to global dissemination. It simply increases the chance.

For there to be further spread, a significant number of infected birds would need to be healthy enough to start their migration. They would need to establish a "chain of transmission" in the migrating flock, with new birds acquiring the virus as the infected ones died or recovered. At their destinations, they would have to make contact with poultry, igniting a new chicken outbreak and again putting the virus into contact with human beings.

The likelihood of any of these steps is unknown.

"What would migratory birds contribute to the possibilities of disease outbreak? That is the question we don't know the answer to," said David E. Swayne of the U.S. Department of Agriculture's Southeast Poultry Research Laboratory in Athens, Ga.

How the Qinghai Lake birds acquired H5N1 influenza is unknown.

There are chickens in Qinghai Province, but "there is no H5N1 infection in those chickens -- they don't have it," George F. Gao of the Institute of Microbiology of the Chinese Academy of Sciences said in a telephone interview from Beijing. He is the lead author of the paper that was published online by Science.

Both his team and one from the University of Hong Kong, whose report is published online in Nature, detected in the Qinghai Lake samples the three genetic defects and mutations found in the H5N1 strains responsible for high mortality in chickens and humans.

According to the two reports, the wild-bird strain bears genetic features of the virus found in chickens in China in 2003 and 2005 and in a peregrine falcon in Hong Kong in 2004. It is not identical to any of them, however.

The leader of the Hong Kong team, Yi Guan, a microbiologist at the University of Hong Kong, said the Chinese Ministry of Agriculture closed the Qinghai Lake area to his colleagues in mid-May.

"We hope they will open the door and let us in to do long-term surveillance," he said yesterday from Hong Kong. "There are a lot of questions waiting for answers."