A UCSF researcher who found the SARS virus in 2003 and later won a MacArthur Foundation "genius grant" for his work thinks he has discovered a culprit in the alarming deaths of honeybees across the United States.

Tests of genetic material taken from a "collapsed colony" in Merced County point to a once-rare microbe that previously affected only Asian bees but might have evolved into a strain lethal to those in Europe and the United States, biochemist Joe DeRisi said Wednesday.

DeRisi said tests conducted on material from dead bees at his Mission Bay lab found genes of the single-celled, spore-producing parasite Nosema ceranae, which researchers in Spain have recently shown is capable of wiping out a beehive.

"It is wise to strike a conservative note, because this is early data, but it is interesting,'' he said.

Government scientists who have been tracking the phenomenon they call Colony Collapse Disorder were skeptical, however, saying the parasite had been an early suspect in the bee die-off but that they had concluded it probably was not responsible.

With a mounting sense of urgency, agricultural scientists are trying to find out just what has caused the disappearance of as much as a quarter of the nation's 2.4 million honeybee colonies since November, when the die-off was first observed by a Pennsylvania beekeeper.

It's not just bad news for beekeepers and honey lovers. Growers of fruits, nuts and many vegetables rely on honeybees to pollinate their crops, which contribute $15 billion to the nation's agricultural output, according to a Cornell University study.

DeRisi is a specialist in the rapid identification of killer germs. In March 2003, he played a key role in helping the federal Centers for Disease Control and Prevention identify the cause of SARS, or severe acute respiratory syndrome, the viral illness that claimed 774 lives and wreaked havoc for a time on the Asian economy.

Using a laboratory tool called a microarray -- which can instantly match a sample to gene sequences from more than a thousand viruses -- he found that SARS was caused by a previously unknown variant of coronavirus, a microbial family responsible for a variety of ailments including the common cold.

The following year, he was awarded a $500,000 MacArthur Fellowship, the prize given by the foundation to individuals who have no idea they were nominated until they win. The awards are popularly known as genius grants.

In researching the bee die-offs, DeRisi's team evaluated samples of potential bee pathogens supplied by the Army's biodefense laboratory, the Edgewood Chemical Biological Center at the Aberdeen Proving Ground in Maryland.

Scientists there had developed a technique to concentrate possible pathogens into a sample that could be run through a rapid genetic screen test such as DeRisi's. Samples taken from dead bees in a collapsed colony from Le Grand (Merced County) were shipped via overnight mail to DeRisi's San Francisco lab last week.

DeRisi used a technique that allows rapid reading of the genetic code of the suspect bug. It is the same approach, known as "shotgun sequencing," that has been used to read the genomes, or the genetic code, of creatures ranging from bacteria to human beings.

The strips of genetic code are then matched to computerized libraries of known genes from thousands of germs. It was this test that pinpointed Nosema ceranae.

"The bees must have been loaded with this stuff,'' said DeRisi, who collaborated in the experiment with Dr. Donald Ganem of the UCSF Department of Microbiology and Immunology.

Fueling the UCSF scientists' interest in the parasite is a recent paper, published by the Journal of Invertebrate Pathology in January, in which a team of Spanish researchers infected hives of European honeybees with Nosema ceranae. Within eight days, the colonies were wiped out.

The federal government's leading honeybee scientists, however, are not ready to conclude that DeRisi has found anything significant. Jeffery Pettis, research leader for the U.S. Agriculture Department's Bee Research Laboratory in Beltsville, Md., said reports suggesting that this parasite has recently appeared in the United States are simply wrong. "There are historical samples from the mid-1990s,'' he said.

Before then, the parasite was seldom seen outside Asia, where it favored a species of honeybee found only there. It did not cause colony collapse in Asia.

Now, Pettis said, tests have shown that Nosema ceranae has displaced a related strain that had been the dominant form of the parasite in the United States, Pettis said. However, large quantities of the microbe have been found in bee colonies that are healthy, as well as in those that have collapsed, he said.

Pettis said the parasite could simply be taking advantage of a newly developed weakness in the insects' immune systems. "Mostly we think of Nosema as a stress disorder of honeybees,'' he said.

It is possible that a more virulent strain of Nosema ceranae has evolved in the United States, but Pettis doubts it. "We can't rule it out completely,'' he said.

Evan Skowronski, senior team leader for biosciences at the Army lab and a friend of DeRisi's, said that because the stake are high, every important lead in the search for the cause of the honeybee deaths needs to be pursued.

"We're not ready to say this is it, but it is a pathogen of interest,'' he said.

Skowronski said there is no reason to think that the cause of Colony Collapse Disorder is "anything other than Mother Nature.'' However, he said that any natural threat to honeybees has major implications for the United States. "This needs a high level of attention,'' he said.

DeRisi agreed that more tests will be needed to prove or disprove the parasite's role in the disappearance of the bees.

"In our results, the control bees did not have it, and the sick ones were loaded with the stuff,'' he said. "It is going to take a lot of time to figure out.''

E-mail Sabin Russell at srussell@sfchronicle.com.