The disappearance of large numbers of U.S. honeybees is so odd that it's attracted Ian Lipkin. Since last fall, beekeepers in at least 35 states have reported colonies that shrank rapidly for no apparent reason. Adult bees just go missing, leaving behind young bees in need of tending. This colony-collapse disorder (CCD), as it's now called, has got bee researchers coast to coast stirred up and looking for causes and remedies.

Honeybees pollinating crops (here, canola) add an estimated 5 billion to U.S. agriculture by boosting yields and quality. Hence the concern when beekeepers in most states (inset) reported mysterious colony collapses.

Lipkin, however, had never studied a bee disease until now. He works in the epidemiology department of Columbia University's Mailman School of Public Health - human health, that is. He's solved mysteries, though, and he says that his methods are yielding results this time too.

Lipkin is the pathogen hunter who in 1999 figured out that a cluster of people with encephalitis in New York had caught a then-obscure virus called West Nile. Since then, his lab has refined ways to use high-speed genetic sequencing to search for novel pathogens worldwide. What involved him in this insect-disease case, he says, is "the same thing that has captured the imagination of the public - the notion that there's been this inexplicable loss of bees."

Following last winter's losses, beekeepers have had some success in rebuilding their hive numbers. But they remain concerned that next winter, their colonies may again suffer unexplained collapses.

©iStockphoto; USDA

It's a good mystery all right, with any number of hypothetical culprits: mites, bad bee food, cell phones, bee AIDS, pesticides, genetically modified (GM) crops, overwork. Jeff Pettis, based in Beltsville, Md., as head of the U.S. Department of Agriculture's network of laboratories devoted to bees, even suggested to the Washington Post that bees had worn themselves out making crop circles, thus explaining two mysteries at once.

Joking aside, Pettis, his government-bee-lab colleagues, Lipkin, and other researchers have been working in earnest on the problem. So far, they've eliminated several hypotheses. Now, they're mixing old-fashioned case study epidemiology with modern genetics. It now looks, says Pettis, as if "more than one factor may be coming together" in the mystery of the missing honeybees.

When to worry

Beekeeper Dave Hackenberg, of Lewisburg, Pa., rents out his bees as migrant laborers, shipping hives around the country as crops come into flower and need pollination. Like baseball teams, his and many other big bee operations retire to Florida for the winter to get their workforce in shape for the stresses of the upcoming season.

But when Hackenburg checked on his hives there last November, he found that they were doing the opposite of recovering. He called Pennsylvania State University in University Park and reported that adult bees were disappearing without obvious cause.

Hackenberg's concern alerted researchers at Penn State and elsewhere to the problem, though Frazier says that once they started asking around, earlier cases turned up. In the typical case, colonies dwindled in a matter of weeks. Large numbers of bees simply vanished, and the few that remained showed a loss of appetite. Bees in neighboring colonies reacted oddly too. Instead of raiding an abandoned store of honey as soon as possible, they left the afflicted hive alone for days.

To start tracking down the cause of Hackenberg's troubles, state apiarist Dennis vanEngelsdorp, based at Penn State, and his colleagues undertook detailed case studies of affected beehives.

In mid-December of last year, vanEngelsdorp's team released its first results. From interviews with seven beekeepers in four states, the researchers learned of substantial losses. One beekeeper expected to lose all but 9 of his 1,200 colonies. A hive can empty and collapse in as little as a week, the beekeepers said.

These interviews, and others that followed, didn't disclose an obvious cause for CCD, but they "ruled out some things," says Frazier.

Beekeepers said that they had procured queens from a variety of suppliers, so it seemed unlikely that a bad batch of queens had spread a disease or genetic problems to the colonies of their offspring. They used a variety of mite-controlling drug regimens on their hives, so the drugs weren't the obvious answer. And the colonies had been given several kinds of diet enhancements, such as corn syrup or protein supplements, so feeding practices didn't seem culpable.

What the bees did have in common, though, was recent stress, such as from a lot of travel between assignments. Stress could have left the colonies vulnerable to some other menace, the researchers speculated.

As news spread about the trouble last winter, bells rang for memories of past cases of honeybee-hive disasters, says Jay Evans of USDA's Beltsville, Md., bee lab. He cites a 1975 paper titled "Disappearing disease of honey bees" in the American Bee Journal. That report cited the paper "Bees evaporated: A new malady" in an issue of Gleanings in Bee Culture from 1897. These old reports raise the possibility that a bee pathogen is always lurking in hives but occasionally flares up in an especially virulent form. "It could be like the Spanish flu," says Evans. Flu is ever present, but the legendary 1918 epidemic killed an estimated 25 million people worldwide.

Maybe not

As winter and spring passed, researchers failed to find a clear answer, but there was no shortage of ideas.

The most infamous, so far, may be cell phones, described in the British newspaper The Independent in mid-April under the headline "Are mobile phones wiping out our bees?" The story referred to a "limited study" at the University of Koblenz-Landau in Germany reporting that "bees refuse to return to their hives when mobile phones are placed nearby."

Pettis wondered why, if this were true, he was having such a hard time getting phone reception when visiting afflicted hives in rural areas. But what deflated this hypothesis most dramatically were the German researchers themselves, who denied that their work had anything to do with colony-collapse disorder. "None of us ever wanted to do research on CCD," says Stefan Kimmel, a graduate student and coauthor of one of the phone-bee studies.

The researchers hadn't even used what Americans would call a cell phone but were experimenting instead with the base of a cordless phone. They were developing a setup to test for effects of electromagnetic radiation on honeybees, and for a source of electromagnetic waves had placed the phone's base unit inside a hive.

It was hardly a realistic test. Evans adds that the bees in the German group's experiment "may have just been offended by having this phone [in their hive]." For the time being, he says, the possible effects of electromagnetic fields on bees have moved down the list of CCD culprits. So has the suggestion that bees are losing their ability to navigate back to their hives because Earth is starting to reverse its magnetic field. Evans asks why that would upset U.S. bees more dramatically than it would bees in other nations, where CCD reports have not been as widespread.

By the end of May this year, Pettis says, he had virtually ruled out two more suggested explanations for CCD. "It's not tracheal mites," he says, referring to tiny parasites that infest the tubes making up the bee's breathing system. "I've personally looked at thousands [of bees from afflicted hives], and less than 10 percent of them have tracheal mites." Nor, he says, is the small hive beetle to blame. These common pests in fact seem unexpectedly rare in hives that have collapsed.


All this nay-saying leaves plenty of other possibilities. "My first instinct was, it has to be caused by varroa mites," says Pettis. Unlike tracheal mites, these pinhead-size, blood-sucking parasites attack the bee from the outside. First noticed in the United States during the 1980s, varroa mites feed on bees of all ages, causing deformities in young bees and weakening adults. Infestations can kill a colony.

©USDA-ARS Nathan Rice
A comb from a troubled bee hive in Georgia shows plenty of the covered cells where larvae develop but far too few workers left to tend them.

To check for them, Pettis washed batches of bees from collapsed colonies and examined the parasites that came loose. He found varroa mites on only about half of the batches of bees, making mites an unlikely culprit for the whole phenomenon. Still, "we can't rule out varroa," he says.

Pettis has begun experiments to see whether some infective agent remains within afflicted hives. He collected combs from hives that had collapsed and irradiated half of them at a facility normally used to sterilize medical equipment. He then introduced new bees into both the irradiated and untreated structures. He's keeping track of these colonies throughout the season to see if differences show up. All he can say so far is that bees in the unsterilized hives "didn't die instantly."

Another hypothesis is that pesticide exposure has either wiped out the colonies directly or contributed to their demise by enfeebling them. Rental hives travel extensively, and the bees visit fields treated with a wide variety of substances.

Bees may have encountered crops that have been genetically modified to produce their own pesticide, a bacterial toxin nicknamed Bt. The toxin is intended to target the caterpillars of moths and butterflies rather than bees, which belong to a different taxonomic order. In any case, one of the major Bt crops, corn, relies on wind for pollination and isn't a top pollen choice for foraging bees.

A new lab experiment found no effect on the weight and survival of honeybees fed for 35 days on pollen of a Bt sweet corn variety. Still, say Robyn Rose of the USDA's Animal and Plant Health Inspection Service in Riverdale, Md., and her colleagues, that doesn't rule out all possible risks from pesticide-bearing crops. In reporting their results, now online for a future issue of Apidologie, they suggest protocols to look for more-subtle effects.

A USDA research plan, released in July, raises another question about the genetically modified crops: European beekeepers have now reported die-offs too, and GM crops aren't grown there.

Bt crops are far from the only pesticide sources a bee might buzz into. For example, the neonicotinoids, a class of insecticides based on nicotine, are toxic to honeybees. The recent USDA report says that research it has funded suggests that widely used fungicides enhance the effects of the neonicotinoids. The report adds that one of them, imidaclorprid, has been found in pollen from corn, sunflowers, and rape at concentrations potentially harmful to bees.

Frazier is coordinating research on bees and pesticides. So far, she says, one analysis of pollen that bees had stored in affected and unaffected hives has been completed, but the work didn't point to a culprit. The colonies with the greatest variety of pesticides at the highest concentrations "are doing quite well," she says.

That doesn't settle the matter. Frazier says that she and her colleagues are designing direct tests of pesticide effects, for example by exposing caged colonies to chemicals in monitored quantities.

Some of Evans' work, too, has touched on pesticides. He specializes in honeybee genetics, and he's approaching the problem by looking at activity in genes known to kick in when bees encounter certain stressors. He and his colleagues are checking bees from affected and healthy colonies for any heightened responses by genes known to indicate exposure to pesticides or pathogens.

As of the end of June, no clear pattern had emerged in either the pesticide- or pathogen-related genes. "I keep looking," Evans says.

His project and others', he says, are hampered by the limitations of the samples. The disorder doesn't leave heaps of dead bees in the hives. The workers leave and presumably die far from home. Would the bees still buzzing around a depleted colony - the majority of bees researchers have been able to collect for study - still show signs of whatever afflicted their hive mates? Or would the survivors escape with no trace of the malady? "I think we would have solved this months ago if there had been more dead bodies," says Evans.

Lipkin uses genetics, too, but in a different way. He sequences genetic material from samples of bees from afflicted colonies and compares the results with sequences from healthy ones. The samples contain genetic material from the bees themselves plus whatever mites, viruses, fungi, or other creatures were living on and in the bees' bodies.

The critical step is in the analysis of this hodgepodge. Discounting bee DNA plus any traces of genes from the people who handled a colony, Lipkin and his collaborators look for genetic sequences that show up in sick colonies but not in healthy ones. Identifying the source of those sequences could reveal a pathogen.

Starting in March, Lipkin and his epidemiology team have thrown themselves into the search, working with bee labs. "This has been a huge project, absolutely huge," Lipkin says.

After all this work, Lipkin's tight-lipped about what his analyses have revealed. He will say, however, that his lab, with help from others, is closing in on a suspicious infectious agent.

It probably won't tell the whole story by itself, he says. Pesticides or some other insult might need to weaken the bees to make them susceptible to attack by a microbe. He hopes that the agent his work has tagged might at least become "an excellent marker" for colony collapse. In the end, it may be that what bee detectives need to catch their villain is a consult from human epidemiology.