Five years after the mysterious mass deaths of honeybees began, bees keep dying, yet in many ways the mystery seems as deep as ever.

Why can't we solve this, in a country with Canada's scientific resources?

Bees are important to anyone who wants food on the table. It's easy to forget this. Once, wild bees pollinated flowers and fruit trees across the continent. But the Europeans brought honeybees, and these have taken over the job.

Honeybees today pollinate apples, canola, soybeans, broccoli, celery, asparagus, squash, cucumbers, peaches, kiwi, cherries, blueberries, cranberries, strawberries and cantaloupe (though not corn or wheat).

Farmers of many kinds hire beekeepers to bring hives at pollination time. And in 2005-06, about one-quarter of the bee colonies in the central and eastern United States suddenly died - many just a week or so after a checkup found them strong and healthy.

In the past few years, Ontario has been losing one-third of its colonies each winter. The phenomenon is called "colony collapse disorder" in the U.S., but "winter colony mortality" in Canada.

The latest news comes from Ontario. At the University of Guelph, a bee scientist named Ernesto Guzman thinks he has solved a big chunk of the mystery, after studying 408 commercial hives for a year.

He picked five main dangers. With these, especially the top two or three, there's a strong chance of death to the hive. But hives free of these five problems are likely to survive, which strongly suggests that there isn't another mysterious killer lurking out there.

The big one is called the varroa mite. It's about the size of a pinhead, but that's big to a bee. It would be like a parasite the size of your fist crawling around on you, sucking your blood.

Bees are unable to pull them off. Just one mite can shorten a bee's life by half. This is especially bad in winter; bees have to survive longer in winter because the queen stops laying eggs then.

Other factors include colonies that are too small; lack of food stored in the fall; another parasite called the tracheal mite; and a fungus called nosema that gets into the bee's gut and makes it less able to absorb nutrients.

But the varroa mite appears to be the big killer, implicated in 85 per cent of the Ontario bee colonies that died in the Guelph experiment.

In this case, if Guzman is right, it's a mite that has been in Canada for 20 years.

So what's different now?

He suspects varroa has adapted to pesticides. If so, a beekeeper might spray mite killer and think all is well, only to have the mites survive and wipe out the hive.

But there are still indications of further mysteries.

In the winter of 2008-'09, colony collapse accounted for 10 per cent of all U.S. losses, says Dennis van Engelsdorp, the Pennsylvania state apiarist. In these cases the bees are somehow weakened and they all fall sick with several diseases at once.

Most colonies, he says, died from starvation or because the queen died.

"It's tempting to say everything is dying from one cause. But I think there are many things killing colonies," and for three years in a row now, some 30 per cent of the hives die each winter.

"I agree with Guzman that some loss is (caused by) varroa and some loss is nosema. But I'm not convinced that all the losses are those things."

Differences, he says, appear on a county-to-county basis - different land use, different flowers, new pesticides, slightly different climates. A recent French study says bees are healthier on a diet of pollen from mixed sources - even if they're not very good sources - than from one pollen type alone. That's a worry as agriculture shifts toward mass production of single species.

Bee researchers are right to wonder why the research effort is so scanty. Last year van Engelsdorp snapped to one reporter that if one in every three cows were dying, "they'd call out the National Guard."

Food is a must. And for a country with Canada's agricultural base, there's surprisingly little sign of enthusiasm here in finding a cure.