The tree outside my front door is safe for now. It stands there, half its leaves brown and nibbled, looking like a banquet table deserted by guests in the middle of the salad course.

That's pretty much what happened. The guests -- rude, greedy creatures -- were gypsy moth catepillars. They had just eaten their way in from the suburbs and begun on my tree when nature called them to their cocoon.

This time I was lucky. But looking at the leaves I remember the original gypsy moth immigrants, whose descendants have decimated the Northeast. They were brought to this country from France by a scientist who thought they would produce silk.

Anybody could make a mistake.

While my tree stands in Boston, the people in California are dealing with another imported pest, the Mediterranean fruit fly. This week the planes sprayed people as well as land. The public's outrage there was palpable.

How did this new outbreak occur? In part because the 200,000 sterile fruit flies released to mate turned out to be fertile. "We got burned on a shipment from Peru," said a state official.

Sterile. Fertile. Anybody could make a mistake.

On both sides of the country, then, we have examples of that fun couple, scientific method and human error. Call it the Killer Bee Syndrome if you will. Someone sets out to breed a bee that will produce lots of honey; some technician unlatches the cage with the killer bees. p

Anybody who could make a mistake usually does.

The most scientific system in the world with fail-safes and triple-checks and computer backups is devised by people, run by people, used and misused and screwed up by people. To put it as simply as possible, the more dangerous the science, the more terrifying our fallibility.

This is something I would like to see cross-stitched on the walls next week when the leaders of all the major industrial countries get together in Ottawa. One of the chief subjects on the agenda is that ultimate killer bee, nuclear "know-how."

For decades, America has been the chief exporter of the most dangerous scientific species. Not only have we built bombs and used them, we have passed out most of the nuclear information for what we used to call Atoms for Peace. We've exported uranium, exported 60 small research reactors, and loaned $5 billion for 70 commercial reactors for energy production.

While the construction of nuclear plants has slowed in this country -- by the sheer tug of public protest over safety issues -- we continue to sell overseas, the same way we sell banned chemicals. We worry about our own technicians, our own safety standards, our own hazardous waste. But we regard "foreign" problems as if they had some private stock of air and water.

More to the point, each "peaceful" nuclear reactor produces the raw material for nuclear weapons. One 1,000-megawatt reactor produces enough plutonium for more than 20 bombs a year. That is why Israel bombed Iraq's research reactor, which can also make plutonium, though on a much smaller scale.

The Reagan attitude toward countries that want to develop the bomb was expressed best in his campaign quote, "I just don't think it's any of our business."

That's changing now, but slowly. In preparation for Ottawa, the administration has come out with a stronger statement against extending the nuclear "family." But it also promised to remain a "clearly reliable and credible supplier" of nuclear technology for peaceful purposes.

The problem is that no one knows how to control the spread of nuclear weapons while expanding the market for nuclear energy. Even the Reagan government offers only some vague idea about monitoring or retrieving the plutonium from foreign countries.

In short, we still seem to be dealing with nuclear bombs the way we've dealt with handguns. I can almost see the bumper sticker: "A-Bombs Don't Kill People. People Kill People."

But the more bombs we build and store, and the greater the number of people involved, the greater the risk. The risk is multiplied by each country -- with its own enemies and instabilities, its own leaders and technicians -- that gets nuclear knowledge.

Surely, nobody wants to blow up the world. But, as I can tell from my tree, anybody can make a mistake.