The Breathing of Bumblebees

Hovering seems like a strenuous activity, and common sense suggests it should require greater exertion to stay aloft in one spot than to move forward at a moderate pace. Only at higher speed, when atmospheric drag is also higher, should forward flight and hovering cost the same amount of energy.

A standard theory of animal aerodynamics does hold that a flying animal uses about as much energy to hover as it does in fast flight. Theories exist to be tested, but how could one test this hypothesis?

With great difficulty, it turns out. The standard means of quantifying an animal's energy consumption is by measuring its oxygen uptake. Now, if your study animal is a human being, measuring oxygen consumption becomes a pretty straightforward problem. You need merely convince your experimental subject to wear a device like a gas mask, covering nose and mouth, while walking on a treadmill or otherwise exercising. The automated measurements of oxygen consumption are made by well-tested equipment that's easily available.

But if your subject is a bumblebee--a handy hovering animal--you've got a very different problem. First, not only would any bee-fitting mask need to be very, very small, it would have to deal with a couple dozen nostril-equivalents. Bumblebees, like other terrestrial arthropods, respire through spiracles, breathing holes arranged along the animal's thorax and abdomen. And to measure the effects of a bee's breathing, the equipment would have to be so sensitive it could detect changes in atmospheric oxygen concentrations down to five parts per million or less. Nobody's got equipment like that on the shelf.

Then there's the problem of convincing the bee to hover while you make the measurements. A bumblebee startled into flying won't want to hang around once it's airborne. Explaining what you want bees to do doesn't affect what they will do.

Now you have some idea of the challenge that Charles Ellington took up. Ellington believes he succeeded, but he needed the help of colleagues at his own University of Cambridge in Britain and at Rutgers University in new Jersey.

What this international team came up with amounted to pairs of small, sealed wind tunnels. One served as a control: in all dimensions it was identical to its mate, but it never held any bees. The other was the working chamber. The plan was to insert a bumblebee into that chamber and get it flying into an artificial head wind. The system would permit testing a range of bee speeds, from hovering to really buzzing off, by varying the wind speed in the tunnel. The extremely small difference in gas pressure between the working chamber and its empty twin showed the bees' rate of oxygen consumption.

The researchers' painstaking measurements during the wind-tunnel tests indicated that hovering bumblebees consumed about the same amount of oxygen, and thus presumably expended similar amounts of energy, as bees flying forward at a moderate speed of four meters (13 feet) per second, which is nowhere near bee maximum. The bees in the wind tunnel had apparently overturned the accepted theory.

Hot on the trail of a new theory, Ellington and graduate student Robert Dudley recently presented an aerodynamics analysis suggesting that moderate, rather than fast forward speed, actually should require about the same mechanical power as hovering, the two activities thus taking about the same energy.

Meanwhile, other researchers are just as hot to prove them wrong. Tests continue, with other animals capable of hovering. Bumblebees are so uncooperative that Ellington's team had few subjects to measure. They put more than 100 bees, one at a time, into the wind tunnel. Only six bees would fly at all.

The whole affair demonstrates how raising meaningful questions is only one part of a scientist's job. Figuring out a good way to test an idea may be as great a challenge.