GENETICS MAY BE THE BUG WARS' big-ticket strategy—a kind of high-tech entomological bunker buster—but battles can also be won from unexpected corners. While Jacobs-Lorena was breeding mosquitoes in Baltimore, doctors working in malaria-plagued rural villages made an interesting discovery of their own: Some houses attract hundreds of mosquitoes, while others attract almost none. In part, this is a matter of location—the closer the bogs, the thicker the bugs. But it also seems to depend on who lives there. According to recent studies, it now looks like about 20 percent of the people attract 80 percent of the bites. (If you've always felt like the most bitten person in the group, in other words, you might be right.)

What makes a person a pied piper of the mosquito world is a subject of great interest to epidemiologists,

Kline shows me his right forearm, which is dotted with lurid red welts—products of an Australian colony's morning meal. "Technically, you're not supposed to feed them off yourself," he says, shrugging. "But sometimes it's the only way they'll eat."

and understanding "the 20-80 problem," as it's known, has become one of the oddest fronts in the mosquito war. In public-health circles, pied pipers are known as "superspreaders," because they are more likely both to get malaria and to pass it along, by infecting the many mosquitoes that bite them. Identifying superspreaders would be a boon to prevention, since they're responsible for the vast majority of new infections: Cure them and you help break the cycle.

The difference between an average person and a superspreader is subtle, although there are clues. Old Africa hands have long noticed that people who drink heavily tend to attract more mosquitoes, as do people who rarely bathe. Ten years ago, Austrian researcher Bart Knols decided to do some tests. He released four species of mosquito into a tall, screened box, then stripped to his underwear and took notes on where they bit.

Tastes varied, he found. The European anopheles, for instance, gravitated to his face, while the African anopheles hung around his feet. Knols then analyzed the bacteria from his feet and experimented by offering the insects a wedge of Limburger cheese, which is cultured with the same odor-producing bacteria. The African bugs swarmed it. But when Knols reported his findings in a journal article, the scientific community was doubtful. "Unfortunately, the article happened to be published in the April 1 issue," Knols says. "Everyone thought it was a joke, attracting mosquitoes with a dairy product."

Nonetheless, the idea—that subtle variations in smell contribute to appeal—was tantalizing. Solve the odor mystery and we could find the superspreaders or create irresistibly scented traps and lure millions of bugs to their deaths. So far, though, the details have proved hard to pin down. The problem is that mammals, crudely speaking, tend to smell the same. Two chemicals that humans exude copiously—lactic acid and carbon dioxide—are so abundant they tend to drown out fainter scents. "It's tough enough to figure out what distinguishes a person from a cow," says Laurence Zwiebel, a professor of biological sciences at Vanderbilt University. "Figuring out what distinguishes one person from another—bearing in mind that people's odors change daily, based on things like stress level and diet—that's really tricky."

Knols eventually left to pursue other projects; he's now overseeing the construction of the sterile-male-mosquito-breeding facility for Sudan. Replacing him on the odor front is Ulrich Bernier, a slight man who smells faintly of sport deodorant. A chemist with the USDA's Mosquito and Fly Research Unit, in Gainesville, Florida, Bernier became interested in smell as a grad student and has spent the past ten years attempting to compile a comprehensive catalog of human odors and their effects on mosquitoes. One recent sample population included a lab technician, a departmental secretary, and Bernier's immediate superior, each of whom spent half an hour with their arms and feet inside bags full of nitrogen gas. When Bernier piped the three mixtures through a gas chromatograph—a machine that identifies the presence and abundance of chemicals—the resulting graph revealed 350 compounds. Bernier combined the 30 most dominant compounds into a batch he felt sure mosquitoes would flock to. Such a concoction could be used to bait a trap—hang one in the yard!—and could provide clues when it came to creating a truly effective repellent. "We thought, This is going to be great!" Bernier recalls.

Bernier and his colleague Dan Kline put a capful of the solution in an olfactometer, an apparatus that consists of two traps and a Plexiglas box full of mosquitoes. Based on the number of mosquitoes that fly into the traps, one can measure a scent's attractant power. Ten minutes passed, then twenty. "It was like Silent Spring," Kline says, grimly. "We didn't get a single mosquito."

Shortly after I meet Kline, he takes me on a tour of the USDA insectary, where different mosquito species are bred. It's a humid space lined with finely woven mesh cages, one of which has a live chicken lashed on top. Typically, Kline explains, larvae are fed yeast and liver powder, while adults get a sugar cube and, once a week, a tube of fresh beef blood suspended in a natural-membrane condom. Some species, however, are such picky eaters that they will feed only on a living animal. Kline shows me his right forearm, which is dotted with lurid red welts—products of an Australian colony's morning meal. "Technically, you're not supposed to feed them off yourself," he says, shrugging. "But sometimes it's the only way they'll eat."

Such finickiness indicates that it may be even harder to fool mosquitoes than anybody expected. Bernier has used his tests to develop more successful compounds, several of which he is in the process of patenting. Even with his best mix, however, 95 percent of mosquitoes will still choose a person over the artificial attractant. "The fact is, we're hard to beat," says Bernier. "We have all the chemicals, in just the right ratio. Mosquitoes can tell that." And while he has yet to discover the blend of chemicals that mark a person as particularly enticing, he did recently come close to identifying several that appear to be off-putting, produced by the departmental secretary (who revealed that she rarely gets bitten).

Bernier thinks that altering our smell even a little could muddle a mosquito's radar, effectively rendering a person invisible—or at least undesirable. This, in turn, could spare us from more toxic chemicals like deet, which is strong enough to eat away at plastic in high concentrations, and permethrin, a neurotoxin that shouldn't be applied directly to the skin. Bernier recently began testing an extract made from chicken feathers that he hopes will confuse mosquitoes by making a person smell like a large, unusually malodorous bird. He admits that a chicken-scented repellent is unlikely to catch on in a place as deodorant-obsessed as the U.S., but he holds out hope that it could flourish somewhere. "It might work for a species like the anopheles, that bites only people," he says. "You wouldn't want to wear it around mosquitoes that feed on birds, of course."