She looked perfectly normal. But what was she doing roaming around at night on the Google campus in Mountain View, Calif?

She’d been drawn out of her home, following the light, and now was taking mincing steps across a white bed sheet. Had she just taken “the flight of the living dead”? Was she actually a “ZomBee”?

Her hunter, the genial John Hafernik, was clearly excited about the possibility but, as a scientist, would only acknowledge what we had observed directly so far. “Well, leaving a hive at night is definitely not normal behavior,” said Hafernik, a biologist at San Francisco State University and trustee and president of the California Academy of Sciences, sounding almost cheerful.

We watched as another honeybee left the Google Crittenden Campus hives and buzzed toward the light traps Hafernik had set up. The insect dropped onto the slippery metal surface fitted into a plastic bucket below a light, sliding down through a hole and making a small plunk when she hit the base. We had captured another possible ZomBee—an Apis mellifera honeybee that might have been parasitized by the Apocephalus borealis, a type of phorid fly. The fly’s brood eats the living tissue of their soon-dead hosts, leaving empty, sometimes headless husks. As we watched, other bees gradually joined the first two in the bucket atop a white sheet. I shivered with feelings of horror and a kind of delight at the discovery. Or maybe it was just a cool evening.

It was Saturday, August 4, and a group, including Fred Guterl, Scientific American’s executive editor, and me, had followed Hafernik from our evening meeting site at SciFoo, the annual “unconference” organized by the O’Reilly Media, Nature Publishing Group, Digital Science and Google on its California campus. We were out to conduct some citizen science by finding the zombie-like parasitized bees. In a short time, our hunting party snared six honeybees acting suspiciously. I asked Hafernik: How do you know if they’re actually infected? “We watch them,” he said. Maggots will emerge after about two weeks.

From the six bees we caught, 11 pale, wriggling maggots later appeared and grew into 11 brown, pill-shaped pupae. Although the parasitizing fly creates ZomBees, the maggots are apparently not much interested in brains. The fly lays their eggs in the abdomen, or back end, of the bees. The maggots eat their way through the abdomen and thorax (middle) and exit below the head. Sometimes the head falls off.

As Hafernik explained, the honeybee is not a native species, so the A. borealis fly’s parasitization is a relatively new phenomenon, occurring only within the past few hundred years. The fly’s other known hosts are bumblebees and paper wasps. In their study, Hafernik and colleagues found more than three quarters of the sites sampled in the San Francisco area had infected honeybees. (Read the paper Hafernik and colleagues published in January in PLoS ONE, “A New Threat to Honeybees, the Parasitic Phorid Fly Apocephalus Borealis”).

Honeybee colonies have suffered from die-offs in recent years from colony collapse disorder, or CCD, where the insects abandon their hives. The bees are also subject to mites and other parasites. Studies of the phorid infection may help in understanding hive-abandonment behaviors in CCD.

The ZomBees that we had found in early August, Hafernik later reported in an e-mail to SciFoo attendees, were the southernmost record of an A. borealis infection in the San Francisco area. In his e-mail, he updated us about more hives: “On August 13, Googler and ZomBee hunt participant Bart Locanthi noticed unusual behavior around the honey bee hives at the Google West Campus. Apparently, honey bees were being attracted to a glass door behind the hives at night as Googlers burned the midnight oil inside. When I checked the areas, there were a number of dead bees with A. borealis pupae near them and a few maggots slithering around as well. Thus, these hives are also infected.” Here are Hafernik's photos of the bees, maggots and pupae. Hafernik hopes to continue to work with the Google beekeepers to monitor the hives.

After about an hour, the Google sprinklers came on, driving the hunting party away from the chilling scene of the living dead and back to the warmth of indoor conversations and the hope that we had contributed in some small way to science. As Hafernik later reassured us in his note: “The hunt was not only fun, but also produced useful data for the ZomBee Watch citizen science project.”

Want to help the scientists find parasitized bees? You can learn more at the ZomBee Watch citizen-science project.

Update: Hafernik shared this photo he shot this morning of more bees and pupae from Google hives.