Can Bees Find Landmines?

The popular press has published numerous stories about our recent work. Many reporters have not talked to us, yet quote us extensively. One publication claimed to have interviewed a person who said he worked with us. Just one problem, we've never met him and don't even know whether he exists.

Can bees find landmines? We think so, and our data supports this idea. On military sites, we have detected traces of explosives brought back to the hives by bees. Our success in measuring explosive vapors in the air inside a hive is due to the development of specialized sampling tubes by Mike Guerin, Mike Sigman, and their associates at Oak Ridge National Laboratories. Our ability to detect explosives in bees and pollen is a result of methods developed by Philip Rodacy, Susan Bender, and their colleagues at Sandia National Laboratories.

Placing hives on a site and measuring chemicals brought back to the hive is a technique that we call passive monitoring. This approach works for monitoring a variety of chemicals, including trace elements and heavy metals (e.g., arsenic, lead, cadmium), radionuclides (e.g., radioactive cesium, cobalt), and organic chemicals, including volatile solvents (e.g., trichloroethylene, carbon tetrachloride), pesticides, and explosives. Recent work conducted by our group indicates that bees may also be useful for detecting microbes in the environment (e.g., microbial pesticides, biowarfare agents).

Bees can be conditioned to search by odor, sight, or a combination of odor and visual markers. Using bees like flying bloodhounds is a technique that we call active monitoring. We use a three step approach to determining what bees can find by odor and at what concentrations. We can test the ability of a bee to smell an odor by simply constraining a bee in a straw, presenting the bee with a puff of air that carries the odor, and then rewarding the bee with a drop of sugar syrup. The next time that a bee detects the smell in a puff of air, she will extend her mouth parts, expecting a syrup reward. The bee is not harmed, and she can be released at the end of the trials. This test is termed the proboscis extension response test (PER).

If the PER demonstrates that bees can detect an odor of interest to us and at a reasonably low concentration, we then move to greenhouse trials. In these tests, free-flying bees have to find targets marked by an odor. This is a sophisticated version of the game of finding the pea under a shell. Again, the bees are conditioned by first marking syrup feeders with an odor and then removing the feeder, causing bees to search for the odor which they associate with a sweet reward. Finally, we move outside for field trials.

To date, bees have proven to be able to detect the odor of DNT, an explosive, at levels below 20 ppb. They can also find soils containing DNT, both in the greenhouse and outside. Currently, we are investigating the lower limits of a bee's ability to detect DNT. Landmines are designed not to leak explosives, although most eventually do. In addition, mines are buried. As such, the chemical signature of a landmine is difficult to detect. A dog's nose is generally considered to be the gold standard, exceeding the ability of most field-portable instrumentation. Whether a bee's antenna is as good as the dog's nose is the subject of our current research.