Bees can be conditioned to seek out materials. We have found that bees can be trained to find explosives, drugs, and military unique chemicals. We are investigating whether they can detect explosive vapors at concentrations given off by leaking landmines. We know that they can detect the odor of DNT, a common explosive, at sub part per billion levels, and possibly lower. Training bees to find things by odor is similar to the training that dogs receive to find drugs or explosives. Dogs are conditioned to expect praise and attention or play when they find an item of interest, bees are given a rich syrup reward.

Bees can find things by odor or visual clues or both. One of the videos on this web site shows a bee flying through a maize. She is following a series of dots in the expectation of finding a syrup reward at the end of the maize. We have successfully flown bees over a residential area to a site about 1 1/2 miles from their hives, and then were able to get them to search for and find targets marked by a drop of odor. In greenhouse trials in Montana and field trials in Montana and New Mexico, bees have proven to be capable of finding targets marked by explosives (DNT, TNT, RDX).

Because explosives are not very volatile, they do not provide much of a signal. Even dogs and sophisticated instruments have problems detecting trace amounts of these substances. Detecting more volatile chemicals is easier. Drugs present a far less challenging problem, mainly because aromatic, volatile chemicals are often used in their manufacture. For example, dogs may be able to find cocaine by the odor of methyl benzoate. We are currently testing the responses of bees to chemicals associated with the manufacture of drugs.

In order to train a bees to seek out a substance, the bee must be able to detect the material's odor at a level likely to be encountered (i.e., in the soil, a suitcase, a box, from a vent, etc.) and not be repulsed by the odor. Bees can be easily conditioned to odors that they respond positively toward such as many floral scents, can often be trained to substances that they naturally ignore, and can even be conditioned to seek out some things that they might normally avoid. But, there are some chemicals that induce a strong negative response in bees. Beekeepers use that knowledge to drive bees out of honey boxes. The beekeeper pours a chemical on an absorbent pad called a fume pad, sets the pad on top of the box of honey, and the bees quickly leave. Usually, the chemicals that drive bees out of hives have a very strong odor that is easily perceived by the human nose and we also find the odor to be repugnant.

Determining that bees have found something of interest presents a different challenge. If the material of interest can be presented to the bees (such as using restrained bees and the proboscis extension response bioassay), detection is simply a function of whether the bee extends its mouthparts. When using free-flying bees to canvas large areas, detection may rely on one or more approaches:

• Chemical residues brought back to the hive
• Chemical or physical markers (taggants) brought back to the hive
• Visual or photographic confirmation of bees at the target (odor location)
• Tracking device attached to the bee

All of these approaches have and can be successfully used. Tracking of bees over a kilometer has been accomplished by Joe Riley in England. However, his system is not readily available and requires the attachment of a diode with a long antennae to each bee. DARPA has funded a project to develop a tracking chip that weighs less than 1 mg, has a self-contained antennae, and is approximately the size of a grain of salt. The first prototypes are expected to become available in the summer of 2001.