the plant. Others, including citrus plants, tomatoes, and aromatic herbs like sage, thyme, and mint, produce chemicals that either repel or poison insects. Decoding the genetic control of these protective chemicals could potentially let bioengineers alter the quantity or type of chemicals produced, or add a built-in defense to other plants. These natural defenses, boosted by genetic engineering, may be all the pest protection that some plants need.

As well as causing billions of dollars' worth of direct damage to crops by feeding on them, many sucking insect pests, such as whiteflies, aphids, and leafhoppers, transmit viruses and bacteria that cause devastating plant diseases. One new approach to controlling these particular pests depends on the fact that sucking insects have vital symbiotic bacteria in their bodies. The bacteria provide essential amino acids to their insect hosts, benefiting them in the same sort of way that symbiotic bacteria in the stomachs of cows help the cows digest grass. By exploring ways to inactivate these little-studied microbes, either by manipulating their genes or by engineering an antimicrobial agent into the plants the insects feed on, scientists would have a powerful way of indirectly controlling the pests.

One of the most successful agents of biological control, first discovered in the early 1980s, is Bacillus thuringiensis (Bt), a bacterium that makes insecticidal chemicals. When ingested by insects, the bacterial spores germinate and produce their toxins, eventually killing the insect as part of their own life cycle. Different strains of the bacterium make their own toxins, each of which has its own range