DELAYING INSECT ADAPTATION TO TRANSGENIC PLANTS - SEED MIXTURES AND REFUGIA RECONSIDERED

Genetically engineered plants expressing insecticidal proteins from the bacterium Bacillus thurinpiensis can provide safe and effective pest control, but they will also select intensely for insect adaptation to the toxins. Mallet and Porter (Proc. R. Sec. Lend. B 250, 165 (1992)) used a model to compare two planting strategies for delaying crop failure: 'seed mixtures' of toxic and toxin-free plants within fields, and 'refugia' of toxin-free fields. They concluded that 'if insects can move from plant to plant, seed mixtures may actually hasten insect resistance compared with pure stands of toxic plants'. In the present study, the model of Mallet and Porter was used to evaluate pure stands of toxic plants, seed mixtures, refugia, and mixtures+refugia. Across a broad range of conditions, seed mixtures delayed development of insect resistance compared with pure stands of toxic plants. The relative merits of seed mixtures and refugia depend upon untested assumptions about pest movement, mating, and inheritance of resistance. Minimizing exposure of pests to toxins in space and time is the surest way to preserve the efficacy of transgenic