PREDICTING THE SPREAD OF INSECT RESISTANCE: A TOOL TO EVALUATE ECOLOGICAL RISK

Joy Bergelson and Juliette Winterer

Department of Ecology and Evolution, University of Chicago, Chicago, IL 60637, juliette@pondside.uchicago.edu

Mathematical models can make predictions about the ecological and evolutionary consequences of using pesticidal genes in crops. Such models may be useful tools in evaluating ecological risk. We are developing such a model to compare three strategies (pyramids, spatial mosaics, and rotation) for deploying two pesticidal genes. A successful strategy delays the rate of evolution of resistant pests and minimizes pest damage. The model predicts that when pesticidal genes have only mild expression in the crop, pyramids are the best use of the genes. When expression is strong, mosaics and rotations are more effective. To test the predictions of the model and evaluate its usefulness as a tool to assess ecological risk, we have engineered Brassica napus, an oilseed crop, to express two pesticidal genes: Cry 1A (c), for Bt-endotoxin expression, and potato proteinase inhibitor I. We will deploy the engineered Brassica in the three strategy arrangements,with appropriate controls, and allow diamondback moths (Plutella xylostella) to cycle through several generations on these treatments. We will measure moth population growth and test for changes in moth resistance to the crop over time.

Because pesticidal genes often come from natural communities, it is possible that pests that develop resistance to transgenically pesticidal crops will be able to exploit natural communities previously inaccessible to them. To assess whether resistant diamondback moths have an altered effect on natural plant dynamics, we will compare the ability of the susceptible and resistant moths to damag Brassica napus when it is in solitary stands and when it grows in competition with other plants. Differences in B. napus performance (growth and fruit set) between moth treatments indicate that resistant moths can alter plant community dynamics.