Optimal refuge strategies to fight pest resistance to GM crops
Résumé
We consider the use of a genetically modi ed crop to ght a pest population that feeds on the crop. We use an entomological model that captures the diversity of the pest population's gene pool, as well as the level of pest invasion itself. A refuge area is used as an instrument to control the evolution of the susceptibility of the pest's gene pool to the genetically modi ed crop. We characterize the refuge area that minimizes the sum of discounted costs related to the crop damage caused by the pest as well as the supplemental cost of the genetically modi ed crop. The model is calibrated for the use of Bt-corn to ght the European corn borer. Because of the linearity of the objective function, the optimal refuge consists of a bang-bang and a singular control. For the calibrated parameters, as well as reasonable variations of them, the bio-economic system tends to an interior steady state where the level of pest-susceptibility is renewable.
However, when the control is restricted to being constant over time, as is currently done in the United States, the system generally tends to a steady state where the susceptibility is completely exhausted. In that case, it takes very particular parameter constellations for the system to reach an interior steady state. We are able to assess,
for the calibrated model, the cost reduction attained by using a refuge area that varies over time instead of a time-invariant one.