Nontarget effects of transgenic insecticidal crops: implications of source-sink population dynamics.

Widespread planting of transgenic insecticidal (TI) crops for pest control has raised concerns about potential harm to nontarget arthropods. Because the first generation of TI crops produce single Bacillus thuringiensis (Bt) toxins causing little or no harm to most nontarget arthropods, they are not likely to cause such negative effects. However, varieties of transgenic crops with multiple Bt toxins or novel toxins might be more harmful to nontarget arthropods. Field studies assessing nontarget effects typically compare the relative abundance of nontarget arthropods in TI crop fields to non-TI crop fields. However, for nontarget arthropods that are killed by TI crops, such analyses may miss important effects. Results from simulations of a spatially explicit population dynamics model show that large-scale planting of TI crops could cause three types of negative effects on nontarget arthropods that suffer mortality caused by TI crops: (1) lower abundance in TI fields than non-TI fields with little or no effect on abundance in non-TI fields, (2) lower abundance in TI fields than non-TI fields and decreased abundance in non-TI fields, and (3) loss of the arthropod from TI and non-TI fields. Simulation results show that factors increasing the potential for negative effects of TI crops on nontarget arthropods in non-TI fields are low reproduction, high emigration, high adoption of TI crops, high mortality in TI fields, insecticide sprays, and rotation of TI and non-TI fields. The results suggest that risk assessment should consider the regional distribution of transgenic crops and the life history traits of nontarget arthropods to identify the most vulnerable regions and nontarget species.