摘要：With the cukivation of Bt cotton, the produced insecticidal Cry proteins are ingested by herbivores and potentially transferred along the food chain to natural enemies, such as predators. In laboratory experiments with Bollgard II cotton, concentrations of CrylAc and Cry2Ab were measured in Lepidoptera larvae （Spodoptera littoralis, Hello- this virescens）, plant bugs （Euschistus heros）, aphids （Aphis gossypii）, whiteflies （Bemisia tabaci）, thrips （ Thrips tabaci, Frankliniella occidentalis）, and spider mites （Tetranychus urticae）. Tritrophic experiments were conducted with caterpillars ofS. littoralis as prey and larvae of ladybird beetles （Harmonia axyridis, Adalia bipunctata） and lacewings （Chrysop- erla carnea） as predators. Immunological measurements （ELISA） indicated that herbivores feeding on Bt cotton contained 5%-50% of the Bt protein concentrations in leaves except whiteflies and aphids, which contained no or only traces of Bt protein, and spider mites, which contained 7 times more Cry 1Ac than leaves. Similarly, predators contained 1%-30% of the Cry protein concentration in prey. For the nontarget risk assessment, this indicates that Bt protein concentrations decrease considerably from one trophic level to the next in the food web, except for spider mites that contain Bt protein concentrations higher than those measured in the leaves. Exposure of phloem sucking hemipterans is negligible.

摘要：In assessing an insect-resistant genetically engineered （IRGE） crop before its commercialization, researchers normally use so-called ＂Tier-1 assays＂ as the initial step to determine the effects of the crop on non-target organisms. In these tests, the insecticidal proteins （IPs） produced by the IRGEs are added to the diets of test organisms in the laboratory. Test organisms in such assays can be directly exposed to much higher concentrations of the test IPs than they would encounter in the field. The results of Tier-1 assays are thus more conservative than those generated in studies in which the organisms are exposed to the IPs by feeding on IRGE plant tissue or in the case of predators or parasites, by feeding on invertebrate prey or hosts that have fed on IRGE plant tissue. In this report, we consider three important factors that must be considered in Tier-1 assays： （i） methods for delivery of the IP to the test organisms; （ii） the need for and selection of compounds used as positive controls; and （iii） methods for monitoring the concentration, stability and bioactivity of the IP during the assay. We also analyze the existing data from Tier-1 assays regarding the toxicity of Bt Cry proteins to non-target arthropod species. The data indicate that the widely used Bt proteins have no direct toxicity to non-target organisms