Abstract.  1. To assess the risks of an insect‐resistant transgenic plant for non‐target arthropods, it is important to investigate the exposure of non‐target species to the transgene product. Exposure of predators in the field depends on the toxin levels in food sources, their feeding ecology and that of their prey.2. To verify the transmission of Cry1Ab toxin through the food chain, and thus exposure of predators in the field, samples from different plant tissues, herbivores, and predators in Bt maize fields in Spain (Event 176) were collected at different periods over the season and the toxin content was measured using ELISA. Complementary laboratory studies were performed with the omnivorous predator Orius majusculus to assess the toxin uptake and persistence after feeding on variable Bt‐containing food sources.3. Field results revealed that toxin content in some herbivores was negligible (aphids, thrips, leafhoppers) compared with those in spider mites. The latter herbivore only occurred after pollen shed and contained three times greater toxin levels than Bt maize leaves.4. Data confirmed that the Bt toxin can be transferred to predators, that is to say to Orius spp., Chrysoperla spp., and Stethorus sp. This only applied when Bt maize pollen or spider mites were available. The passage of Bt toxin to O. majusculus via these two food sources was also confirmed in the laboratory. Contrastingly, some predators in the field (hemerobiids, Nabis sp., Hippodamia sp., Demetrias sp.) contained no or negligible toxin levels even when pollen or spider mites were present.5. Besides essential information for exposure assessment of numerous arthropod predators, this study provides an insight into the feeding ecology of different arthropods in the maize system.

A major concern regarding the deployment of insect resistant transgenic plants is their potential impact on non-target organisms, in particular on beneficial arthropods such as predators. To assess the risks that transgenic plants pose to predators, various experimental testing systems can be used. When using tritrophic studies, it is important to verify the actual exposure of the predator, i.e., the presence of biologically active toxin in the herbivorous arthropod (prey). We therefore investigated the uptake of Cry1Ab toxin by larvae of the green lacewing (Chrysoperla carnea (Stephens); Neuroptera: Chrysopidae) after consuming two Bt maize-fed herbivores (Tetranychus urticae Koch; Acarina: Tetranychidae and Spodoptera littoralis (Boisduval); Lepidoptera: Noctuidae) by means of an immunological test (ELISA) and the activity of the Cry1Ab toxin following ingestion by the herbivores. Moreover, we compared the activity of Cry1Ab toxin produced by Bt maize to that of purified toxin obtained from transformed Escherichia coli, which is recommended to be used in toxicity studies. The activity of the toxin was assessed by performing feeding bioassays with larvae of the European corn borer (Ostrinia nubilalis (Hu¨bner); Lepidoptera: Crambidae), the target pest of Cry1Ab expressing maize. ELISA confirmed the ingestion of Bt toxin by C. carnea larvae when fed with either of the two prey species and feeding bioassays using the target pest showed that the biological activity of the Cry1Ab toxin is maintained after ingestion by both herbivore species. These findings are discussed in the context of previous risk assessment studies with C. carnea. The purified Cry1Ab protein was more toxic to O. nubilalis compared to the plant-derived Cry1Ab toxin when applied at equal concentrations according to ELISA measurements. Possible reasons for these findings are discussed.

As a part of a risk assessment procedure, the impact of Bt maize expressing Cry1Ab toxin on the thrips Frankliniella tenuicornis (Uzel) (Thysanoptera: Thripidae) was investigated, and the potential risks for predators feeding on thrips on Bt maize were evaluated. The effects of Bt maize on F. tenuicornis were assessed by measuring life-table parameters when reared on Bt and non-Bt maize. The content of Cry1Ab toxin in different stages of F. tenuicornis reared on Bt maize and the persistence of the toxin in adults where determined in order to evaluate the possible exposure of predators when feeding on thrips. In addition, Chrysoperla carnea (Stephens) (Neuroptera: Chrysopidae) was used as a model predator to assess how the behaviour of prey and predator may influence the exposure of a natural enemy to the Bt toxin. Life-table parameter results showed that F. tenuicornis was not affected when it was reared on Bt maize. This indicates that the potential for prey quality-mediated effects on predators is low. Bt content was highest in thrips larvae and adults, and negligible in the non-feeding prepupal and pupal stages. The persistence of the Cry1Ab toxin in adult F. tenuicornis was short, resulting in a decrease of 97% within the first 24 h. Predation success by young C. carnea larvae varied among the thrips stages, indicating that exposure of predators to Bt toxin can additionally depend on the prey stage. When combining the current knowledge of the susceptibility of major thrips predators with our findings showing no potential for prey quality-mediated effects, relatively low toxin content in thrips as well as short persistence, it can be concluded that the risks for predators when feeding on thrips in or next to Bt maize fields are negligible.