Interaction of glucosinolate content of Arabidopsis thaliana mutant lines and feeding and oviposition by generalist and specialist lepidopterans.

The diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae), is an insect specialized on glucosinolate-containing Brassicaceae that uses glucosinolates in host-plant recognition. We used wild-type and mutants of Arabidopsis thaliana (L.) Heynh. (Brassicaceae) to investigate the interaction between plant glucosinolate and myrosinase content and herbivory by larvae of the generalist Helicoverpa armigera Hübner (Lepidoptera: Noctuidae) and the specialist P. xylostella. We also measured glucosinolate changes as a result of herbivory by these larvae to investigate whether herbivory and glucosinolate induction had an effect on oviposition preference by P. xylostella. Feeding by H. armigera and P. xylostella larvae was 2.1 and 2.5 times less, respectively, on apk1 apk2 plants (with almost no aliphatic glucosinolates) than on wild-type plants. However, there were no differences in feeding by H. armigera and P. xylostella larvae on wild-type, gsm1 (different concentrations of aliphatic glucosinolates compared to wild-type plants), and tgg1 tgg2 plants (lacking major myrosinases). Glucosinolate induction (up to twofold) as a result of herbivory occurred in some cases, depending on both the plant line and the herbivore. For H. armigera, induction, when observed, was noted mostly for indolic glucosinolates, while for P. xylostella, induction was observed in both aliphatic and indolic glucosinolates, but not in all plant lines. For H. armigera, glucosinolate induction, when observed, resulted in an increase of glucosinolate content, while for P. xylostella, induction resulted in both a decrease and an increase in glucosinolate content. Two-choice tests with wild-type and mutant plants were conducted with larvae and ovipositing moths. There were no significant differences in preference of larvae and ovipositing moths between wild-type and gsm1 mutants and between wild-type and tgg1 tgg2 mutants. However, both larvae and ovipositing moths preferred wild-type over apk1 apk2 mutants. Two-choice oviposition tests were also conducted with P. xylostella moths comparing undamaged plants to plants being attacked by larvae of either P. xylostella or H. armigera. Oviposition preference by P. xylostella was unaffected as a result of larval plant damage, even in the cases where herbivory resulted in glucosinolate induction.