Failure-time Analyses of the Effectiveness of Larval Shield Defenses in Tortoise Beetles (Chrysomelidae: Cassidinae)1

Plant chemistry and predation are thought to be major factors responsible for the recurrent evolution of dietary specialization in herbivorous insects. However, their relative importance and the degree to which they interact to drive diet evolution remain unknown. The present study aims to test predictions of the ‘nasty host plant hypothesis’, which posits that an herbivore’s diet becomes more restricted as its reliance upon novel host plant compounds that confer protection from predators increases. The tortoise beetle larval shield system affords a unique opportunity to examine how predation and host plant chemistry interact. Shields can be micro-manipulated, including removal, chemical modification and reattachment, without harm to the larvae. We subjected larvae of different diet breadths produced from basal and derived hosts to a predation bioassay and compared the relative effectiveness of their shields under different treatment conditions. Failure-time analyses, the most appropriate statistical approach for right-censored temporal data, revealed that specialist larvae were consistently less susceptible to predation than were generalists feeding on the same plant. Although generalists were as competent as specialists at handling non-polar host chemistry, specialists were better at manipulating more polar host-derived compounds, which are more likely to include novel chemistry. Host shifts may be constrained to only those plants that possess novel, polar compounds. The interaction between plant chemistry and beetle diet evolution may be one of escalation driven by predation, wherein specialists are increasingly more effective than generalists in the assimilation of host plant polar compounds into shield defenses.