Many true parasites and parasitoids modify the behaviour of their host, and these changes are thought to be to the benefit of the parasites. However, field tests of this hypothesis are scarce, and it is often unclear whether the host or the parasite profits from the behavioural changes, or even if parasitism is a cause or consequence of the behaviour. We show that braconid parasitoids (Glyptapanteles sp.) induce their caterpillar host (Thyrinteina leucocerae) to behave as a bodyguard of the parasitoid pupae. After parasitoid larvae exit from the host to pupate, the host stops feeding, remains close to the pupae, knocks off predators with violent head-swings, and dies before reaching adulthood. Unparasitized caterpillars do not show these behaviours. In the field, the presence of bodyguard hosts resulted in a two-fold reduction in mortality of parasitoid pupae. Hence, the behaviour appears to be parasitoid-induced and confers benefits exclusively to the parasitoid.

Adaptation to novel host plants is a much-studied process in arthropod herbivores, but not in their predators. This is surprising, considering the attention that has been given to the role of predators in host range expansion in herbivores; the enemy-free space hypothesis suggests that plants may be included in the host range of herbivores because of lower predation and parasitism rates on the novel host plants. This effect can only be important if natural enemies do not follow their prey to the novel host plant, at least not immediately, thus allowing the herbivores to adapt to the novel host plant. Hence, depending on the speed with which natural enemies follow their prey to a new host plant, enemy-free space on novel host plants may only exist for a limited period. This situation may presently be occurring in a system consisting of the herbivorous moth Thyrinteina arnobia Stoll (Lepidoptera: Geometridae) that attacks various species of Myrtaceae, such as guava ( Psidium guajava L.) and jaboticaba ( Myrciaria spp.), in Brazil. Since the introduction of eucalyptus (Myrtaceae) species into this country some 100 years ago, the moth has included this plant species in its host range and frequently causes outbreaks, a phenomenon that does not occur on the native host plant species. This suggests that the natural enemies that attack the herbivore on native species are not very effective on the novel host. We tested this hypothesis by studying the searching behaviour of one of the natural enemies, the omnivorous predatory bug Podisus nigrispinus (Dallas) (Heteroptera: Pentatomidae). When offered a choice between plants of the two species, the predators (originally collected in eucalyptus plantations) preferred guava to eucalyptus when both plant species were clean, infested with herbivores, or damaged by herbivores but with herbivores removed prior to the experiments. The bugs preferred herbivore-damaged to clean guava, and showed a slight preference for damaged to clean eucalyptus. These results may explain the lack of impact of predatory arthropods on herbivore populations on eucalyptus and suggests that eucalyptus may offer an enemy-free space for herbivores.

Most herbivorous arthropods are host specialists and the question is which mechanisms drive the evolution of such specialization. The theory of antagonistic pleiotropy suggests that there is a trade-off between adaptation of herbivores to a novel host and their native host. The mutation accumulation hypothesis proposes that herbivores on a novel host lose their adaptation to the native host through the accumulation of mutations with negligible effects on performance on the novel host. Experimental evidence for either of the two hypotheses is scarce. We compared the fitness of two sympatric moth strains from an introduced host and a native host. The strain from the novel host did not perform better on this host than the strain from the native host. The strain from the novel host performed less well on the native host than did the strain from the native host. Hence, selection on the novel host did not result in noticeable gain in performance, but adaptation to the native host was lost. These results are more readily explained by the mutation-accumulation hypothesis than by the trade-off hypothesis.

The ‘enemy‐free space’ hypothesis predicts that herbivorous insects can escape their natural enemies by switching to a novel host plant, with consequences for the evolution of host plant specialisation. However, if natural enemies follow herbivores to their novel host plants, enemy‐free space may only be temporary. We tested this by studying the colonisation of the introduced tree Eucalyptus grandis (Hill) Maiden (Myrtaceae) by insects in Brazil, where various species of herbivores have added eucalyptus to their host plant range, which consists of native myrtaceous species such as guava. Some herbivores, for example, Thyrinteina leucoceraea Ringe (Lepidoptera: Geometridae), cause outbreaks in eucalyptus plantations but not on guava, possibly because eucalyptus offers enemy‐free space. We sampled herbivores (mainly Lepidoptera species) and natural enemies on eucalyptus and guava and assessed parasitism of Lepidoptera larvae on both host plant species during ca. 2 years. Overall, predators were encountered more frequently on guava than on eucalyptus. In contrast, parasitoids were encountered equally and parasitism rates of Lepidoptera larvae were similar on both host plants. This indicates that herbivores may escape some enemies by moving to a novel host plant. However, this escape may be temporary and may vary with time. We argue that studying temporal and spatial patterns of enemy‐free space and the response of natural enemies to host use changes of their herbivorous prey is essential for understanding the role of natural enemies in the evolution of host plant use by herbivorous arthropods.