Breaking and entering: predators invade the shelter of their prey and gain protection

Abstract

Many herbivorous arthropods construct shelters on their host plant that offer protection against natural enemies. This has resulted in selection on natural enemies to enter these shelters, where they can feed on prey that are inaccessible for competing predators and parasitoids. The spider mite Tetranychus evansi produces a shelter consisting of a dense web that is impenetrable for most predators; the only known natural enemy that can penetrate the web and can forage efficiently on this pest is Phytoseiulus longipes. We show that this predator preferentially foraged and oviposited in the web of its prey. Moreover, intraguild predation on juveniles of these predators was significantly higher outside this web and in the less dense web of a closely related prey species (T. urticae) than inside the web of T. evansi. Although the production of shelters by herbivores may be profitable at first, their adapted natural enemies may reap the benefit in the end.

Fig. 1

Preference of the predator Phytoseiulus longipes for web of its prey Tetranychus evansi. One half of a leaf disc contained eggs and web of T. evansi (grey bars), the other half contained eggs but no web (white bars). Predation is shown as the proportion of the total number of T. evansi eggs preyed on each disc half. Oviposition only occurred on the disc half with web. Asterisks indicate significant difference in proportion of prey eggs killed or predator eggs laid between leaf discs (GLM: ***, P < 0.001)

Fig. 2

Preference of the predator Phytoseiulus longipes for web and eggs of two spider mite species. One half of a leaf disc contained eggs of Tetranychus evansi (grey bars), the other half contained eggs of T. urticae (white bars). Top two bars both disc halves contained web of the spider mites (grey bars: web of T. evansi, white bars: web of T. urticae). Bottom two bars the spider-mite web was removed from both disc halves. Oviposition is expressed as the average proportion of all predator eggs on either disc half. Predation is shown as the proportion of the total number of eggs preyed on each disc half. Asterisks indicate significant difference in proportion of prey eggs killed or predator eggs laid between leaf discs (GLM: **, P < 0.01 and ***, P < 0.001), ns no significant difference

Fig. 3

a Intraguild predation of juveniles of Phytoseiulus longipes in the presence of an intraguild predator, an adult female P. macropilis. Entire discs contained eggs of either Tetranychus evansi or Tetranychus urticae and were either covered with web (grey bars) of the same species or without web (white bars). Shown are average numbers (+ s.e.) of juveniles of P. longipes preyed during 3 days. There was a significant difference in intraguild predation on leaf discs with and without web of T. evansi (***, P < 0.001), and no such difference on leaf discs with and without web of T. urticae (ns). There was a significant interaction between the presence of web and the species that produced the web. Without web (white bars), there was no difference in intraguild predation on leaves with eggs of T. evansi or T. urticae, with web, there was a significant difference. b Average egg production (+ s.e.) by the predator P. macropilis during the intraguild predation experiment per 3 days. Predators oviposited significantly more on leaf discs with eggs of T. urticae than on discs with T. evansi eggs (***, P < 0.001)