Parasite manipulation: how extensive is the extended phenotype?

Abstract

The many ways in which the genes of one organism can have an extended phenotypic influence on the living body of another organism is no better illustrated than in the subtle and varied ways in which parasite genes find phenotypic expression in the behaviour of their hosts. For over 30 years, biologists have been fascinated by the adaptive manipulation of host behaviour by parasites, and the debate over its extent and magnitude shows no sign of diminishing, as illustrated by two recent papers.Although the existence of the manipulation of behaviour by parasites is not in any doubt, Robert Poulin1xManipulation of host behaviour by parasites: a weakening paradigm?. Poulin, R. Proc. R. Soc. London Ser. B. 2000; 267: 787–792Crossref | PubMedSee all References1 argues that the magnitude of its effect might have been overestimated by biologists, owing to publication biases that were prevalent in the early years of this developing paradigm. In support of his assertion, he shows that the magnitude of parasite-induced changes in behaviour has declined significantly since the first papers on the subject were published in the 1970s and early 1980s.Meanwhile, William Eberhard2xSpider manipulation by a wasp larva. Eberhard, W.G. Nature. 2000; 406: 255–256Crossref | PubMed | Scopus (81)See all References2 has added an important contribution to the debate, with a report in Nature on a fascinating discovery made in oil palm plantations in Costa Rica. He reports on an ichneumonid parasitoid (Hymenoepimecis sp.) that manipulates the behaviour of an orb-weaving spider (Plesiometa argyra) to increase its own survival chances during rainstorms. The wasp larva spends the first two weeks of its life feeding on the blood of its host, while the spider continues with its normal web-building activities. Then, on the night that it will finally kill its host, the larva induces the spider to build a web that is totally different from any web it has built previously. Instead of a typical orb web, the new construction comprises a dense central hub anchored by between two and eight robust guys or drag lines, making the web appear more like a canopy or awning. Once the spider has built this ‘cocoon web’, the wasp larva moults, kills and eats the spider, then spins the pupal cocoon in which it will hang by a line beneath the new web. Eberhard argues that the modifications to the spider’s web make it much stronger and more resistant to the impact of heavy rains, which are a significant mortality factor in a related wasp species.This instance of parasite manipulation of behaviour is notable for two reasons. First, unlike many of the examples analysed by Poulin, there is no question that the change in behaviour is due to the influence of the parasite and is detrimental to the host (although the relative advantage of the cocoon web over the orb web for wasp survival has yet to be explicitly demonstrated). Second, it might be possible to determine the precise mechanisms that are generating the behavioural change; a feat that has rarely, if ever, been achieved for such a complex manipulation. Already, it is clear that the behavioural change is mediated by a rapid-acting, long-lasting biochemical injected into the spider by the wasp. Thus, it should be possible to isolate and characterize the active substances. If this could be done, then light would be shed not only on the evolution of spider web-making behaviour, but also on the tools used by wasp genes to extend their phenotype within their spider hosts.