Sources of variation in herbivore preference: among-individual and past diet effects on amphipod host choice
- First Online:
- Cite this article as:
- Poore, A.G.B. & Hill, N.A. Mar Biol (2006) 149: 1403. doi:10.1007/s00227-006-0307-3
- 188 Downloads
Understanding which factors affect the feeding preferences of herbivores is essential for predicting the effects of herbivores on plant assemblages and the evolution of plant–herbivore interactions. Most studies of marine herbivory have focussed on the plant traits that determine preferences (especially secondary metabolites), while few studies have considered how preferences may vary among individual herbivores due to genetic or environmental sources of variation. Such intraspecific variation is essential for evolutionary change in preference behaviour and may alter the outcome of plant–herbivore interactions. In an abundant marine herbivore, we determined the relative importance of among-individual and environmental effects on preferences for three host algae of varying quality. Repeated preference assays were conducted with the amphipod Peramphithoe parmerong and three of its brown algal hosts: Sargassum linearifolium, S. vestitum and Padina crassa. We found no evidence that preference varied among individuals, thus constraining the ability of natural selection to promote increased specialisation on high-quality S. linearifolium. Most of the variation in preference occurred within individuals, with amphipod preferences strongly influenced by past diet. The increased tendency for amphipods to select alternate hosts to that on which they had been recently feeding indicates that amphipods are actively seeking mixed diets. Such a feeding strategy provides an explanation for the persistence of this herbivore on hosts in the field that support poor growth and survival if consumed alone. The effects of past diet indicate that herbivore preferences are a function of herbivore history in addition to plant traits and are likely to vary with the availability of algae in space and time.