Concluding Remarks: Trophic Exploitation and Community Structure

Abstract

All the empirical contributions in this issue seem to agree about the generally exploitative nature of trophic interactions. This, indeed is an explicit premise of Holt's (1987), Hansson's (1987) and Fretwell's (1987) theories and also implicitly assumed in the contribution of Pimm and Kitching (1987). (Pimm's (1982) theory on stability problems of long food chains and food chains with much omnivory is based on exploitation models.) Moreover, exploitative trophic interactions seem to be essential for generating sustained multiannual population fluctuations (Stenseth and Oksanen 1987). Herbivores are indeed selective foragers (Danell et al. 1987, Ericson and Oksanen 1987) but plants seem nevertheless to be vulnerable to their consumers if not protected by predators, and phenol-based inducible defense does not seem to work reliably (Jonasson et al. 1986, Laine and Henttonen 1987, Oksanen et al. 1987). The idea of Pimm (e.g., 1982, see also Pimm and Kitching 1987) that primary productivity is practically always high enough to support three trophic levels is not contradicted by any contribution. Even barren Fennoscandian highlands and genuinely arctic tundras are invaded by predators during lemming peaks (especially by snowy owls and jaegers; see Andersson 1976, Cernjavskij and Tkacev 1982, Pitelka 1973), and one resident and a chiefly terrestrial predator the arctic fox is practically restricted to relatively barren tundra areas. On the other hand, there seems to be profound between-habitat differences in the intensity of predation (Hanski 1987, Henttonen et al. 1987) and herbivory (Oksanen and Ericson 1987). These differences might even explain a large part of the vegetational characteristics of different habitats (Ericson and Oksanen 1987).