Evolutionary consequences of species interactions

John Thompson [10] argues that species interactions must be an integral part of conservation efforts directed at maintaining communities that will be viable. Four properties of species interactions seem particularly important.

1. Many species are composed of populations specialized to different interactions. Even if the same two species are interacting across a broad geographical range, the details of that interaction (e.g., its genetic basis) may differ from population to population. Thus, minimizing the extinction of populations within species is important (compare [4,3]). Moreover, long-term persistence of the interaction may depend on time lags and asynchronous responses in different populations, as appears to be the case in some gene-for-gene interactions between plants and fungal pathogens.

2. Some interactions can evolve rapidly under changed ecological conditions. Host-pathogen interactions may evolve particularly rapidly,² but we also saw an example with the checkerspot butterfly in the Sierra Nevada. Populations occupying clear-cut meadows, which were created in the late 1960s, switched their food plant preferences from Pedicularis semibarbata and Castilleja disticha to Collinsia torreyi. The Hawaiian i'iwi provides a particularly striking example [9].
   • The i'iwi has a long, curved bill, presumably to match the long, curved corollas of lobeliod flowers that used to be common in the Hawaiian understory.

   • They now feed predominantly on ohia (Metrosideros polymorpha, Myrtaceae), which has relatively flat open flowers.

   • The upper mandible of contemporary i'iwis appears to be detectably shorter than that of i'iwis collected in the late 19th century.

3. There is no inherent directionality in how interactions will evolve. Because the genetic basis of interactions may differ from one population to the next and because the ecological details of the interaction are likely to differ from one population to the next, natural selection is likely to produce different results in populations that evolve independently of one another. Habitat fragmentation may contribute to evolutionary divergence among populations that were once similarly adapted.

4. The impacts of interaction often depend on the density of the interactors. The more frequently individuals encounter one another, the more intense their interactions. Pathogens tend to evolve greater virulence in dense populations with high rates of horizontal transmission, for example, than in sparse populations with high rates of vertical transmission.

As Thompson argues, these effects reinforce the importance of conserving large, relatively undisturbed tracts of land. Only over these broad expanses can we study and understand how species interactions evolve in relatively unmanipulated conditions. Only by understanding how they evolve in those circumstances can we discover how to manage them more effectively in human-dominated landscapes.