Abstract

Coevolution in trophic interactions is often considered as a major factor underlying diversification in both interacting species. Empirical examples of tight cospeciation are however rare. We argue that this lack of examples may reflect reality, and one reason is that speciation through bottom-up and top-down processes are not symmetric, partly because enemies are able to select the best host whereas hosts are not able to choose their enemies. In the paper, we discuss how the oscillation hypothesis through network rewiring and geographic range shifts have different consequences for the two interacting partners, leading up to differences in their capacity for local adaptations. We also argue that these processes have been poorly explored in theoretical models that focus mainly on the gene dynamics and then overlook trait asymmetries.