Coevolution: a history of the macroevolutionary approach to studying host-parasite associations.

The year 1991 marked 100 yr of coevolution research. I have reviewed the first 90 yr of this history. Three chronological phases are apparent: recognition of predictable associations among hosts and their parasites; search for patterns of association and their underlying causes, emphasizing either correlated biogeographic patterns or correlated phylogenies; and development of objective and repeatable methodologies for reconstructing and interpreting these patterns of association. Von Ihering, an outspoken anti-Darwinian, was undoubtedly the first to recognize and make use of predictable host-parasite associations. Kellogg and Fahrenholz, however, had more profound influence on subsequent generations, but in different directions. Kellogg attempted to meld natural selection with speciation by isolation. He also considered host specificity a component of coevolution, important but variable. His work laid the foundation for future research concentrated on biogeographic interpretations of host-parasite relationships. This emphasis and Metcalf's failed attempts to provide adequate mechanisms for reconstructing phylogenies reduced the biogeographic approach to an empirical research program in the hands of Manter. Fahrenholz, on the other hand, exposed to a strong anti-Darwinian sentiment, emphasized the importance of strict host specificity. This led to Eichler's formulation of the first 3 coevolutionary rules and the conclusion that host specificity was not a component but the cause of coevolution and ultimately the tautology inherent in the phylogenetic approach. All had to rely on 1 assumption, that host and parasite phylogenies were reflected in the taxonomic hierarchy. Hennig criticized this assumption and provided a method whereby phylogenies are reconstructed independently. Brooks melded this new phylogenetic method (cladistics) with an equally new biogeographic method (vicariance biogeography), providing the foundations for the modern macroevolutionary approach to studying host-parasite associations.