The perils of using host relationships in parasite taxonomy: phylogeny of the Degeeriella complex

The taxonomy of lice (Insecta: Phthiraptera) is often heavily influenced by host taxonomy. The use of host information to define genera of avian lice in the widespread Degeeriella complex has been prevalent but has created problems. Several workers have suggested that genera defined on the basis of host association are not monophyletic. We used sequences of nuclear (elongation factor-1alpha) and mitochondrial (cytochrome oxidase I) genes to test the monophyly of several genera in the Degeeriella complex. Parsimony and likelihood analyses of these data indicated that many genera in the Degeeriella complex are not monophyletic, such that species occurring on the same host groups do not form monophyletic groups. Biological features of hosts (including predaceous habits, brood parasitism, and hole nesting) for species in the Degeeriella complex likely provide opportunities for switching of lice between host groups. In addition, dispersal of lice via phoresy on hippoboscid flies also likely provides opportunities for host switching in the Degeeriella complex. This study indicates that the overuse of host taxonomy in louse taxonomy can result in classifications that do not reflect phylogenetic history.