Integrating molecular phylogenetic results into ant taxonomy (Hymenoptera: Formicidae)

Ward, P.S.

Abstract: The ongoing molecular revolution in ant systematics, of which Ross Crozier was an early proponent and practitioner, has led to remarkable progress in our understanding of ant phylogeny. In this review I consider the interplay between molecular and morphological evidence, and the integration of molecular phylogenetic results into ant classification. New phylogenetic findings indicate that most ant subfamilies and genera are monophyletic, but there are some significant exceptions, including the subfamily Cerapachyinae and several large and species-rich ant genera (Amblyopone, Aphaenogaster, Camponotus, Cerapachys, Messor and Pachycondyla, among others). These non-monophyletic groups are manifestations of two different phenomena: (1) convergence in morphology and (2) heterogeneity in rates of morphological evolution. Granivorous ants in the genus Messor are an example of the former: molecular phylogenetic analysis indicates that the New World and Old World Messor are separate lineages that independently evolved similar derived morphology. Heterogeneity in rates of evolution is exemplified by highly divergent army ants and certain other dorylomorphs – currently assigned to different genera, tribes and subfamilies – that are nested phylogenetically within a group of morphologically plesiomorphic species, most of which are placed in a single genus (Cerapachys). This presents a classificatory conundrum if we wish to maintain a Linnaean (i.e., ranked) classification system in which all named taxa are monophyletic and easily diagnosed. Nevertheless, despite the conceptual appeal of a rank-free classification (as embodied in the PhyloCode), there are practical advantages to maintaining a ranked phylogenetic taxonomy, at least for groups such as ants that are in relatively recent and species-rich branches of the tree of life. These benefits include explicit information about the inclusivity and exclusivity of clades, and identification (through binomial nomenclature) of those taxa that are considered to represent species. It is important to recognize that the assignment of family-group or genus-group rank to clades is arbitrary and that ant taxa of a given rank are not equivalent, except in a very approximate sense. A new modus operandi is emerging in ant taxonomy above the species level, wherein molecular (Dna sequence) data from multiple nuclear genes are used to generate a well supported phylogeny, and the resulting tree serves as a framework for evaluating the informativeness of morphological traits and for identifying (and naming) clades that can be diagnosed morphologically. In following this protocol an attempt should be made to maintain continuity in nomenclature and taxon concepts with the preexisting classification, to the extent that this is possible. A similar set of principles has been followed by plant systematists developing a ranked phylogenetic classification of flowering plants (Angiosperm Phylogeny Group system).

Myrmecol. News 15: 21-29; printable