Stability, ranks, and the PhyloCode

Abstract

Current codes of biological nomenclature define taxon names using types and ranks: the type determines the mini− mal membership of a named taxon, and the rank is supposed to determine its limits. Homo is “the taxon including the type species Homo sapiens that is assigned to the rank of genus”. However, there is no “genus concept ” (analogous to a spe− cies concept), and thus no way of empirically determining the limits of a particular genus, even in the context of a single agreed phylogeny. The same problems also apply to higher taxa at all other ranks under current codes, leading to great taxonomic instability. All proposed objective criteria for de− termining membership of taxa at a particular rank (e.g., geological age, genetic divergence) are shown to be problem− atic. In contrast, the clades named by phylogenetic defini− tions are objective and stable. Node−based and branch− based definitions are most precise; however, apomorphy− based definitions can be ambiguous due to difficulty in de− fining alternative character states, and optimisation uncer− tainty. A major benefit of ranks (information about relative nesting of taxa) can be achieved even more efficiently using standardised but rankless suffixes already widely used in phylogenetic taxonomy. Finally, in situations where the phy− logeny is poorly known, phylogenetic nomenclature appears to be superior to the Linnean system. Phylogenetic nomen− clature does not force one to officially name poorly corrobo− rated groupings, whereas Linnean codes compel users to erect and name genera even when relevant supraspecific re− lationships are poorly known.