What to do with Hibiscus? A proposed nomenclatural resolution for a large and well known genus of Malvaceae and comments on paraphyly

The generic classification of Hibisceae has long been unstable. A new understanding of the phylogeny of Hibisceae has found that genera from three tribes (Decaschistieae, Hibisceae and Malvavisceae) are nested within Hibiscus. We discuss issues that impinge upon the classification of Hibiscus in a general sense, including the genus concept, monophyletic and paraphyletic taxa, the use of characters and phylogenies to define taxa, and the current ranked system (based on the International Code of Botanical Nomenclature) v. a rank free alternative (the PhyloCode). We conclude that genera are subjective constructs that are only real in the sense of their phylogenetic origins (i.e. the taxa they denote may be real discoverable monophyletic groups), that paraphyletic taxa are not necessary in classification, that overemphasis of 'distinctive' characters in preference to phylogenies leads to recognition of paraphyletic taxa, and that there is no objective fixed and discoverable rank for any taxon. While the PhyloCode avoids some of these problems, it fails to adequately provide an alternative to the mnemonically powerful and information rich Linnean binomial and its advantages do not, in our opinion, outweigh the utility of the ICBN system when the latter employs only monophyletic taxa. With these conclusions in mind, we offer a brief set of guidelines for higher level classification and apply this to Hibiscus. The severe paraphyly in Hibiscus means that no classification using the ICBN system with exclusively monophyletic taxa will be free of major nomenclatural changes. We argue that including over 200 species from several genera within a broadly defined Hibiscus causes fewer nomenclatural changes overall than do alternative schemes, while promoting stability and attempting to minimise change to well known species. A hybrid formal ranked and informal rank free system is discussed and proposed for this group. A series of rank free names that are nested within Hibiscus s.l. are proposed to convey information about membership of distinctive clades within Hibiscus s.l. in lieu of a complete ranked subgeneric classification that awaits more investigation.