UNCORRECTED PROOF © The Norwegian Academy of Science and Letters • Zoologica Scripta, 00, 0, January 2005, pp000–000 1 Z S C 1 9 6 Operator: Zhou rong Dispatch: 21.02.05 PE: Julian Staddon Journal Name Manuscript No. Proofreader: Ni xiuqiang No. of Pages: 3 Copy-editor: Jonathan Lewis Traditional (‘Linnaean’) definitions of taxon names ‘have types and no circumscriptions’ (Moore 1998: 565), and the contents of taxa thus defined are inherently unstable. For instance, the Linnaean taxon Hominidae contains the type species Homo sapiens and all other species that ‘deserve’ to be in the same family as the type. However, different people will have different interpretations of just how much divergence or variability is permitted within the Linnaean rank of ‘family’ (for instance), resulting in endless and inconclusive argu- ments over the limits of this family, even if there is complete agreement over phylogenetic relationships (e.g. Cantino et al. 1997; Laurin & Anderson 2004). In contrast, phylogenetic definitions tie taxon names to clades that are precisely circumscribed through a definition and phylogenetic hypothesis, and these well-defined taxon boundaries improve taxonomic stability (de Queiroz & Gauthier 1992; Cantino & de Queiroz 2003). However, since the boundaries of named taxa are defined with respect to an inferred phylogeny, instability in taxon composition within this system can occur with changes in hypothesized phylogenetic relationships. If the composition of taxa with phylogenetic- ally defined names is to be maximally stable, the boundaries of named taxa should be buffered against phylogenetic uncertainty. The choice of reference taxa can greatly influence the sta- bility of phylogenetic taxonomies. The three most common types of phylogenetic definitions use one or more reference taxa or ‘specifiers’ (analogous to Linnaean ‘types’) to delimit named clades. For instance, ‘the least inclusive clade contain- ing species A and species B’ (node-based), ‘the most inclusive clade containing species A but not species B’ (stem-based), and ‘the most inclusive clade exhibiting character state X synapomorphic (‘homologous’) with that in species A’ (apomorphy-based). Here, building upon previous suggestions (e.g. Schander & Thollesson 1995; Lee 1998; Sereno 1999; Cantino & de Queiroz 2003), I suggest how such instability can be mini- mized through choosing phylogenetically stable reference taxa. In particular, it is possible to quantify taxon stability using simple (but seldom-used) indices, and then choose only the most stable terminals to use as reference taxa in phylo- genetic definitions. Using this approach enables one to apply names only to stable clades in the best-supported ‘backbone’ subtree within a phylogeny. Unstable taxa can float in and out of these named clades causing only minimal nomenclatural disruption. The example discussed below employs node-based defini- tions and morphological data, but the conclusions apply to all types of phylogenetic definitions and all types of character data. Also, while example definitions below use one or two reference taxa, the suggestions apply to definitions that use more reference taxa (e.g. Cantino et al . 1997; Joyce et al . 2004). Consider the data set in Table 1 and the resultant phylogeny (Fig. 1). As will be clear, there is a strong signal (congruent characters 1–10) suggesting relationships among taxa A to E as follows: A(B(C(D(EF)))). The relationships of taxon W are less certain: the most parsimonious tree of 18 steps places it deeply nested within the ingroup as the sister to taxon F, on the basis of six characters (7–12). However, there are five con- Table 1 Example data matrix containing a highly unstable, ‘wildcard’ taxon (W). O = outgroup. itap0 Taxon Characters O 0 0 0 0 0 0 0 0 0 0 0 0 0 A 1 1 0 0 0 0 0 0 0 0 0 0 1 B 1 1 1 1 0 0 0 0 0 0 0 0 0 C 1 1 1 1 1 1 0 0 0 0 0 0 0 D 1 1 1 1 1 1 1 1 0 0 0 0 0 E 1 1 1 1 1 1 1 1 1 1 0 0 0 F 1 1 1 1 1 1 1 1 1 1 1 1 0 W 1 1 0 0 0 0 1 1 1 1 1 1 1 Blackwell Publishing, Ltd. Point of View Choosing reference taxa in phylogenetic nomenclature MICHAEL S. Y. LEE Lee M. S. Y. (2005). Choosing reference taxa in phylogenetic nomenclature — Zoologica Scripta, 00, 000–000. Michael S. Y. Lee, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia; and Earth Sciences Section South Australian Museum, North Tce, Adelaide 5000, Australia. E-mail: lee.mike@saugov.sa.gov.au Accepted: ?????? 2000 doi:10.1111/j.1463-6409.2005.00196.x 1