Nothing in Taxonomy Makes Sense Except in the Light of Evolution: Examples from the Classification ofRanunculus1

Abstract

Biological classification aims at establishing ordering systems for organisms. Principles of classification, however, differ in their criteria and in their information content. Cladistic classification emphasizes information on descent, but the strict application of logical inclusiveness leads in practice to disregard of modification and to a lack of information on evolutionarily relevant features. Phenetics provides information on similarity regardless of descent. Evolutionary classification maximizes information on evolution by combining information on descent and modification, but it relaxes the requirement of inclusiveness. In practice, this means accepting holo- and paraphyletic taxa, but rejecting polyphyletic groups. Review of a recently published case study of the species-rich and cosmopolitan genus Ranunculus demonstrates how evolutionary classification can be performed in practice. A hypothesis of descent was reconstructed by phylogenetic analysis of DNA sequence markers plus morphological and karyological characters. Based on this backbone phylogeny, information content on morphology, karyology, and ecology was used as a criterion for delimitation of infrageneric taxa. This concept resulted in the subdivision of a more basal, paraphyletic Ranunculus subg. Auricomus and the holophyletic Ranunculus subg. Ranunculus. On a sectional level, 14 holophyletic and two paraphyletic sections plus one monotypic section were classified. Holophyletic sections mostly reflect extinction gaps, while paraphyletic groups appear in clades that have reticulate evolution and/or ecological shifts. Classification of paraphyletic and monotypic sections preserves information on morphology, ecology, and evolutionary processes. This pluralistic approach is justified as it best reflects the diversity of the genus. The principle of broadening criteria maximizes information on descent and modification. Evolutionary classification facilitates practicability and stability of taxonomic work, as the broadening of criteria restricts the number of equally valid options for classification. For users, preserving information content on phenotypes aids practicability, because the connection to traditional literature and to modern information systems is optimally maintained.