In support of a broad concept of Paepalanthus (Eriocaulaceae)

Abstract

We here evaluate the conflicting recent splitter and lumper strategies concerning the generic classification system of the Eriocaulaceae, and whether Paepalanthus s.str. and its proposed segregate genera are morphologically recognizable. To this end, we examine evidence from inflorescence branching patterns, seed coat, and phylogeny in detail, and contrast circumscriptions of genera with the available data. We also indicate possible consequences of the proposed system for herbarium curators, keepers and users of large databases, and generalist users of identification keys. We argue that the morphological circumscriptions of Paepalanthus s.str. and of most of the other 11 segregate genera lack coherence. The majority of characters used as potential synapomorphies or in diagnoses are not sufficient to support the genera because they conflict with or do not represent the morphology of all the included species. The information regarding inflorescence branching patterns and seed coat are interpreted ambiguously or represent preliminary results, and some characters are known for only a limited number of species. Inflorescences are difficult to interpret in herbarium specimens, and the seed coat mostly requires SEM analysis for proper visualization, thus hindering identification. On the other hand, Paepalanthus s.l. can be identified by a combination of floral characters, as can the remaining genera of the family. We therefore endorse the recognition of Paepalanthus s.l. as a single, nomenclaturally stable, and monophyletic entity. This is the best-fitting solution to reflect both evolutionary and morphological aspects of this recently diversified lineage. Finally, we provide a new identification key for the 7 genera of the Eriocaulaceae, characterizing Paepalanthus s.l. by floral characters that can be observed in any flowering specimen.