Making Carex monophyletic (Cyperaceae, tribe Cariceae): a new broader circumscription

Abstract

Carex (Cyperaceae), with an estimated 2000 species, nearly cosmopolitan distribution and broad range of habitats, is one of the largest angiosperm genera and the largest in the temperate zone. In this article, we provide argument and evidence for a broader circumscription of Carex to add all species currently classified in Cymophyllus (monotypic), Kobresia (c. 60 species), Schoenoxiphium (c. 15 species) and Uncinia (c. 70 species) to those currently classified as Carex. Carex and these genera comprise tribe Cariceae (subfamily Cyperoideae, Cyperaceae) and form a wellsupported monophyletic group in all molecular phylogenetic studies to date. Carex as defined here in the broad sense currently comprises at least four clades. Three are strongly supported (Siderostictae, core Vignea and core Carex), whereas the caricoid clade, which includes all the segregate genera, receives only weak to moderate support. The caricoid clade is most commonly split into two clades, one including a monophyletic Schoenoxiphium and two small clades of species of Carex s.s., and the other comprising Kobresia, Uncinia and mostly unispicate species of Carex s.s. Morphological variation is high in all but the Vignea clade, making it extremely difficult to define consistent synapomorphies for most clades. However, Carex s.l. as newly circumscribed here is clearly differentiated from the sister groups in tribe Scirpeae by the transition from bisexual flowers with a bristle perianth in the sister group to unisexual flowers without a perianth in Carex. The naked female flowers of Carex s.l. are at least partially enclosed in a flask-shaped prophyll, termed a perigynium. Carex s.s. is not only by far the largest genus in the group, but also the earliest published name. As a result, only 72 new combinations and 58 replacement names are required to treat all of tribe Cariceae as a single genus Carex. We present the required transfers here, with synonymy, and we argue that this broader monophyletic circumscription of Carex reflects the close evolutionary relationships in the group and serves the goal of nomenclatural stability better than other possible treatments.