No abstract

With approximately 2000 species, tribe Cariceae (Cyperaceae) comprises a morphologically distinctive cosmopolitan clade, with holocentric chromosomes (N = 6 to 56), complex biogeographical patterns, and habitat diversity ranging from rainforests to deserts. Such a remarkable combination of characteristics should make Cariceae an ideal model for studying the evolution of biodiversity, although they also obscure their relationships in Cyperaceae, complicating attempts to identify the contributing factors to diversity of Cariceae. Recent molecular studies place Cariceae in a strongly supported clade consisting of tribes Dulichieae, Scirpeae s.s, and the enigmatic monotypic genus Khaosokia, although relationships in this clade are unresolved. Using the plastid genes matK and ndhF and a greatly improved taxonomic sampling covering 16 of 17 genera and 55% of the species outside Cariceae, our analyses firmly position Dulichieae and Khaosokia (79% and 85% bootstrap support) as successive sisters to a clade consisting of five major lineages (Calliscirpus, Trichophorum + Oreobolopsis + Cypringlea, Cariceae, Scirpus + Eriophorum, and Amphiscirpus + Phylloscirpus + Zameioscirpus), the first four of which receive good to strong support (> 80% bootstrap support). Cariceae are sister to the Trichophorum clade, although topological tests cannot exclude either Calliscirpus or a Scirpus clade + Zameioscirpus clade as sister to the tribe. Trichophorum appears to be paraphyletic and Eriophorum is firmly nested in Scirpus. There appears to be a trend in the increase of chromosome numbers in Scirpus and Eriophorum and a trend in the reduction and proliferation of the inflorescence throughout the major Cariceae-Dulichieae-Scirpeae clades.

In its broadest sense, Scirpus consists of a heterogeneous assemblage of up to 250 species, but modern circumscriptions suggest that only 40–50 species are part of the genus. Despite a narrower definition of the genus, atypical species continue to be segregated from Scirpus with a common pattern being the removal of Southern Hemisphere taxa to other genera and tribes. In South America, the morphology of remaining Scirpus species also suggests that they are not closely related to Scirpus s.str., but most of these taxa are only known from their types, making a detailed analysis of their generic affinities difficult. One notable exception is Scirpus asper, a species that is relatively common in the mountains and adjacent lowlands of Peru south to Argentina. Although this species possesses features used in the circumscription of Scirpus, such as cauline leaves, flat leaf blades and anthelate inflorescences, it is known to differ from Scirpus s.str. by its Schoenus‐type embryo, and most of its presumed allies are now placed in different genera (e.g., Scirpus analecti ≡ Cypringlea analecta; Scirpus giganteus ≡ Androtrichum giganteum). In this study, we use DNA sequence data from the plastid (matK, ndhF) and nuclear (ETS‐1f) genomes to demonstrate that Scirpus asper is not closely related to Scirpus s.str., but sister to Phylloscirpus within the predominantly South American Zameioscirpus clade (Amphiscirpus, Phylloscirpus, Zameioscirpus). When combined with morphological, anatomical and embryological data, results indicate that S. asper is best treated as the sole species of a new monotypic genus, Rhodoscirpus. The implications of these results on the taxonomy of tribe Scirpeae are discussed.

No abstract