|
Within the Schoeneae (Cyperaceae), the relationships and limits of 'Carpha' have been unclear. Further, some species limits within 'Carpha' appeared to be uncertain. There are conflicting interpretations of spikelet structure in Schoeneae and 'Carpha'. This study examines spikelet structure in Schoeneae, determines species and generic limits of 'Carpha', and estimates phylogeny in and around 'Carpha'. Spikelet morphology of 250 specimens of 47 species of Schoeneae was examined using light microscopy and scanning electron microscopy. The study confirmed that spikelet structure in Schoeneae is cymose with a sympodial 'rachilla'. Monopodial spikelets, as described by most of the current literature, were not found in Schoeneae. The definitions of spikelet and rachilla are revised; problems of previous interpretations of spikelet structure in Schoeneae are discussed and clarified. The study highlights the need to reinvestigate spikelet structure in other tribes of Cyperaceae. Phenetic analyses of morphology were undertaken to identify and test limits of species of 'Carpha': 16 species in 'Carpha sensu lato' (i.e. 'Carpha sensu stricto' and 'Asterochaete') were identified. Three, 'Asterochaete acuminata', 'A. discolor' and 'A. ulugurensis', are new species; the taxonomic ranks of 'A. angustissima', 'A. capitellata' var. 'bracteosa', 'Carpha nivicola' and 'C. schoenoides' are clarified; 'C. perrieri' is synonymised with 'A. capitellata'; and the definitions of 'A. schlechteri' and 'A. glomerata' are amended. Cladistic analyses were undertaken on morphological and molecular (the 'trn'L intron and 'trn'L-'trn'F intergenic spacer sequence) data to estimate phylogenetic relationships. On the basis of these analyses, 'Carpha sensu stricto' and 'Asterochaete' are reinstated. Within 'Carpha sensu lato', phylogenetic relationships of the species are fully resolved, but with variable support. Phylogenetic relationships are suggested for 'Carpha' and its relatives. The analyses also indicate some problems with the limits of 'Schoeneae' and the status of 'Costularia', 'Schoenus' and 'Tricostularia'; support separation of 'Capeobolus' from 'Costularia' and 'Tetraria'; and support the inclusion of Schoenoides back in 'Oreobolus'. Lack of support for some clades indicates additional sources of data are needed to corroborate these relationships. |
|