Phylogenetic relationships in the genus <i>Crinum</i> (Amaryllidaceae) with emphasis on tropical African species: evidence from <i>trnL-F</i> and nuclear ITS DNA sequence data

Fifty‐three species of Boophone Herb., Ammocharis Herb., Cybistetes Milne‐Redh. and Crinum L. were analysed cladistically using nuclear ITS and plastid trnL‐F sequences, with Amaryllis belladonna L. as an outgroup. Boophone disticha (L. f.) Herb. is sister to the subtribe Crininae (i.e., Crinum, Ammocharis, Cybistetes ). Two main clades were revealed in the Crininae. The first comprises Ammocharis, Crinum baumii Harms and Cybistetes longifolia Milne‐Redh. &amp; Schweick. Cybistetes longifolia appears in a sister relationship to Ammocharis angolensis (Baker) Milne‐Redh. &amp; Schweick. Crinum baumii and Cybistetes longifolia are thus both referred to the genus Ammocharis under the names Ammocharis baumii (Harms) Milne‐Redh. and Ammocharis longifolia (L.) Roem. respectively. The second main clade is constituted by all other examined species of Crinum , and is split into four subclades. The first subclade includes populations of the newly discovered Zambian species, C. jasonii Bjorå &amp; Nordal (2006), which has bell‐shaped flowers. The second subclade comprises all species with star‐shaped flowers (“ Stenaster ”) occurring in the southern and eastern part of Africa, Madagascar, and Australasia. The Angolan “ Codonocrinum ” (i.e., with bell‐shaped flowers), C. fimbriatulum Baker, is in a sister relationship to this subclade. The third subclade includes species with bell‐shaped flowers with main distribution in southern Africa, in addition to the Australian C. flaccidum Herb. The fourth subclade includes North African and tropical species with bell‐shaped flowers including Asiatic taxa. A monophyletic group with star‐shaped flowers distributed in West Africa and America is nested within this subclade. Interestingly, a narrowly endemic species, C. binghamii Nordal &amp; Kwembeya, occurring in swamps in western Zambia is sister to this “western Stenaster ” group. There is no support for the taxonomic recognition of subgeneric delimitation based on flower morphology. On the other hand, there are strong geographical and ecological trends in the phylogeny.