Linnaeus's folly – phylogeny, evolution and classification of Sedum (Crassulaceae) and Crassulaceae subfamily Sempervivoideae

Abstract

Sedum, containing approximately 470 species, is by far the largest genus of Crassulaceae. Three decades of molecular phylogenetic work have provided evidence for the non-monophyly of Sedum and many more of the 30 genera of Crassulaceae subfam. Sempervivoideae. In this study, we present a broadly sampled and dated molecular phylogeny of Sempervivoideae including 80% of all infrageneric taxa described in Sedum as well as most other genera of the subfamily. We used sequences of one nuclear (ITS) and three plastid markers (matK, rps16, trnL-trnF). The five major lineages of Sempervivoideae (i.e., Telephium clade, Petrosedum clade, Sempervivum/Jovibarba, Aeonium clade, Leucosedum plus Acre clades) were resolved as successive sister to each other in the phylogenetic analysis of the plastid markers, while in the ITS phylogeny the Petrosedum clade is the closest relative of the Aeonium clade. Our dating analysis of ITS suggests that Sempervivoideae diversified rapidly throughout the Paleocene and Eocene, possibly in the area of the former Tethys and Paratethys archipelago. A biogeographic pattern emerges in which migration out of this ancestral area is linked to pronounced morphological evolution resulting in several distinct lineages recognized as segregate genera thought to be derived from Sedum. These segregate genera, however, have been defined on the basis of strongly homoplasious characters such as degree of petal fusion, petal colouration or flower merism. Moreover, all character states currently used for the delimitation of segregate genera seem to be homoplasious, and each of them can be found in at least one species of Sedum. Extensive literature work led to the conclusion that only few of the monophyletic clades found by us can be defined unambiguously by morphological characters. Mainly for these two reasons, we believe that combination of all 14 genera currently recognized in tribe Sedeae (= clades Leucosedum plus Acre) into Sedum might be the most stable solution of the “Sedum problem”. This new Sedum s.l. would then comprise approximately 755 species.