Phylogeny and generic delimitation in Molluginaceae, new pigment data in Caryophyllales, and the new family Corbichoniaceae

Abstract The circumscription of Molluginaceae has changed radically in recent years, with Corbichonia being moved to Lophiocarpaceae, Limeum to Limeaceae, Macarthuria to Macarthuriaceae and all species of Hypertelis , except the type, to Kewa in Kewaceae. In a broad analysis of core Caryophyllales using plastid trnK‐matK and rbcL sequences, the position of Molluginaceae in a strict sense as sister to the Portulacineae clade is corroborated, as are the positions of Corbichonia , Limeum and Kewa outside the family. The phylogeny of Molluginaceae is reconstructed based on trnK‐matK and nuclear ITS sequences of about half of the currently recognized species in the family and with representatives from all recognized genera. Mollugo is found to be polyphyletic and a new taxonomy for the family with 11 genera is proposed. Mollugo in its new restricted sense is a mainly American genus of about 15 species, including M. ulei comb. nov., previously placed in the monotypic Glischrothamnus . The Australian and Asian genus Trigastrotheca is resurrected for T. molluginea , T. pentaphylla comb. nov. and T. stricta comb. nov. The name Paramollugo nom. nov. is proposed for the Mollugo nudicaulis group and the combinations P. angustifolia comb. nov., P. cuneifolia comb. nov., P. decandra comb. nov., P. deltoidea comb. nov., P. navassensis comb. nov. and P. nudicaulis comb. nov. are made. Hypertelis is expanded to include, besides the type H. spergulacea , also H. cerviana comb. nov., H. fragilis comb. nov., H. umbellata comb. nov. and H. walteri comb . nov. In Pharnaceum , the new combination P. namaquense comb. nov. is made, Hypertelis longifolia is treated as a synonym of P. lineare and Mollugo tenella as a synonym of P. subtile . Corbichonia is proposed to be treated as a family of its own, Corbichoniaceae fam. nov. Several names are lectotypified, including the Linnaean Mollugo pentaphylla and M. stricta . An anthocyanin is reported for the first time from Simmondsiaceae. The detection of anthocyanins in members of Kewaceae and Molluginaceae agree with previous reports and corroborate the view that these families represent reversals from betalains to anthocyanins. The report of an anthocyanin in Limeaceae, previously regarded as unpigmented, apparently represents a newly detected reversal from betalains to anthocyanins in this family.