Linnaeosicyos (Cucurbitaceae): a New Genus for Trichosanthes amara, the Caribbean Sister Species of all Sicyeae

The Old World genus Trichosanthes has flowers with strikingly fringed petals, and Linnaeus therefore placed a species from Hispaniola that he only knew from an illustration (showing such fringed petals) in that genus. The species remained hidden from the attention of subsequent workers until acquiring new relevance in the context of molecular-biogeographic work on Cucurbitaceae. Based on molecular data, it is the sister to all Sicyeae, a New World clade of about 125 species in 16 genera. We here place this species in a new genus, Linnaeosicyos, describe and illustrate it, and discuss its phylogenetic context using molecular and morphological data. Judging from Dominican amber, elements of the flora of Hispaniola date back 15–20 my, and the occurrence on the island of at least five endemic species of Cucurbitaceae (Linnaeosicyos amara, Melothria domingensis, Sicana fragrans, and the sister species Anacaona sphaerica and Penelopeia suburceolata) points to its long occupation by Cucurbitaceae. Keywords—Flora of Hispaniola, fringed petals, lectotypification, Linnaeus, Plumier. With about 100 accepted species, Trichosanthes L. is the largest genus of the family Cucurbitaceae (Rugayah and De Wilde 1999; Huang et al. 2007). It is endemic in Asia and Australia, except for T. amara L., a species from the Caribbean island of Hispaniola in the Greater Antilles. Linnaeus (1753) described T. amara based on an illustration from Plumier’s Description des plantes de l’Amerique (1693, pp. 86–87) and placed it in the genus Trichosanthes because of its striking fringed petals, otherwise then only known from Trichosanthes. Wild species of Trichosanthes are restricted to eastern Asia, tropical Australia, and Fiji (Jeffrey 1980, 1990; Rugayah and De Wilde 1997, 1999), and Linnaeus only knew them from illustrations. Thus, he described T. cucumerina and T. nervifolia from plates in Rheede’s Hortus Indicus Malabricus (Jarvis 2007). These species have white flowers with deeply divided petals, similar to the fringed petals in the Hispaniola species. Of the latter, Linnaeus apparently only saw Plumier’s plate, a plate also cited by Ray (1704). Because T. amara L. was treated in very few floras (with the notable exception of Liogier 1986) and was never studied in a broader context, it was never formally excluded from Trichosanthes. The most recent monographer of the family, Cogniaux (1881), placed it under Species dubiae, and the broadest modern specialist on the family, Charles Jeffrey, suspected that it did not belong in Trichosanthes but could not decide in which genus it belonged (C. Jeffrey, St. Petersburg, personal communication, March