INTRODUCTION

The genus Gloriosa L. was proposed by Linnaeus in 1737 and validated in 1753 (Linnaeus 1753). The type species for the genus is G. superba L., described from a specimen collected in southwest India (Malabar, present-day Kerala and part of Tamil Nadu). During the next [+ or -] 260 years, a further 40 species of Gloriosa were described (IPNI 2012), although the majority were subsequently reduced to synonymy. The most recently described species of Gloriosa is G. sessiliflora Nordal & Bingham (Nordal & Bingham 1998), a paper where the generic delimitation between Gloriosa and Littonia Hook. was first questioned. The connivent tepals of G. sessiliflora are similar to those of Littonia, although the colour, shape and undulation of the tepals strongly resemble those of some forms of G. superba (Nordal & Bingham 1998). The slightly bent style of G. sessiliflora also appears to be an intermediate trait.

The genus Littonia (Hooker 1853) differs from Gloriosa in its straight, not bent style and connivent, not reflexed tepals (Nordal & Bingham 1998), but there are many similarities between the two genera. Both have tuberous corms (Buxbaum 1937; Dyer 1976; Thulin 1995; Demissew 1997; Nordenstam 1998), their leaves frequently develop tendril-like tips (Queva 1899; Dyer 1976; Thulin 1995; Demissew 1997; Nordenstam 1998) and colchicine occurs in both (Hegnauer 1963; Wildman & Pursey 1968; Raffauf 1970; Vinnersten & Larsson 2010). Queva (1899) also noted that crystals of calcium oxalate were lacking in both Gloriosa and Littonia. The pistils of Gloriosa and Littonia are generally tricarpellate and alike (Sterling 1975). Because of such a series of resemblances, most investigators have been inclined to treat these genera as being closely related and have placed them in the same tribe (e.g. Krause 1930; Hutchinson 1934, 1959; Buxbaum 1936; Nordenstam 1982, 1998). Recent molecular phylogenetic studies on family Colchicaceae using three non-coding sequences from cpDNA retrieved a well-supported clade (100% jacknife support) in which Littonia species were nested among Gloriosa species (Vinnersten & Reeves 2003). Consequently, the genus Gloriosa has been expanded by including Littonia (Vinnersten & Manning 2007), rendering it monophyletic. The genus Gloriosa (including Littonia) is now classified as a member of the tribe Colchiceae (Vinnersten & Manning 2007).

This paper reports on the presence of G. sessiliflora in the Bie Province, central Angola. Gloriosa sessiliflora was described as endemic to Western Zambia (Nordal & Bingham 1998). The second known set of specimens (Bingham 12717) were collected in 2003...