Cleistopetalum, H. Okada, 1996

Turner, I. M., 2010, A consideration of Cleistopetalum and a new combination in Polyalthia (Annonaceae), Phytotaxa 8, pp. 41-45 : 41-42

Species of Cleistopetalum

The species, Cleistopetalum borneense Okada (1996: 4) and C. sumatranum Okada (1996: 5) , were described as new and not transferred from other genera. Although it has not been possible for me to gain direct access to the type material of either species, by studying the original protologues, paratype material and other collections, and photographs of type specimens, I have been able to identify the two species of Cleistopetalum as taxa already published.

Cleistopetalum borneense was designated as the type species of the genus, so I will deal with it first. The collections reported as being C. borneense , both from East Kalimantan, are sufficiently similar to what was first described as Polyalthia macropoda King (1892: 60) from the Malay Peninsula to be considered conspecific. Given that C. borneense is the type of Cleistopetalum it is necessary to consider if the genus should be maintained or reduced to the synonymy of Polyalthia Blume (1830: 68) . Polyalthia is a large and distinctly unnatural genus of Old World Annonaceae that has been something of a dumping ground for nonclimbing species bearing relatively open flowers with little difference between the corolla whorls. Molecular analysis has provided strong evidence of the polyphyletic nature of Polyalthia sensu lato ( Mols et al. 2004). Polyalthia macropoda King , with its single-seeded monocarps, clearly belongs in Polyalthia section Monoon (Miq.) Benth. , though this is itself something of a mixed bag of species and still remains to be typified. Other genera, such as Enicosanthum Becc. and Woodiellantha , need to be considered as well as Monoon Miq. The molecular analysis of Mols et al. (2004) distinguished clade F4 which included species of Enicosanthum and Polyalthia section Monoon (though species of Cleistopetalum and Woodiellantha were not sampled). The characters emphasised by Okada (1996) in describing Cleistopetalum - cauliflory, relatively closed flowers with the petals remaining more or less erect at anthesis and the base of the inner petals pressed over the reproductive structures (more evident in C. borneense than C. sumatranum ) - are common in Enicosanthum , and present in Woodiellantha (distinguished by its basally connate petal whorls), as is the single basal ovule per carpel. Other species of Polyalthia section Monoon , such as Polyalthia lateriflora (Blume) Kurz (1874: 52) , have petals reflexed at anthesis. There is, as Mols et al. (2004) concluded, considerable variation in floral form in clade F4, but more consistent fruit morphology. Molecular analysis to ascertain the phylogenetic position of Cleistopetalum and Woodiellantha is required in order to improve resolution of the membership and any subdivisions of clade F4. Meanwhile, I consider Cleistopetalum to be a synonym of Polyalthia sensu lato.

Unfortunately P. macropoda King is a later homonym of P. macropoda (Miq.) Mueller (1877: 95) , a species from New Guinea. I therefore provided an avowed substitute Polyalthia sinclairiana Turner (2007: 275) . I was not aware at that time of the identity of Cleistopetalum borneense . As a synonym of Polyalthia macropoda King , C. borneense should provide the correct name for the species. However Polyalthia borneensis is already occupied by a different species ( P. borneensis Merrill, 1929: 65 ), so Cleistopetalum borneense cannot be transferred to Polyalthia and Polyalthia sinclairiana continues to be the correct name.